# High Variance – All Transcripts

This file contains all transcripts from the High Variance podcast. Paste it into your LLM of choice to ask questions across episodes, search for themes, or explore what guests had to say.

High Variance is an interview podcast about a world that has become harder to read — more uncertain, more volatile, stranger. Host Danny Buerkli speaks with public intellectuals, entrepreneurs, and technologists to ask what is going on and how we should respond.

- [Apple Podcasts](https://podcasts.apple.com/us/podcast/high-variance-with-danny-buerkli/id1839231559)
- [Spotify](https://open.spotify.com/show/19vjjI6amjfur4u0i6UeLh?si=9c6f274910aa43fe)

## About the host


*Danny Buerkli*

I'm a Director at Windfall Trust, a new organization focused on the economic implications of transformative AI, and advise organisations on frontier strategy.

I also host [High Variance](https://www.dannybuerkli.com/podcast), a long-form interview podcast, which you should subscribe to.

Windfall is the third venture I'm building. Previously I co-founded [staatslabor](http://www.staatslabor.ch), a state capacity consultancy, and the [Centre for Public Impact](https://www.centreforpublicimpact.org/). Before that I was at BCG.

I'm currently based in Zurich and spent time at Stanford, in Berlin, London, and Geneva.

Email me at *dannybuerkli at gmail dot com* (I like receiving email!) or find me on [LinkedIn](https://www.linkedin.com/in/danny-buerkli-a5216326/).

## Table of contents

- **The Quest for the Perfect Weapon – with Jeff Stern** (2026-03-25): Jeff Stern — reporter and author of The Warhead — joins Danny Buerkli to talk about the long quest to build the “perfect weapon.” They discuss how a company mostly known for calculators revolutionized warfare, the paradox of precision-guided munitions, what disappears behind the language of “surgical” force, and how covering conflict changed Jeff’s understanding of his own role as a reporter.
- **Intelligence Saturation and the Economics of AI – with Ioana Marinescu** (2026-03-04): Ioana Marinescu — associate professor at the University of Pennsylvania, research associate at the NBER, and member of Anthropic's Economic Advisory Council — joins Danny Buerkli to discuss her 'intelligence saturation' paper, which divides the economy into an intelligence sector and a physical sector to model what happens as AI automates cognitive work. They discuss why a recent scenario piece that moved markets gets the economics wrong, what parameters to watch to understand where we're headed, why being first to AGI may matter less than people think, her policy proposals for AI Adjustment Insurance and a Digital Dividend, what UBI experiments do and don't tell us about a world without work, Ioana's favorite theory of divorce, and what it feels like to be an intelligence worker watching AI get better at your job.
- **France, Germany, and the State of Europe – with Joseph de Weck** (2026-02-15): Germany and France have historically formed the core of Europe and the European Union. Joseph de Weck - political analyst, historian, author, columnist, and Europe Director at Greenmantle - knows both exceptionally well and joins Danny Buerkli to unpack the state of Europe. They discuss the influence of Paul Ricœur on Emmanuel Macron, why French baguette bread is so standardized, what may happen at the German elections in 2029, and the big shifts under way in Europe's geopolitical posture.
- **Vienna: How the City of Ideas Created the Modern World – with Richard Cockett** (2026-01-21): Every aspect of modernity was shaped by individuals with intellectual roots in Vienna, argues Richard Cockett, author of 'Vienna: How the City of Ideas Created the Modern World'. Richard, historian and senior editor at The Economist, joins Danny Buerkli to discuss his latest book. They cover the story of the stunningly productive and creative Viennese emigrés, discuss implications for progress today and trace how the intellectual feud between Vienna and Frankfurt still reverberates today.
- **Labor Market Impacts of AI – with Bharat Chandar** (2025-12-29): Which effects of AI are we already seeing in the labor market? And what might be coming down the line? Bharat Chandar, postdoc at Stanford and co-author of the 'Canaries in the Coal Mine' paper, joins Danny Buerkli to discuss what we know about the impacts of AI on the labor market and where the jury is still out.
- **Improv Wisdom – with Patricia Ryan Madson** (2025-12-10): Patricia Ryan Madson, professor emerita at Stanford and author of 'Improv Wisdom', joins Danny Buerkli to talk about how she got into improv, how she starts a class, how status works, Keith Johnstone's dark side, and the four A's of improv: attention, acceptance, appreciation, and action.
- **Scenario Planning – with Jamais Cascio** (2025-11-26): Jamais Cascio, futurist, scenario expert, and author of Navigating the Age of Chaos, joins Danny Buerkli for a deep dive into scenario planning. They discuss how the discipline has evolved since the days of Herman Kahn at RAND and Pierre Wack at Shell, whether the military or the private sector do it better, why geoengineering might lead to predictable trouble (and why we might do it anyway), and whether today's AI is more or less weird than Jamais once imagined. Jamais also reflects on his time working with Ken Waltz and shares the story behind his BANI framework, which captures how many now perceive the world: brittle, anxious, nonlinear, and incomprehensible.
- **Building an Open LLM – with Antoine Bosselut** (2025-11-05): Antoine Bosselut, Assistant Professor at the Swiss Federal Institute of Technology (EPFL) in Lausanne, joins Danny Buerkli to explain how he and his team built Apertus, the 'open' LLM. Antoine and Danny discuss why taxpayers should fund this work, which constraints bite hardest when creating an LLM outside one of the large labs, and which public investments may be needed now.
- **The Art of Facilitation – with Vishal Jodhani** (2025-10-22): Vishal Jodhani, a master facilitator, joins Danny Buerkli to talk about what makes facilitation work. They discuss what makes for a good question, how to know the difference between productive chaos and unproductive confusion, and what is underappreciated about the Berlin club scene.
- **State Capacity and Government Reform – with Don Kettl** (2025-10-08): Don Kettl — prolific scholar of public administration — joins Danny Buerkli to talk about state capacity and government reform. They discuss what DOGE got right (and what it didn't), whether gradual government is possible at all, why Operation Warp Speed was so unreasonably effective, and what lessons we should learn from Paul Volcker.

---

## The Quest for the Perfect Weapon – with Jeff Stern

2026-03-25 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/the-quest-for-the-perfect-weapon-with-jeff-stern/id1839231559?i=1000757196219) | [Spotify](https://open.spotify.com/episode/7jexIYqPyCr0wtMt47m7s0?si=82454b8907f043bd)

Jeff Stern — reporter and author of The Warhead — joins Danny Buerkli to talk about the long quest to build the “perfect weapon.” They discuss how a company mostly known for calculators revolutionized warfare, the paradox of precision-guided munitions, what disappears behind the language of “surgical” force, and how covering conflict changed Jeff’s understanding of his own role as a reporter.

**Danny Buerkli:** My guest today is Jeff Stern. Jeff is the author of five books. His most recent is The Warhead, The Quest to Build the Perfect Weapon in the Age of Modern Warfare. Jeff is also a reporter. His work on the Ebola outbreak, the death penalty, Yemen, and much more has appeared in places such as Vanity Fair, The New York Times Magazine, and The Atlantic.

Jeff has also written and produced movies, and his second book, The Fifteen Seventeen to Paris, was turned into a motion picture directed by no less than Clint Eastwood — and I could go on. Jeff and I also went to grad school together, so this is extra delightful. Jeff, welcome.

**Jeff Stern:** Thank you, Danny. Great to see you, and thanks for having me.

**Danny:** Jeff, in your latest book, The Warhead, you open with this cockamamie story of Joe Kennedy Junior's fatal mission in 1944 — a first heroic attempt at precision-guiding a bomb, as it were. Tell us that story.

**Jeff:** It's a little-known story about Joe Kennedy, who was supposed to be the scion of the Kennedy dynasty. He was supposed to be the first Catholic president and all that. Towards the tail end of World War II, he volunteered for what was known to be a really dangerous mission: to try to take out one of the newer vengeance weapons that Hitler was developing in Nazi-occupied France. It was an underground bunker with a weapon from hell that was considered capable, if it got up and running, of obliterating London.

The Allies couldn't figure out how to get enough firepower close enough to strike this bunker and still get back safely. So as they were trying to figure out how to destroy it, they broke down the problem to its elemental form: if the problem is we can't get enough firepower close enough and still return safely, what if we remove "return safely" from the equation? From there came this idea to outfit war-weary bombers with torpedo explosives. They packed these things with 30,000 pounds of explosives.

They figured out a way to rig rods and levers to control and steer the plane, and then, using what was still pretty novel technology — television — mounted a camera on the front and figured out how to fly this giant plane remotely. So they had a steerable bomb. The problem was that taking off was still too complicated a maneuver to do with rudimentary rods and levers. The solution was to have a pilot and a copilot take off, and then once the plane was airborne and being controlled by the mothership, they would parachute out.

They tested it a little — you're parachuting at something like 200 miles per hour, which you and I both have some experience with parachuting out of planes, but not going anywhere near that fast. What happened is that on the first mission, the plane detonated — this giant bomb went off before they could parachute out.

I set that up as the prologue for a few reasons. One is that it's just dramatic. Two, starting a book with a Kennedy is a little bit of a cheat. But the real reason is that the arc of the story gets into the unintended consequences of developing what seems like a really perfect weapon. I didn't want it to feel like an anti-war book. I didn't want your first confrontation with this thing to be about how terrible it is or how evil the people who built it are — which I don't think is true. What better, more noble, more unambiguous mission than trying to destroy a terrible weapon that was going to kill a lot of civilians during World War II? I wanted you to be leaning forward and rooting for a better weapon to be invented.

**Danny:** And then you go on to portray a number of characters. One of the main protagonists is Weldon Word, the Texas Instruments engineer who goes on to invent Paveway, the first precision-guided munition. He's a very sympathetic character in the book. You really made me want to hang out with the guy — he's absolutely brilliant and interesting. And I don't think that's wrong. But surely that was a conscious choice.

**Jeff:** A conscious choice, but also — the people I spoke to who knew him adored him. He was a very selfless, ingenious, personable guy who was not interested in glory or taking credit. He was really interested in promoting the people around him. Very quiet, but apparently had a very loud, booming voice, so when he did decide to speak up, people listened.

It was convenient that he seems like such a humane, likable person, because I think there's a temptation, when something causes harm, to simply categorize the people responsible as evil villains. That's rarely the case. Very few people who develop weapons are thinking, "I hope this kills more civilians." In his case, he was initially animated by the risk both to civilians and to American pilots who didn't have a tool they could use for the job they were trying to do in Vietnam. A lot of them were getting shot down, brought into detention, or killed. What he was trying to do was invent something that would both save the lives of pilots and cause less collateral damage on the ground.

**Danny:** You mentioned the war in Vietnam, and there's another story in the book that forms the genesis of the quest for precision-guided ammunition — a bridge called the Dragon's Jaw. What's that story about?

**Jeff:** The Dragon's Jaw Bridge was a really strategically important bridge that connected the North and South, and it was one of the ways the Viet Cong were able to move men and materials southward. It took on an almost mythical reputation because it was one of the important targets the Americans were trying to destroy. It was a steel suspension bridge, so very strong. It's also a bridge — narrow, hard to strike. And it was very well defended. It became the locus of morale for both sides.

A lot of American pilots were sent to try to bomb it. There were other attempts at guided munitions that didn't work well. The principal ways they would bomb were either to drop a ton of bombs from very high or, if they were trying to aim, to get very low, which brought them into range of all sorts of antiaircraft munitions. It was that bridge that really provided the impetus to finally try to solve the precision bombing problem.

**Danny:** So what did Weldon Word figure out?

**Jeff:** At the time, one of the weapon systems that the Army had some interest in was the laser. Einstein had theorized about it, and it had appeared in science fiction comics — this idea that if you could focus a powerful enough light beam, you could use it to destroy something. The Army, along with Martin Marietta, one of the defense contractors, developed some preliminary lasers and found that making them man-portable required so much energy that it was impractical. You couldn't do significant damage.

But Weldon had that in the back of his head and was thinking: what if we use this new laser not to destroy something, but essentially to point at something? He combined that idea with basically off-the-shelf components to design the first prototypes of a bomb with a little photon sensor on the front. It just said: the photons are hitting this side, steer to correct; photons are hitting that side, steer the other way. You would aim a laser at the target, and the bomb would follow the laser reflection. Other attempts at guided bombs had used TV footage to steer, or required steering with a joystick while in the plane. This one was really easy to use — you just had to keep a laser pointing at the target.

It was also really inexpensive, which was partly a function of the fact that Texas Instruments was a sort of upstart electronics company. No one really took them seriously, and they couldn't get the resources other defense contractors had. As they developed it, one of the people who came on board, a Navy veteran, kept saying: we have to keep this thing simple. It could be the best bomb ever, but if it's complicated to use, the real decision-makers — not the generals and colonels, but the ground crews and the pilots — aren't going to want to use it. They were able to develop a weapon that was very easy to use, very inexpensive, and very effective.

**Danny:** What's fascinating about Weldon as a person, and you chronicle this in the book, is that not only does he come up with this genius series of insights and actually make it work, but over the years and decades he does two things: he keeps developing the bomb and making it more sophisticated — and of course this is a dynamic system, so the people on the receiving end adapt in response. Something that struck me, if you abstract away a bit from what we're talking about, is just his apparent incredible ability to see into the future and keep repeating this trick. What accounts for his ability to do that?

**Jeff:** I don't know. I think that's a really good question. It may be that he had that stereotypical engineer brain, looking at the world as a series of problems and how to solve them. His outlook is consistent with the lore of Texas Instruments. The company that became Texas Instruments was founded as an oil prospecting company — Geophysical Services International. Their main product was a magnetic anomaly detector, a ball you held out of a car that would detect magnetic anomalies in the ground indicating the presence of hydrocarbons. But they incorporated the day before the Japanese attacked Pearl Harbor, and oil was basically nationalized. They founded a company that became obsolete the very next day.

So in company lore it was: we need to be looking for tomorrow's market, not today's. We're willing to take a bet on something daring because we want to be in good shape for whatever comes down the pike. I don't know whether Weldon internalized that at the company, since he came on board much later. But he personifies that ethos. He was constantly looking at different trends and seeing how they might converge in ten or twenty years.

**Danny:** Tell me more about the genesis of the book. You've done an incredible amount of reporting from war zones, Afghanistan and Yemen included. You had an incredible piece where you traced a piece of shrapnel from someone's face in Yemen back to the unionized shop in the US where the ammunition had been assembled. Was that the nucleus of the idea for the book?

**Jeff:** Good catch, Danny. There were a few things. The most superficial was that in reporting the Yemen story, I found out the bomb used in the strike I was writing about was made by Raytheon but had been invented by Texas Instruments. Like everyone else, I just think of the calculator when I hear Texas Instruments. So this idea that the bomb was invented by the calculator company was a revelation.

The other thing was that what I was attempting to do with that story was collapse the distance a little bit. We're launching strikes all over the world now. It's very easy for one or two people in the Oval Office to say "let's attack there" — we've got precision weapons on planes or long-range missiles, you flip a switch, and something's exploding somewhere. It's also very easy to view the people we're attacking as two-dimensional, black-and-white figures on a screen, infrared output. But the thing is still exploding somewhere, and there are real people and real lives and real communities being destroyed.

I was trying to find a way to use this weapon — which symbolizes that remove — actually as a tool to bring people together, even if only as a literary device. Telling the history of how a calculator company revolutionized warfare, but also using this device of disconnection to connect a series of stories of people on every end of war over the past sixty years.

**Danny:** You've just touched on the central tension the book revolves around: are precision-guided munitions actually more humane because they let us do things surgically and precisely, or do they just make it easier to wage war? Each generation of the Paveway and similar weapons seems to expand the political appetite for using force.

**Jeff:** That is one of the hypotheses of the book — whether, if you have access to a weapon that is ostensibly humane enough, inexpensive enough, and precise enough, it begins to argue for its own use. One of the anecdotes I cite a lot is how, pretty quickly in the second Gulf War — the American invasion of Iraq, which was called the most precise air war in history, and by the proportion of precision munitions used, that is accurate — within six months or a year, more civilians had been killed in Iraq than in Hiroshima and Nagasaki combined. The statistics vary depending on who's counting, but you have the nuclear weapon, which is the exact opposite of a precision weapon, and the precision weapon gives us the political cover to do so many strikes that we end up causing way more damage than we would if we were deciding whether to level a city with a nuclear bomb.

**Danny:** One of the counter-questions would be: what about the dog that didn't bark? What about the wars that were not fought because of the dissuasive effect of those weapons?

**Jeff:** Do you mean nuclear weapons or precision weapons?

**Danny:** Precision-guided munitions.

**Jeff:** Another thing to think about is that if nuclear weapons set the terms of the Cold War, it was precision weapons with which we actually fought it and were allowed to fight it. There was this idea of "we can't escalate, we need to be careful because this could lead to nuclear exchange." Well, if you have the opposite of a nuclear weapon — a very precise, small weapon — then you can still go and meddle and intervene, ostensibly without as much risk of escalating. It's another way having this weapon gives permission to engage when we might not otherwise have.

**Danny:** Given that they exist and it does seem difficult to put the genie back in the bottle, what is the correct way of using these weapons?

**Jeff:** With restraint. There probably are cases where being able to strike precisely is a net benefit. But it requires a lot of restraint that we are not demonstrating right now, because there is this illusion that we can fight wars very cleanly and surgically. One of the ways this manifests — and I think we're seeing this now in Iran — is that yes, a lot of these weapons are astoundingly precise. However, even the most precise weapon is only as precise as the intelligence on the ground. What often happens is that aspect gets forgotten or ignored, and we fall into this almost intoxicating idea that the weapon is so precise it will necessarily hit what it's supposed to hit.

In a place like Iran, where we don't have a lot of intelligence on the ground — no embassy, not as much political reporting, very few spies who can go there under diplomatic cover — very early on, the day the strikes started, we saw this horrible thing where an elementary school was bombed. Even if you have a weapon that can go into a car window and kill just a terrorist leader and no one around him, you have to know that that's a terrorist leader. Even though everyone knows logically that a bomb isn't evaluating a person's biographical details, there does seem to be this thing where the hardware capability almost convinces us that we don't need the human intelligence.

**Danny:** Does that make you a technological determinist?

**Jeff:** I guess so. Because I really do think the trigger can pull the finger. If you have the thing, you kind of want to use it.

**Danny:** The thing I might push back slightly on is the adversarial dynamic — you do write about this in the book. Precisely for the reasons you describe, the opponent will react by trying to make it harder for you to collect that intelligence, or attempting to lure you into very precisely attacking an object you should not. And then, precisely because these are precision-guided munitions, you don't have a good excuse for why you did that — it couldn't have been an accident, precisely because it's very precise indeed.

**Jeff:** That's a really good point. It reminds me of two things. In the early days, when I was in Afghanistan, I would hear a lot of conspiracy theories about how the US was supporting Bin Laden. The reason was that everyone had heard stories of the mythical power of the American military — "you guys can read license plates from space." So if you haven't killed Bin Laden, it's because you don't want to.

I also think of the war in Gaza, where you'd often hear: "What are we supposed to do? Hamas is hiding weapons under hospitals." I always thought that was the reverse of the framing. Often it's not hiding weapons under hospitals to protect the weapon — it's to invite the hospital being struck, for the exact reason you're talking about.

**Danny:** And then military theorists would say you've moved the war from the physical realm to the information realm — the old "you may win the battle but lose the war." I also want to say, we're talking about this in very clinical terms, and it is interesting to be analytical about it, but as your reporting does really well: these are not just abstractions. These are not big analytical categories. As you just said, it does explode somewhere, ultimately at some horrific cost.

**Jeff:** It's really hard to see that because of how we see footage of these wars. The first time we saw footage, in Desert Storm, we were seeing mostly targeting pod footage. It's very clinical, looks like a video game. Even now, with cell phones and such, when you watch the news you're seeing a building explode in the distance — smoke, dust, debris. You're seeing violence, but it's often violence to structures, to buildings. It's very hard, when that is the footage available to you, to picture that happening in your own backyard to people you know.

We've been allowed this remove from the real impact. When these bombs go off — the Paveway's warhead is often high-ferritic iron, really heavy iron that breaks into sharp shrapnel. You hold these tiny bits and they're so heavy, flying through the air at thousands of miles per hour. The human body is no match for it. You can imagine how horrible it is to have these sharp, heavy pieces of superheated shrapnel flying through — the kind of horrible things they do to a human body. We don't really see that. We don't really understand that. We think of a roof collapsing on a terrorist leader, a bunker being destroyed. I don't think we often attach the impact on human matter and minds to what we're doing.

**Danny:** You strike me as some kind of mixture — in the best possible way — between Hunter Thompson and Ryszard Kapuscinski. Who's your idol when it comes to reporting?

**Jeff:** Sebastian Junger. He has a profile of the kind of young hothead I was, just going to whatever place for glory. But he's a brilliant writer and incredibly humane, which I admire and envy and am surprised by. A lot of journalists covering really horrible things have to be clinical, have to remove the humanity. It's hard to be both in touch with victims as real people and do really excellent work. And he routinely does both.

Chivers is another. But most of the people I steal the most from are fiction writers. I'm trying to make things read more engagingly than clinical reporting, often to my editor's chagrin — "why are you trying to use all these big words?"

Colum McCann, who is a brilliant novelist, a beautiful writer, but all of his novels are based on either a real person or a lot of research. There's a saying a friend with a startup told me: the best technology disappears. If you develop a really good UX, you want the user to almost forget they're using technology. I feel that applies to research in storytelling too. There are times you read something and think, this person is just flexing how much they know. People like Colum McCann and the best novelists — you know they've done a whole bunch of research, but it feels effortless. It feels like you're just reading a story. That's the goal.

**Danny:** How do you do this? Because your books have this very cinematic, novelistic quality, particularly the later ones. You tell us what's happening in people's heads, what they're thinking, you recount dialogue that people had and that you cannot possibly have any record of. All of that cannot possibly be, strictly speaking, accurate. How do you deal with documentary accuracy versus atmospheric truth?

**Jeff:** We did aspire to accuracy. I went through a very rigorous fact-checking process with a fact-checker whom I really despised during the process — which meant he was good, and extremely patient.

One example is the section you helped me with, which was probably the most difficult. It's a spy-versus-spy story set largely in East Germany in the early eighties. Of the two main characters, one was dead and the other was unavailable for comment. Both were working for spy services — heavily confidential, redacted — and both operated in languages I don't speak. So I did what I often do: called my friend Danny and asked, "Can you help me find someone who can get access to some of these documents?" I ended up getting hold of some Berlin criminal court trial proceedings related to one of these terrorist attacks. There was a lot of information there, in German and redacted. So I called you again and had you connect me to another friend who could translate it.

Then there was a lot of cross-referencing, because many characters were referred to with code names, aliases, or were just redacted — sometimes all three. But pairing secondary materials like news reports with documentary evidence and heavily redacted documents, you can piece things together. For dialogue, there's no made-up dialogue. For what was going on in the White House and the National Security Council, there are minutes — also redacted, but you can see this person reportedly said this, that person reportedly said that. You don't have a whole conversation, so I have to provide filler. But all the dialogue can be backed up by either an interview or documentary evidence.

As for what's going on in people's minds — to have character-driven storytelling, if someone walks out a door and turns right, you need to know why they turned right. So it's important to try to get into people's heads. Even when it's a person available for follow-up interviews, it's really hard to double-check that. But we tried.

**Danny:** Speaking of the cinematic quality, it does feel like your latest book, and the one before it, The Mercenary, are constructed like movies. You've been involved in writing and co-producing films. Is that intentional — the structure, beats, and rhythm having that quality?

**Jeff:** I really appreciate you saying that. I strive hard to do it. It's not because I hope some Hollywood person will take interest — that's not the reason. It's that I live in fear of someone getting bored. I'm not a great reader myself. If there's too much technical information or if I can't envision it, I get expelled from the narrative. So I'm constantly redoing things to make it tactile, to make it so someone can imagine themselves in the world. Anytime someone says they think I've partially achieved that, it's way more validating than it probably should be. You need a spoonful of sugar to make the medicine go down. That's why I start with Kennedy and Nazis and all that red meat.

**Danny:** You wrote in a very personal article that it was the end of a war that did you in. What happened there?

**Jeff:** I always prided myself — or maybe I should say I always had the illusion — that whenever I went to these different places, I was different from everyone else. I was really trying to identify with the people I was meeting and to see myself as no different or better or worse than them. I didn't quite appreciate the fact that I had a passport; I could always leave. Even if there were times it felt like I was stuck somewhere, I wasn't going to be there forever.

The place I spent the most time in, had the most people I loved, and most identified as a second home was Afghanistan. When the troops pulled out in 2021 and the Taliban were about to take over, everyone I knew was in danger. I wasn't there, but it was all happening on my phone. I realized for the first time: I can't leave. I'm not there. This is happening in my imagination. Of course it was real, but I had to picture all the people I really loved and was worried about and what they were dealing with.

That experience of trying to get some of those people out while the whole thing played out on my computer and my phone — it didn't matter where I was, it was with me all the time. In a way, I think that was a bit of karmic retribution for all the places I'd gone and told myself I was helping by asking people to recount their most traumatic experience.

It was a crisis, both of conscience and of humbling, because I think — like a lot of journalists who cover war or tragedy — I licensed myself to meddle in other people's trauma because I was a journalist, doing the Lord's work. My attitude has changed, largely because of Afghanistan, where I was seeing the long tail of trauma. This time I wasn't going and living with people for a few weeks to recreate something horrible and then leaving. It was just with me — the cost of conflict — and continues to be.

My attitude has shifted. Now it's not that some horrible thing pops off somewhere and I think, "I'll go and figure out why later." Now it's: is there actually something not being discussed that I can, in a very modest, marginal way, add a little insight to? The bar to clear for me to really want to take something on has become a lot higher.

**Danny:** Do you miss the adrenaline?

**Jeff:** I think I miss the adrenaline. I also miss problems that are external problems — how to get across the border, how to get the rebel group to not get in the way. Things that are real and external and proximate. Someone told me, because of the experience of trying to evacuate people during the collapse of Afghanistan: "You're really good at making order out of chaos." And I thought, thank you, but I'm realizing that if there isn't chaos, I'm pretty good at making chaos out of order.

In my daily life, I create problems for myself. When you're out there and the threats are real and external, and the problems are real and external, it's almost a more natural place for me to be, where my mind can grab onto those things and try to solve them rather than the existential problem of what color shirt to wear.

**Danny:** You have an interesting sideline as a credited ghostwriter for — flippantly, you could put it — Republicans who get shot at. You wrote a book with Steve Scalise, the Republican congressman who was almost fatally shot by a domestic terrorist. And you also wrote The Fifteen Seventeen to Paris, the retelling of a 2015 story when three young Americans chanced upon an ISIS terrorist on a train from Amsterdam to Paris — a very unlikely story. You helped both Scalise and these three men write down their stories. How come?

**Jeff:** The 15:17 came first. These three young guys stopped a terrorist on a train and became mini-celebrities overnight. A talent agency signed them — speaking, TV shows, a book, a movie. For a minute it was pretty high-profile. The comparable at the time was Thirteen Hours, the book and movie about the annex security team in Libya where these contractors held off gunmen for thirteen hours. That book came about because those guys walked into publishing houses and told their story — it was clearly a book, there was a big advance, and they found a Pulitzer-winning writer.

This was different because it was ninety seconds on a train. It wasn't clear how it was going to be a book. The publisher wasn't going to give a big advance without seeing what the book would be. They couldn't just get the most accomplished writer and throw money at them because there was no deal. So they needed someone young and hungry enough to do a fair amount of work without getting paid — writing a book proposal, figuring out how to tell the story. Because I was young and similar in profile to the guys, the agency asked if I'd take a crack at putting together a proposal. I did.

Through a long, convoluted course, that book became the movie. And the movie was just coming out when that same agency was trying to shop a book by Scalise — a memoir of him surviving. Because I was associated with this other "Republicans who get shot at" book, my agent was able to drop in our first call with Scalise: "Oh, and by the way, apropos of nothing, I just saw the billboard for the Clint Eastwood movie. Anyway, back to you, Steve."

I took on the Scalise project because — and this may be another example of me deluding myself — the little I'd heard about the story involved all these different kinds of people coming together to save him. His Capitol Police security detail was a Black man and a Black woman. The woman was married to her partner — she was a lesbian. She literally took a bullet for him, and later they were both in rehab together. And there were all these other people: far-right Republicans, far-left liberals, the trauma surgeon who was this interesting philosophical guy, the helicopter pilot — all coming together for one purpose.

I thought: this is a way I can reach an audience I can't reach through The New York Times or The Atlantic, and this could be an important, helpful story at a moment when we're seeing all these people come together. I don't know that it quite had that effect in the end. But when I pitched doing it as much about all the other people as about him, Scalise was really into it and supportive. He was really wonderful to work with, which made it a little weird to be with this lovely, thoughtful, caring guy in person and then see this other side of him on the news. It was a cognitive dissonance.

**Danny:** For those who haven't read it — what's the CliffsNotes version of what happened to him?

**Jeff:** There's a tradition of the congressional baseball game, where every year Democrats play Republicans. It's a big fundraiser. For a couple of months leading up to it, there are practices. Scalise was at practice when a gunman opened fire. He was shot in a way that should have been fatal — the injury ruptured a vessel and he was bleeding out internally. That's an injury that's really hard to identify because he was also bleeding externally, but the real problem was internal.

It happened that another member of the Republican congressional baseball team had been a combat medic in Iraq and had lost a soldier because of this exact injury — a grievous wound that was not visible. Something about Scalise allowed this guy, Brad Wenstrup, to identify that he had this other thing going on.

Then there were all these other coincidences. Members of Congress don't have security unless they're in leadership. There are only two or three who have a small security detail. Because Scalise was Majority Whip, he was in leadership and had two security officers. It's only for that reason the gunman wasn't able to just walk out and mow everyone down — Scalise being there saved all the other congressmen. Then there were all these other people being in the right place at the right time, and some genuine heroics that saved his life.

Again, a lot of politicians who write books do it as a branding and campaign thing — how great their ideas are, their own story. The book is set up as many profiles of all these different people and how they intersected to save him. The fact that Scalise was excited to do this in a way that featured him less — I thought that was pretty telling.

**Danny:** I'll admit freely, it's probably not a book I would have read had I not known you, but it's an incredibly compelling read.

**Jeff:** A lot of people haven't read it even though they do know me.

**Danny:** But what I wonder is: you tell this story in a really crisp way, but I assume at the outset it wasn't obvious at all that that was the story, and there are a million ways of telling it. How do you find the thread that becomes the spine of something compelling?

**Jeff:** I've always been drawn to Rashomon-type stories. For some reason, I've always liked reading, watching, and trying to write stories where there's some unifying thing or place or event, and then we see different people coming at it from different angles — in this case, literally from different angles. I'm often trying to find a way to do that and then backing into a reason for doing it.

With the 15:17, the problem was: there's this really compelling event, but it's a minute or two on a train. How do we turn this into a book? I was struggling with that, and simultaneously with: how do you write autobiography when there are three "autos," three different people? At the same time, for some reason, I was watching The Affair, which does a Rashomon thing — the first half of every episode is from the man's perspective, the second half from the woman's.

At some point it clicked: tell the story from three different perspectives and you get a 360-degree view. Plus, when there's trauma, there are literal physiological changes, including to the shape of your eye — when people describe tunnel vision, the lens is literally changing shape. So these three people see the same event from their different perspectives, very similar but with key differences. I tried to use that to play up the fact that when you experience trauma, it's like a distorted video camera.

The same thing applied with Scalise. I couldn't tell a story from his perspective about how the trauma surgeon saved him — he wouldn't know that, and he was unconscious at the time. Being faithful to point of view required doing it from different viewpoints. And I always find it fascinating: how does the whole ambulance dispatch system work? This was an excuse to figure out how these different mini-worlds operate.

**Danny:** On The Fifteen Seventeen to Paris, we should briefly explain: the terrorist is in the bathroom, he has a rifle, he gets out and essentially immediately gets overwhelmed by the three men. His gun mysteriously doesn't fire, so they don't get shot, they overwhelm and subdue him, and then the police take over. Is that a reasonably accurate rendering?

**Jeff:** Yeah. Two other things: he also had a pistol and a knife, and he was able to shoot one person in the neck. That person survived partially because Spencer, one of the guys, was an Air Force medic and was able to identify and stop the bleeding. The other thing is that technically it wasn't that the gun jammed — it was what I think is called a bad primer. That distinction matters because when a gun jams, there's a way to clear it. In this case, he pulled the trigger, the hammer hit, everything happened like it was supposed to, and the bullet just didn't go off. That provided extra time for Spencer and the other guys, who were on the other side of the train car, to get up, run down, and tackle him. Spencer's hobby was jiu-jitsu, so he knew how to put the guy in a chokehold.

**Danny:** It's a wild story. These guys are brave beyond imagination.

**Jeff:** I should also tell you, Danny: the guy who played the terrorist in the movie — who is, weirdly, a wonderful, kind, awesome guy — read the audiobook for The Mercenary and reads the audiobook for the part you helped with in The Warhead, the section about Wappen und Vastasi.

**Danny:** There we go. That brings us to The Mercenary, the book you published before The Warhead, which also has that Rashomon quality you've described. You tell the same story twice: your experience in Afghanistan, then retold from the perspective of your friend who was with you and helped you along. How therapeutic was writing that book?

**Jeff:** That's a really interesting question. In some ways it was therapeutic. I think the reason it didn't fully feel that way is because of how it came about. I love this guy, Amal. He was my first friend there, saved my life a bunch of times. He really was a brother, is a brother.

Early on I would write about some of our experiences without knowing what for. In 2021, when Afghanistan came back in the news and Biden announced we were going to withdraw — but before I realized how bad it was going to be — I started talking to publishers: "Afghanistan's in the news, you might be looking for an Afghanistan book, here's the story." The publisher that did the 15:17 was interested.

My thought was I'd secure the deal and do it after The Warhead, which I was still in the middle of. Because I was still under contract for The Warhead, we couldn't go out to different publishers. This one said: okay, we'll do it, we're going to give you a tiny advance, and it has to come out on the one-year anniversary of the collapse of Afghanistan.

I ended up trying to write two books while in the midst of evacuations. In a way, it was a real-life manifestation of what became part of what I'm struggling with in that book: here I am taking time away from evacuations to move my words around. The gulf between what we often license ourselves to do as journalists and our actual impact can be fairly large, and it felt really large there. I had a lot of ambivalence working on it — I need to finish this, it needs to be good, but also, why am I spending time on this and not that?

I think that preempted what could have been a pretty therapeutic process. There were aspects that were therapeutic — especially when we get into my mom, putting her in a memory care unit at the same time we're trying to evacuate people. But the overall experience was a real-life manifestation of one of the things I'm struggling with in the book. The book is called The Mercenary, ostensibly because my friend becomes an arms dealer, but it's also trying to hold a mirror up. I'm probably more the mercenary. I go to this foreign place and treat it like a theater for my benefit — to write about and make a career out of — and the book is me struggling with that.

**Danny:** Why did you choose to put your mom in both of the books?

**Jeff:** For The Mercenary, it was my agent's idea. By that point, I knew the structure: the same rough period of time from my perspective, then the same period from Amal's perspective. I didn't know where it went from there. There was a period when I was supposed to be reporting the book that I could not get in touch with Amal. He'd disappeared. It turned out he was in jail — accusations of domestic abuse.

My agent, at some point when I asked him how to end this thing, said: "I think it's Amal in literal jail and you in this kind of mental jail because of what you're going through." That set me off, because part one is from my perspective, part two is from his, and this allowed part three to go back and forth in a sinusoidal rhythm — me, him, me, him — me with my mom and him in jail. It became a way to visualize and articulate what is arguably an indirect result of the trauma we experienced, and a lot of the trauma he experienced because of me.

**Danny:** What are you working on now?

**Jeff:** Doing podcasts about The Warhead. I have way too many ideas for the next thing, which I'm trying to stop myself from diving into because I have family obligations I've been putting off for years. But you'll be the first to know when I begin putting pen to paper, because I will probably need your help with some aspect of it. I should also say that the Yemen story too — you're my first call. Why I thought you would have a connection there, I don't know, but you did. You connected me to Ahmed, who turned out to be a very good friend and extremely helpful. So this whole thing goes back to Danny.

**Danny:** Thank you. Thank you, Jeff. This was a true pleasure.

**Jeff:** Thank you so much, Danny. I hope I get to see you in person soon.

**Danny:** I hope so too.

**Jeff:** Take care, brother.

---

## Intelligence Saturation and the Economics of AI – with Ioana Marinescu

2026-03-04 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/intelligence-saturation-and-the-economics-of/id1839231559?i=1000753008033) | [Spotify](https://open.spotify.com/episode/2p6V1TswOkANXuihTWsg82?si=3ae4c9393d6249e3)

Ioana Marinescu — associate professor at the University of Pennsylvania, research associate at the NBER, and member of Anthropic's Economic Advisory Council — joins Danny Buerkli to discuss her 'intelligence saturation' paper, which divides the economy into an intelligence sector and a physical sector to model what happens as AI automates cognitive work. They discuss why a recent scenario piece that moved markets gets the economics wrong, what parameters to watch to understand where we're headed, why being first to AGI may matter less than people think, her policy proposals for AI Adjustment Insurance and a Digital Dividend, what UBI experiments do and don't tell us about a world without work, Ioana's favorite theory of divorce, and what it feels like to be an intelligence worker watching AI get better at your job.

**Danny Buerkli:** My guest today is Ioana Marinescu. Ioana is an associate professor at the University of Pennsylvania and a research associate at the NBER, the National Bureau of Economic Research. She was a principal economist at the US Department of Justice antitrust division from 2022 to 2024. She sits on Anthropic's Economic Advisory Council and has done much work on labor markets, antitrust, and AI. Ioana holds not one but two PhDs, one from LSE and one from the School for Advanced Studies in the Social Sciences, EHESS, in Paris.

Ioana, welcome.

**Ioana Marinescu:** I'm very happy to be here.

**Danny:** Ioana, is the "lump of labor" fallacy a fallacy, or should we call it the "lump of labor fallacy" fallacy?

**Ioana:** I think it pretty much is a fallacy. It's really easy to get carried away with simplistic reasoning about the economy, especially when you think about a sophisticated economy like we have today.

**Danny:** We're recording this in February 2026. A small research firm, Citrini, published a 7,000-word scenario report just two days ago and ended up — maybe surprisingly or not — moving markets down.

**Ioana:** Crazy.

**Danny:** Surprising. I did not see that coming. Maybe to get us started, what did you think of the scenario they described?

**Ioana:** I read that and thought it was interesting — they're thinking through all of the bad things that could happen, and many of those things could happen. But I felt the economics links were often somewhat tenuous and not fully thought through. That's why we need to sit down and think through every piece of the mechanism. Could that happen? Under what conditions? And so on. It is useful in terms of pointing out possible adverse effects, but I don't think they're necessarily going to happen.

To be fair to them, they did frame it as a scenario. And I think the markets are generally feeling a little nervous right now about AI. The scenario somehow hit a nerve and triggered a panic.

**Danny:** If I were to summarize the core thesis of the scenario: AI gets really good, it displaces a lot of humans, and that leads to a drop in aggregate demand — and that's a bad thing. What exactly is wrong about that story?

**Ioana:** The fundamental problem is that they don't really think about how workers reallocate. Their story is largely about intelligence workers ending up as, say, Uber drivers — that was one of their examples. They seem to think there will necessarily be a wage decline for these people. But if you think about it in a more economically principled way, that doesn't necessarily follow, because the progress of AI in some sectors will often have a positive effect on what we call the marginal productivity of workers in other sectors.

As a generic example: the Uber driver, thanks to AI, might now be able to use more different services. The service itself becomes higher value to customers, so the driver's value could go up. They don't think through that side of things, which seems quite plausible. The question is: which effect dominates? That's part of what we're trying to explore in our paper.

**Danny:** Speaking of your paper — you have a recent paper out, the intelligence saturation paper. One of the novel bits is that you posit we may wish to think of the economy as having two parts: the intelligence part and the physical part. Maybe introduce the idea briefly.

**Ioana:** Interestingly, the Citrini report talks about the "intelligence crisis," so I was struck by the parallel with my paper. The basic assumption is that AI is possibly going to replace a lot — potentially all — intelligence jobs. What do we mean by intelligence jobs? It's essentially any job you could do remotely at your computer. If you can do it at your computer, plausibly down the line AI can do it. And don't tell me "what about Zoom calls?" — the video generation is getting much better, and you'll be able to have an AI avatar that does calls.

If that's the case, then people have to work in what we call the physical sector, which is everything else — essentially any job that requires some in-person activity. It's not necessarily physical as in "I need to do something with my hands," but I need to be there in person as a flesh-and-blood human being in order to deliver the job, and this has to be a significant part of the job. I might also be working with AI, but fundamentally, I've got to be there in person. Those are the physical jobs as opposed to the intelligence jobs.

**Danny:** How do you think of the separation between the two? One may think this is not necessarily a static boundary and that it's endogenous to AI capabilities in the first place.

**Ioana:** Absolutely. And I should say why my co-author and I — he is an AI and neuroscience expert, very much thinking about physical embodiment and the interaction with intelligence — why we think this framework is useful. If you look at price trends, the price of AI is going down exponentially, and that's probably why it's so exciting. But the price of physical capital — the kind you would need to do something in the physical world, even when using AI — is going down much more slowly, more linearly. In terms of economic incentives, the big incentive is to replace people in the intelligence sector because it's so cheap to do with AI, whereas doing anything with robots remains fairly expensive.

Relatively speaking, what's most advantageous is to expand the use of AI in intelligence tasks because it's so cheap. I found it intriguing that the Citrini scenario had intelligence workers going to drive Ubers — which is a physical job. They have to be there in person to do it.

**Danny:** Another person who's modeled this is Anton Korinek and Donghyun Suh in a 2024 paper, who took a different approach: they parceled out every task in the economy, ordered them by the difficulty of automating, and assumed that AI will progressively work through that list. If Korinek were to critique your model thoughtfully, what would he say?

**Ioana:** Our models are fairly similar. The intelligence part of our model is pretty much the same as theirs — we're also saying AI is going to replace more and more intelligence tasks, plausibly starting with the easier ones and progressing to more difficult ones. But the question is whether there's a physical sector where AI doesn't make much inroads. That's the key contention, and it's an empirical one.

In their paper, they discuss the case of whether there will be some reserved array of tasks for humans. What we're saying is similar, but it's not that it's reserved — it's just technically very difficult to replace people in physical jobs for many reasons. It's not easy, it's not very cost-effective. You can do it, but at what cost? Relatively speaking, it's easier to replace workers in the intelligence sector.

**Danny:** One of the beautiful things about your paper — and this is the explicit aim — is that it allows us to reconcile different intuitions that the more computer-science-oriented AI folks have with the intuitions that classically trained economists have. What kind of reaction have you gotten from the AI community?

**Ioana:** AI people tend to be smart, so they get the point. But usually what I hear is: the robots are going to get so much better, and thanks to AI, the robots themselves are going to get cheaper, faster, and self-improving. Maybe that's possible. We just don't think it's very plausible, because that gets to our point about intelligence saturation. Just adding more and more intelligence to a physical process is unlikely to make it tremendously better. There are multiple pieces of evidence from prior experience suggesting that it's difficult. You can never say never, but the idea that using AI to improve robotics is going to solve all your problems is certainly a high hurdle.

A lot of AI people are also aware that developing robotics is actually not easy. Right now it's still very difficult. One of the funny things is that you see on X and other social media all these cool robots doing human-like things — but often the videos are accelerated because the real speed at which these robots operate is so much slower. You would lose patience watching them do a task. That shows you how powerful human bodies are for many of these tasks — human hands, for example, are incredibly well optimized. It can be done, but at what cost and speed for a robot to achieve the same performance?

**Danny:** David Autor has this distinction between highly qualified tasks and lower qualified tasks. The intuition is: if I automate the lower qualified tasks, that's good for the worker because I'm left with the bundle of highly qualified tasks. But if I do the inverse — automate the highly qualified tasks — I'm left with a bundle of non-qualified tasks. That's a problem. How does that mesh with your model?

**Ioana:** That's an interesting way of looking at it. Our perspective is more macro. It's also more simplified because we only have one type of worker in the economy. But the bottom line is that you really have to think about this at the macro level — what are the interrelations between sectors? Whenever we say workers are displaced from somewhere and go somewhere else, and that somewhere else might currently be lower wage, I want to say: wait — that somewhere else could become much higher wage in the near future because of the general economic growth that AI is generating.

As an intuitive example, take primary school teaching. It's an in-person job — physical, in my definition. We've seen that online education just doesn't work for most people, so the in-person experience is really important. In that sector, there have been very few productivity gains. A teacher fifteen hundred years ago probably didn't do things hugely differently. And yet teachers today are paid a lot more than teachers fifty or a hundred years ago, because the economy is better and that pulls their wages up along with the rest of the economy.

Similarly, jobs that today aren't well paid could become higher paid due to the general productivity growth that AI generates. It's an important positive channel that people often overlook. The partial equilibrium may not be the general equilibrium.

**Danny:** Which parameters would you be watching most closely to understand where we're headed?

**Ioana:** First, you want to look for a decrease in the share of workers in intelligence jobs. If that's not happening, the revolution hasn't started yet — there's got to be worker reallocation. Then the question is whether the size of that reallocation is sufficiently large relative to the productivity gains in the intelligence sector.

The intelligence sector — the one where everything could be done virtually — is where AI is most easily deployed. As we deploy AI, does it raise the output of this sector a lot? And are relatively few people displaced from it? That would be a good scenario where wages are likely to grow. The more you see the productivity effects of AI in the intelligence sector weakening for a given amount of labor reallocation, the more likely that wages decline.

Those are the two core forces, and they depend on parameters. In particular, it depends on how easily you can replace a physical job with a virtual one — for example, how easy is it to replace an in-person primary school teacher with an AI doing the teaching? If it's relatively easy, you could see a fairly large wage decrease as automation progresses. It could even be the case that wages rise at first, buoyed by general productivity gains, but as we realize it's actually pretty easy to make everything virtual, wages could decline as automation continues. So one core factor is how substitutable physical and virtual jobs are — how easy is it to replace an in-person job with a virtual one and achieve about the same result?

**Danny:** A lot of these conversations end up coming back to a bet on future capabilities. Is it ultimately all a function of capabilities, absent maybe some hard physical constraints?

**Ioana:** For sure. But something AI folks don't think enough about is that they often translate technical capabilities of AI directly into economic value, and those could be very different. Just because something is technically possible doesn't necessarily mean it's economically efficient.

**Danny:** To this exact point, there are two ways AI can have an effect on the economy today. One is through diffusion — the technology exists, it spreads, it gets actually used. The other is through capability increases. Which effect is currently dominant?

**Ioana:** That's a good question, and I don't think we have robust empirical evidence to say. What's interesting is that right now it's still a bit unclear if AI is doing much of anything in terms of labor market effects or productivity effects. We've seen some productivity gains, but is it really AI? Is it other things? Empirically, we're not sure where we stand.

However, those two things are very important theoretically and very different. In our paper, we carefully distinguish automation — pure automation — from capability growth. When I think about automation, it's about taking a given stock of AI, however capable it is, and applying it to more and more things. That's where people potentially get displaced from the intelligence sector and must reallocate to the physical sector, and that can have negative wage effects under certain conditions.

What you're talking about with increased capabilities — I think of that as capital deepening or AI scaling. That means we have more AI or better AI, so the total amount of AI available goes up. From economic theory, that's always a good thing in itself. If you just add more capital to the economy, it makes workers more productive. In every basic model, that increases wages. In our model, once we automate everything that could possibly be automated, if we can still add more and more AI, that extra AI is going to benefit workers. But that could come after a catastrophic decline in wages.

**Danny:** In the long run —

**Ioana:** It will all be fine.

**Danny:** Exactly. To butcher a quote. To this point, you've suggested we may wish to smooth this transition. If we were to worry about a potential decline, we might not want to stave it off, but we might want to steer the speed — slow automation. The obvious counter is: how should we think about the trade-off between the smoothness of the transition and international competitiveness?

**Ioana:** This is a difficult question. First, in our model, the model itself is not well set up to think about transition costs because workers are all the same and can costlessly transition from an intelligence job to a physical job. It's simplified to understand what would happen in that situation. In the real world, workers who are forced to change sector typically experience significant wage declines of about 15%. That's significant, and from a policy perspective you might want to guard against it. There is economic theory suggesting you could want to tax AI a little bit just to slow things down — give workers more time, don't have them all get replaced at once.

When you do that, you're also delaying the growth effects of adopting AI. That's an obvious trade-off. And then there's international competition. But here our paper has something original to say: how important is it to be first? There's this race to AGI where everyone thinks that whoever is first will transform the world. Well, maybe not. That's where intelligence saturation comes in — the more intelligence you add, the less incremental effect it has. If that's the case, even if you're the first to achieve greater capabilities, you'll only be a little bit ahead of the next. So it's a trade-off, but it's not this infinite advantage where going first gives you superpowers. To believe in that narrative, you really have to believe that intelligence saturation is not a thing — that having much more capable AI gives you a massive lead over everyone else.

**Danny:** In your model, definitionally, there is no such thing as strong AGI that automates all physical labor. Therefore —

**Ioana:** Therefore, it's a bottleneck. You can always add more, but —

**Danny:** Right. I guess the critique would be that that's an assumption baked into the model, which therefore produces this result.

**Ioana:** Absolutely. And you could see it more generally: maybe you're not happy with my characterization as "physical." Rename it whatever your preferred sector that you think AI has trouble automating, and the mechanics would be the same. We think physical is a useful label, but if we believe there's some kind of work that is difficult or not cost-effective to automate — again, it's about the economics, not necessarily that you couldn't technically do it — then the same reasoning would apply.

We also have to think about different time horizons. There's no doubt that in the medium run, at least, there are sectors that will be incredibly difficult to automate. We say physical, but it could be other more subtle things. Our model remains valid as long as you replace "physical" with "non-automatable" and "intelligence" with "automatable." You would get similar dynamics.

**Danny:** The model holds as long as there's something that we believe may not be practically automatable. And if everything is practically automatable, then we're all better off, as some people have pointed out.

**Ioana:** Exactly. And one more thing. The intelligence saturation assumption hinges on two things. First, there are some tasks that are not automatable. Second, there is a complementarity between this non-automatable sector and everything else. Because if you can just substitute one for the other, then AI could still grow and you wouldn't have the bottleneck — you could have very strong growth. The core of intelligence saturation is that there are two sectors: physical and intelligence. The physical is non-automatable, intelligence is automatable, and the two sectors are complements. That complementarity — the fact that you cannot easily substitute a physical thing with an intelligence thing — is what makes intelligence saturation kick in. You add more and more intelligence, and it only does so much. That also means that if China or whoever gets to AGI first, well, that's nice, but they will only be a little bit ahead.

**Danny:** It will only get you so far. On the more practical policy end, in an essay in the Digitalist Papers volume two, you proposed two policy solutions to this transition dynamic: an AI adjustment insurance and a digital dividend that would fund it. Say more about the AI adjustment insurance mechanism in particular — why that mechanism and not another one?

**Ioana:** The idea with AI adjustment insurance is that workers have to transition from intelligence jobs to physical jobs, and there are still jobs — people just have to switch. As I said earlier, in practice, unlike in my model, this is not costless. People often lose 15 percent of their earnings. They spend time unemployed. We already have programs that help, but to the degree that we make a policy choice not to limit AI and to let it run, there's an argument that we should help people who are adversely impacted. Just as with trade — when we decided to have more free trade, we knew some people would be negatively affected, and we put policies in place. The Trade Adjustment Assistance program existed and worked really well for those who qualified and were displaced by trade. AI adjustment insurance follows the same logic: help people transition, acknowledge that it's difficult, and support workers through it.

The program has three elements. It has longer unemployment benefits, so people have more time to search. It has training. And the third, perhaps most unusual element, is wage insurance. If you take a job that pays less than your prior job, the insurance covers a percentage of the difference. Say you take a job that pays $1,000 less — if the insurance covers 50%, it would pay you $500. That helps people make the transition. With Trade Adjustment Assistance, the wage insurance was shown to be incredibly effective at helping people return to work. And amazingly, it actually made the government money because people returned to work and paid back the subsidy through payroll taxes. It's a really appealing program design to encourage and support worker reallocation, which we generally expect with any big technological change.

**Danny:** Thomas Piketty, whom you know well, pointed out that r is larger than g — the return on capital is larger than economic growth — and that leads to concentration of capital. What happens to r and g in your model?

**Ioana:** What we see in our model is that the labor share goes down. Even when wages increase, AI in this model disproportionately benefits capital. Then it depends on who owns the capital. If capital ownership were somewhat equally distributed, everyone would also broadly benefit. It therefore depends on how broadly capital ownership is distributed.

This gets us into my second policy proposal: the digital dividend. The idea would be a small tax on the AI sector broadly — not just the makers of AI but everyone who uses it, because they get a lot of benefit from using it. We could put this in a fund where people receive the benefits. Or there could be other schemes — the government could take some participation in companies and then distribute the returns. It's really important to think through this transition in terms of its effects on inequality, given that it's likely to benefit capital. Where does the income from capital go? That becomes an important distributional question that one shouldn't neglect when thinking about what sort of world we want to live in.

**Danny:** There are some interesting political economy questions there. Something a lot of people worry about — probably rightly so — is that the timing matters a great deal. There may be a point after which it becomes very difficult to change the regime you're in, because capital has concentrated to such a degree that it would be even more challenging than it already is to do something about it.

**Ioana:** That's why I was so interested in thinking about whether we can set up these policies now, sooner rather than later. With the digital dividend, if the labor market isn't too disrupted and people are able to transition and things are okay, then maybe we don't need it. But if it turns out there's huge job displacement and the physical and intelligence sectors are very substitutable — so people lose their jobs, some can't find new ones — then we could transition this dividend into something bigger, more like a UBI-style program that covers more people, including young people who don't qualify for unemployment benefits. With the AI adjustment insurance I discussed first, you have to have worked enough to qualify. If you're a young person without work history, you don't qualify for anything. If this is going to hit young people hard, the dividend is a way to cover them. Similarly, even if you have benefits, they eventually expire.

A core aspect is that we don't know exactly what's going to happen, so we need a flexible system that can adapt. The fundamental danger is that once winners and losers are clearly known, it can be difficult to persuade winners to give some crumbs to the losers. It's better to have something in place ahead of time.

**Danny:** You want some kind of Rawlsian veil of ignorance before this all hits — which is the elegant thing about your proposal. And as you point out, particularly for the digital dividend, the amount need not be fixed. What you want is to have the rails in place to collect it and the mechanism in place to distribute it, but you want to be able to change the rate depending on the needs.

**Ioana:** Exactly. If most of the economy is automated, you can just put a tax on all firms — that's what I call expanding the base. First, I say let's put the tax on AI-related sectors, because that can also slow things down a little, which has some benefits. But ultimately, if most jobs are replaced anyway, we can expand beyond that. Having a flexible system in place is really important.

I also want to add that it shouldn't only be about redistribution of income. That's very important — people need an income to live — but there are other considerations. For many people, jobs have a lot of non-monetary value. They provide community. They give a sense of excellence in developing your skills. Not having jobs could hit people's well-being in ways that go beyond not having an income. We need to think about what people would do if that were to happen — what could replace the non-monetary benefits of jobs?

And as you were saying about the concentration of capital, it's really important to think about power and who decides. If a lot of the benefits go to capital and we change nothing, then whoever gets those benefits has even more power. In that sense, it's important to think today about the trajectories — what will it look like tomorrow and who has the power to decide. At a very basic economic level, if the owners of capital get most of the benefits, all this AI boom will be directed to satisfy their needs and not necessarily the needs of other people. The economy responds to whoever has the money to buy — that's how the capitalist system works, and it's very effective at that. So it's not just about redistribution; it's about predistribution — the conditions that will shape the economy and society at the next stage.

Another element of my proposal that speaks to this is the training side. That element will influence what jobs are created, because firms want to hire in domains where there are people who want to work and have the right skills.

**Danny:** And they don't want to shoulder the training cost if they can avoid it.

**Ioana:** Exactly. If we had more training in physical jobs — health care in particular seems like a very good sector to invest in, because it's growing for other reasons and it's a physical sector — that would likely encourage the development of the sector. If we invest in training today, it will likely encourage the development of health care tomorrow. These are ways we can shape outcomes. It's not just about making sure people have money tomorrow — it's about how we shape our economy and society so that tomorrow we end up in a better situation.

What counts as "better" is in the eye of the beholder, and that's why we should debate the goal. Then we can run models and scenarios: with this policy, you could get this result; with other policies, you could get other results — and at what cost each time. I'm not claiming it's obvious what you want to achieve, and reasonable people will disagree. But it's important to think ahead about what instruments we have and what we can achieve with them.

**Danny:** You point out in the intelligence saturation paper that investments in capital in the physical sector are a way of shaping the trajectory.

**Ioana:** Exactly. People, according to our model, are going to transition from intelligence jobs to physical jobs. A big reason their wages go down is overcrowding: imagine there's a fixed number of hospitals, and now we have twice as many nurses — that's bad for productivity. So you need more hospitals.

We haven't worked it out in the model, but there will be market incentives to invest there because it's a market that can be captured. But if we think that's not enough — and there are policy decisions and other objectives at play — we might want to encourage the sector through training (which in some sense subsidizes the sector, since firms don't have to train workers themselves) or through other types of subsidies. That changes the structure of our economy by giving extra advantage to certain sectors that are likely to absorb the workforce as people are displaced by AI — ideally sectors with reasonably good jobs that we can train for.

**Danny:** You mentioned predistribution. There's a domestic view of that, but also an international view. In a conversation in London not long ago, something people worry about a lot is that if you're a country that doesn't own frontier models — on the assumption that frontier models will continue mattering — you can imagine a world where a lot of your GDP flows to a place like the US. That generates several complications, including that you would really struggle to collect any meaningful digital dividend. What to do?

**Ioana:** This problem is more subtle than it may seem. First, less developed countries have a much smaller share of their workforce in intelligence jobs. The displacement threat is therefore far lower, and the extent of the labor market problem could be much more limited in these countries. That might also mean they're not getting the same productivity boost, but on the upside, they won't be as severely impacted as a country like the US where the largest share of the labor force works in intelligence jobs.

Second, it really matters how competitive the AI sector is. There are reasons to believe there's a lot of competition — many different models developed by many different actors, some incredibly cheap. They might not be quite as good as the frontier model, and that's where intelligence saturation comes in. If you think being a little bit better is tremendously important — so there's no intelligence saturation — then maybe it's a problem. But if being a little bit better is just a little bit better, then poorer countries can use a cheaper, nearly free model and achieve almost the same results.

The potentially less grim view for developing countries is: first, they have fewer workers who might be affected; second, they can use AI very cheaply if the sector stays competitive and if using a better model doesn't yield huge additional benefits. They have the chance to develop their economies at low cost without getting hit by huge job losses. Some things could go wrong, but it's not obvious that things will be worse there. In some ways, they could be better.

**Danny:** You've done a lot of work on UBI. The underlying question is: is work a good or a bad?

**Ioana:** It depends — the economist's favorite answer. In the very basic model in economics, people work to make money. They don't like working, so you have to pay them. If that's all there is to it, then great — no more jobs, we still get money, wonderful.

But if there are non-monetary benefits of work that aren't easy to replace, then it's more nuanced. Is there other ways we could get those benefits — through volunteering or other social structures that would provide community and skills development? I think the second view is more realistic. Some jobs are really bad and people would rather not do them. But a lot of jobs, even those that might seem bad from the outside — people might have colleagues they really like, even if the job itself isn't great, and that really matters to them.

We can't assume it's necessarily for the best that we eliminate all work. We have to think about what jobs are good for besides income and how we can get those benefits in different ways. Let me give an example. With the digital dividend, one approach — what I'm proposing — is that people simply receive income at the individual level. But you could devote some of the dividend to local community investment trusts, where people are called upon to decide how the money should be invested locally to develop their neighborhood or community. That could be a kind of job for people — administering and deciding what they want to do at the local level. That would restore more power, meaning, and engagement. Those things aren't always easy to design well, but it's worth thinking about. In Europe, there have been schemes like this — transition funds in Spain, for instance, where they closed mines and local communities had to decide how to use the funds to support people who had lost their jobs.

Economists often think about the money, and the money is very important. But there are other elements important for people's well-being that we shouldn't neglect, and we can address them both theoretically and empirically by learning from existing experience.

**Danny:** We have a lot of empirical evidence on the effects of UBI from various experiments. You've done much of this work. One limitation is that these experiments all happen in a world with work. How much of those insights would transfer to a world with no work or almost no work?

**Ioana:** That's a valid question. First, people are really worried that UBI will disincentivize work. In that case — hooray, we don't need to worry about that because there's no work to be disincentivized.

The benefits of UBI don't really depend on work existing. It's rather that the removal of work could cause harm, including non-monetary harm, that we need to think about rectifying. We know that unemployment causes mental health problems. In fact, researchers have shown it increases mortality — people who get laid off are more likely to die. The baseline probability is very low, so don't panic, but it becomes more likely. Unemployment has real adverse psychological impact.

Jobs can be very important to people, and UBI is nice as far as income goes, but it doesn't by itself provide the other benefits of jobs. You probably need some other social structure for that. There's a real policy design question here, because some people say people will just invent that by themselves. I think that's tough. As an individual, it's often difficult to create the kind of structure that would really help you have meaningful activity, and often it has to be social. Then we have the classic coordination problem, where policy could really help.

**Danny:** You can invent it, but it only really works if everyone around you buys into some similar notion of what the alternative is. If it's just you and two friends, that feels difficult.

**Ioana:** Exactly. That won't necessarily happen just because people have incomes. The coordination problems are significant. That's where having the right incentives and social structures in place can help generate meaningful activities for people.

**Danny:** If we go back to a world with employment and to your model — in a world of transformative AI, would you expect employment tribunals to be sympathetic to workers or to firms?

**Ioana:** It's funny you're asking me that because I have a paper about this, where I looked at the impact of economic conditions on how employment tribunals decide. To the extent that the situation is dire for workers, tribunals might have a bit more sympathy for the worker side. But it's hard to tell. If we really think AI is going to displace jobs on a massive scale, this sort of thing might slow things down a little, but it's not going to drastically change things — especially because there's a lot of talk about AI-native companies that start with a completely different model and won't have people to lay off. I think it's important to support people through the process, but there are limitations to how much you can slow down AI deployment. It's more about having a comprehensive approach so that we get good results for society overall, rather than going on a fool's errand of saying there will be no AI in this country.

**Danny:** I thought you might say that employment tribunals might rule in favor of firms, because AI-induced competition has become so intense that they feel for the firm rather than the worker. Do you expect AI to increase or decrease labor market concentration?

**Ioana:** That's a tough question. I think there will be new firm creation due to AI — new business models become possible. But plausibly those firms might not employ many people, almost by definition. So while there might be new firms, which typically helps decrease concentration, if these firms don't have a lot of employment, I don't know how useful that will be.

There are reasons AI could help small firms. Right now, it's hard to do certain core functions — accounting, HR — well when you're really small, and that might be an impediment. But as AI tools automate many of these business functions, it could make it easier to be a small firm. You could have many small firms rather than big behemoths that exist in part to absorb the fixed cost of complex departments. But it's quite speculative, and I'm not sure which way it would go. Big firms and incumbents will naturally try to reinforce their position. How it comes out is fairly unclear.

**Danny:** That would imply that from a labor market concentration perspective, you would expect to see more mergers because they just wouldn't affect labor market concentration. But it's unclear whether anyone would want to merge at all given what you've just described.

**Ioana:** Exactly. It's unclear how the equilibrium develops. This is uncharted territory. But usually when there's a new technology, it reshuffles not only workers but firms — which firms are successful and so on. Whether that leads to more or less concentration is hard to tell at this point.

**Danny:** For something completely different: what is your favorite theory of divorce?

**Ioana:** I have written a paper on that. Why do people divorce? There are two big theories. Either you get to know somebody, they seem great, but then you live with them and it turns out they're not — you made a mistake in assessing compatibility. Or you like them and they are great, but over time people change, and in some cases you don't like how that change happens.

It turns out we can use data to disentangle these theories. In my paper, I showed that the big cause of divorce seems to be that people change. It's not that you chose the wrong person to start with — you probably chose reasonably appropriately. But people change, and sometimes they change for the worse. That seems intuitive, too. Nowadays people usually don't rush to marry. They know each other, they often marry later, so they have a good idea of compatibility. But statistically, we can show that it's mostly not about insufficient information, but rather that people change, and sometimes the changes aren't appreciated by the other person.

**Danny:** I suppose that would imply AI will not change the rate of divorce, because I could imagine a story where AI improves the information I have about a potential partner ex ante. Maybe AI will also make us change more — that seems less plausible intuitively, but who knows.

**Ioana:** It might not change much unless it makes you change more rapidly. I use AI a lot. I don't know if I've changed. Paradoxically, sometimes I feel like it makes my job harder, which is a funny thing to say because it's very useful. But given the type of work I'm trying to do — having deep insights, things that matter, things that are well-checked and well-developed — I use AI to make things even better. In the end, I spend the same amount of time or sometimes even more, because now I have this tool to push things further.

It ups the ante. I have this super tool, so I can't be satisfied with something that's merely pretty good. Now I need to make it even better. Paradoxically, it's the source of more stress than you'd think. At first you think: I can do so many things! But then your bar goes up.

**Danny:** But now you can do so many things, and you can do them almost perfectly.

**Ioana:** Exactly. That's tough. Not to mention that I am in large part one of these intelligence workers whose jobs are at risk. So there's an identity crisis looming. It's not easy to be in the middle of this.

**Danny:** Joking aside, do you believe your job is actually at risk?

**Ioana:** In the short run, no. Adoption is slow, I'm very privileged, and I have my unique strengths. I also teach students in person, so that helps.

**Danny:** So you're also part of the physical sector.

**Ioana:** Exactly, that's the physical side. But as far as my research job, the amount of progress the tools are making is definitely scary. Judgment remains very important, and that's why I think I still have a lot of value. But we don't know where this is going.

Sometimes nowadays I find myself purposefully writing things my own way, even though I recognize that maybe it's worse — less smooth. But that's how I think, that's how I express myself. There's this value in authenticity, but it's warped, because what am I doing? I'm making things almost worse on purpose. It's a bit like pottery — handmade, organic. You could have industrial pottery that's perfect, but it's not authentic. That's definitely driven by an identity threat, because I pride myself on doing really good work, not on being authentic at the core. I think: I have all this expertise, I can do all this cool stuff. And now AI can do a lot of this cool stuff. What becomes more important is my judgment and taste — what I think is important, and maybe my quirks of expression. And I don't entirely like that. I need to contend with a new way of thinking, and it feels uncomfortable.

**Danny:** You've studied philosophy. There's a great book by Lionel Trilling, Sincerity and Authenticity, where he essentially argues that authenticity is overrated in many aspects of our daily interactions.

**Ioana:** That's what I used to think.

**Danny:** But maybe that has now changed.

**Ioana:** Exactly. I used to think: all these people talk about authenticity — who cares? The important thing is to get great results. But now I think: maybe I need to go back to authenticity, if I still need to matter, because AI can do the work. Not quite — there are limitations. But it's getting better by the day. It's definitely an identity threat, and I think it's going to reach more and more people. Things are reorganizing every day, and I'll probably adapt. I'm not in any immediate danger — as I said, I'm very privileged. But it's more a psychological kind of threat. How do you adapt to a somewhat different understanding of what you're for, what you're good for? We live in a weird moment.

**Danny:** My related pet theory is that politically, possibly the most salient group will be highly educated academics in high-prestige, relatively low-wage jobs. They're currently compensated by prestige, and so they put up with a comparatively low wage. They tend to have political influence, and LLMs are really quite good at producing output that looks suspiciously similar to theirs. It goes to the heart of professional identity. If you already don't have a lot of money and then the societal prestige goes away too — historically, that's been a dicey setup.

**Ioana:** I think it's hard. If you see it through the lens of our theory, it's about whether people like me can find meaning in physical jobs — which in my case would most immediately be teaching. But there are other elements, like in-person mentoring and consulting, as long as it's important for it to be in person. Because otherwise, why wouldn't you just ask the AI?

Some roles might remain because we don't want them to be AI — judges and politicians, for example, because we often require them to be in person. Counseling those people might be interesting. But even for them, if it's just about intellectual input, they could use AI more and more. I believe in my own model that most intelligence jobs can eventually be largely replaced, including the purely intelligence side of my own job. That's an uncomfortable thing to sit with.

But if you go back in history and assume nothing cataclysmic happens — if it's just that we have to do different jobs — the artisans who did textiles by hand put immense pride in their expertise. When they couldn't make a living anymore because of factory textiles, that must have been terrible. But in the end, people adapted and did different jobs. It surely wasn't a great feeling to be in the middle of it.

**Danny:** Definitely not. Final question: what should I have asked but didn't?

**Ioana:** I think we really covered a lot of ground, including some surprising topics. Nothing comes to mind right now.

**Danny:** Well, with that, Ioana, thank you so much. This was really fun.

**Ioana:** Thank you.

---

## France, Germany, and the State of Europe – with Joseph de Weck

2026-02-15 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/france-germany-and-the-state-of-europe-with-joseph-de-weck/id1839231559?i=1000749809108) | [Spotify](https://open.spotify.com/episode/0rWn3r3iPWieHyAKxh6LaW?si=f688900e17a243e8)

Germany and France have historically formed the core of Europe and the European Union. Joseph de Weck - political analyst, historian, author, columnist, and Europe Director at Greenmantle - knows both exceptionally well and joins Danny Buerkli to unpack the state of Europe. They discuss the influence of Paul Ricœur on Emmanuel Macron, why French baguette bread is so standardized, what may happen at the German elections in 2029, and the big shifts under way in Europe's geopolitical posture.

**Danny Buerkli:** My guest today is Joseph de Weck. Jo is a historian and senior fellow at the Institut Montaigne. He writes columns for various publications, has published a book about Macron, and serves as the Europe director for Greenmantle, a macroeconomic advisory firm. Jo, welcome.

**Jo de Weck:** Thank you very much for having me.

**Danny:** If we start with France and Macron specifically, what did Emmanuel Macron learn from Paul Ricœur?

**Jo:** Paul Ricœur was a very old man when Macron got to know him. Macron was studying at Sciences Po, one of France's elite universities, and he helped Ricœur edit one of his last books. Ricœur is a special character in the French philosophical pantheon. He's neither left-wing nor right-wing. He's a Protestant, a liberal — a rarity in the French intellectual spectrum. One of his theses was that identity is created through narration, through performative acts — that identity is not something fixed but something that can be made through speech and the use of words.

This is interesting because Macron himself has said that perhaps one of the biggest tasks of his presidency is to create a European identity. He has said that his work is to do ideological work for Europe. He understands that for Europe to come into existence as a polity that can act on the world stage, integrate nation states, and become a geopolitical actor, you need democratic legitimacy. You need an identity that binds people together and makes them agree to have fiscal transfers or act together against a common enemy. He understands that right now there is no true European demos. There are elements of it — we can all feel European in certain instances — but it's incomplete.

If Europe wants to make the big leap forward that Macron envisions, that identity needs to be created. What is really interesting is that he thinks this way, because most European politicians would say their goal is not to do ideology but simply to improve people's well-being — the kind of third-way politics completely void of ideological content. In Germany, even the word ideology is very negatively connoted, because the last time someone tried to do ideology, it went very badly. One of the most interesting aspects of Macron is that he thinks he's in government not only to solve problems — he very much has a technocratic mindset — but also that his job is to build a European identity.

**Danny:** Another person who's written quite a bit about French identity, the question of ascendancy and decline, is Michel Houellebecq — controversially so. What does he get right?

**Jo:** It's very easy to say what he got wrong. He said there was an imminent takeover by an Islamist party in France that would win elections. In his novel *Submission*, the Islamists should have taken over by 2022 or perhaps 2027. We're obviously far from that. It was always a complete fantasy, because anyone who knows the French knows they are completely uncompromising on secularism. France without the croque-monsieur is unthinkable. There would be no France before there was no cheese-ham sandwich.

The idea that France's Muslim population is steadily growing is also wrong. There's always this hyperbole that immigrants would have many more children than the French who have been there before, and that is simply not true. If you look at the statistics, the children of people who moved to France have exactly as many children as everyone else, and polls show that their affiliation to religion is relatively weak.

What Houellebecq gets right in most of his books is that there is a sense of decline in France, especially perhaps for his cohort — white men. The red thread in all his books is a main character who is basically a version of Houellebecq, looking at the world and thinking it was better before. The French like these kinds of stories. There's not only Houellebecq — Virginie Despentes, for example, wrote a very similar trilogy called *Vernon Subutex*, also about decline. The French like this because they've been thinking in terms of decline for a very long time. Peak France was maybe the late nineteenth century, so they've been in this mode for ages, and they're very pessimistic. Polls show that the French are extremely pessimistic compared to other nations about the future of their country, while they are exactly as optimistic about their personal future as any other European. A German is as optimistic about his own future as a Frenchman, but if you ask the French about their country's future, they're pessimistic.

I think Houellebecq caters well to this. There is a demand for a pessimist, depressive narrative. Macron once said of himself that he's a depressive writer, and that in his quality as a depressive writer he really resembles France. I think that's right. In this French depression there's a lot of self-narration, a way of seeing yourself through a pessimism that is a real problem in the country. There's also truth to it — there is a sense of decline that you can see in statistics. But perhaps — and here we're back at Ricœur — the narration is stronger than the reality.

Over the last ten years, on a lot of factors things have gotten better under Macron's presidency. Unemployment is much lower today than when he came into power. Purchasing power has improved — there was a dip with the pandemic, but overall it has improved. There are many indicators that point towards genuine improvement. Not everything has gone badly.

**Danny:** One characteristic of France is its reliance on a very small cadre of elite institutions to produce the leadership of the country — not just political and civil service leadership but also private sector leadership. ENA was the school founded after the sense of French defeat, founded to improve the state of affairs, and Macron closed it down in response to the Gilets Jaunes protests. Is it actually gone?

**Jo:** No, it isn't. It still exists under a different name. The curriculum has changed. People who do this kind of school now have to spend more time in the provinces and less time in Paris. I know people who always wanted to do ENA because it's the royal road, but they always wanted to be in positions of power and responsibility right away — they never wanted to spend time in the provinces doing less glamorous work far from Paris and its cultural life. By changing the curriculum and ensuring students spend more time outside Paris doing less prestigious things before they accede to prestigious posts, the type of people who apply is maybe a bit different. That's a good thing. But the basic problem stays, and Macron is the pure product of it — this idea that you have to build an elite class of enlightened technocrats who can solve problems, deliver for the people, and run the country.

There are many problems, but two are key. The first is that these schools were supposed to be very republican, meritocratic institutions. There is an American dream about being able to rise from bottom to top, and there's a French dream that is very similar — except it's the state, through these schools, rather than the market that allows you to move up. The entry exams are highly competitive. There's no way to buy yourself into these schools. On paper everything is super meritocratic; in reality, not at all. The reproduction of elites is getting worse. In the 1960s and 70s, the narrative of the school as a motor of social ascension was very true; today it is much less so. You can have elites, but it's a real problem when those elites are just the children of past elites. It destroys the French republican dream and undermines the legitimacy of the state and the social model.

The second problem is that French elites have a conception of politics that relies entirely on output legitimacy — the idea that if you get the results right, if pensions are good, if all the KPIs in different policy areas are solid, then you're delivering good politics and the people will reward that. This is a hyper-French conception of politics, one that was dominant in the West more broadly in the 1990s and 2000s. But it completely misunderstands that democracy and politics are also about political participation. As people have become dissatisfied with politics in France, the response has always been to centralize power even more, to give technocratic elites even more competences, in the hope that better delivery would legitimize politics. That has been the wrong response.

Macron is in some ways the climax of this tendency. There's no French president since Charles de Gaulle who has centralized power as much or been as top-down and technocratic. And it has obviously failed in important ways. He managed to get results in many areas — unemployment, the economy — but he has lost a huge part of the population. The far right has risen under his nine years in power, because he misunderstands that politics is about results but also about political participation, giving people a voice, taking them seriously. The Gilets Jaunes protests and other protest movements were about concrete issues, but they were also fundamentally about political participation. As a Swiss person living in France, I can completely understand that. It's hard to live in a country where for ten years one person decides everything, and then you think: the French Revolution — maybe it was about something else as well. He hasn't understood that to this day. And it's not only his problem but the problem of most of France's technocratic elites. That is something France truly has to address.

**Danny:** Historically, what explains this extreme tolerance for centralization?

**Jo:** France is one of the oldest states in Europe. It has a centuries-old history of centralization. The kings were already centralizing politics. You had very early formation of centralized bureaucracies. French as a unified language came into being early. You have a centuries-old history that predates the Republic and the French Revolution, and that said you need a state, you need a bureaucracy, and it has to be run top-down.

There are lots of anecdotes that illustrate this. Take the history of the baguette. Why do we have the baguette in France — one piece of bread, clearly standardized, very recognizable? Whether you're in Corsica or in Brittany, you will always get your baguette, because they standardized exactly how it should look, what kind of wheat is used, and so on. One explanation for this craving for standardization — and with it, excellence and the definition of what is French — comes from economic history. France had a protectionist, mercantilist economic policy even under the kings — Colbertism. One instrument of driving exports was the idea that France should always export excellent products, and that the state should define quality standards for the cloth, the textiles. You had all these standards being imposed while other places had no state enforcing product quality, so products were indeed worse — businessmen would cut corners, add bad textiles to make things cheaper. This standardization comes from an export-oriented policy driven towards excellence, which is funny because it still exists today. France's export economy that works well is the luxury industry — high quality, selling the idea of superior quality.

The other aspect is that the French believe, through the French Revolution and its three core tenets — liberty, fraternity, equality — that the state is the central actor to bring about these promises. And they continue to believe that. It's a paradox I don't quite understand. They are state-centric, and to bring about these promises they're ready to give the state enormous power. We've seen it with anti-terrorism, for example. After the Bataclan attacks, the French state was given sweeping powers to read every email. There is very little sense of data privacy, because the French said: the state has to protect us, so we have to give them everything. To be fair, it worked quite well — there have been very few attacks since. But it shows the reflex: if there's a problem, you call on the state.

**Danny:** If we pivot to Germany — the AfD, the right-wing populist or extremist party, is currently polling around 25%, give or take. That would make it the largest party by share of vote. The elections are a couple of years away, in 2029, but it seems like the system is between a rock and a hard place. The party might get banned, which would disenfranchise a large part of the population, or it might succeed at the polls. How does Germany get out of this conundrum?

**Jo:** I don't know. The thing with the AfD is interesting because the far right in Germany came much later than in other European countries. Most European countries saw a far right emerging in the 1990s or 2000s. Germany never really had that, for historical reasons, and seemed immune to it — but now we realize it is not.

What is unusual is that the AfD was born in the Euro crisis, created by economics professors who were against a deepening of the Eurozone and fiscal integration. It came from an ordoliberal, bourgeois background, and then slowly morphed into a far-right, xenophobic, openly racist, denialist party. The main accelerator was 2015 — the Syrian refugee crisis and Merkel's decision to take in a million Syrians.

Because the AfD formed so much later than most other European far-right parties, it hasn't made the move towards the political center that we see elsewhere. It remains a party that is actually drifting further to the far right. This means that no centrist or center-right party in Germany can work with it, because disagreements are everywhere. On economic policy it's completely unclear what the AfD wants: you have a radical wing that says we need to be Milei-type libertarians, and a more Rassemblement National-style wing that says we need a strong welfare state but only for Germans. On foreign policy, you have people who are pro-Trump and pro-MAGA, and others who are European sovereigntists saying we need to build a Europe that defends European civilization and our enemies are Trump as much as Russia. Some want to abolish NATO and close the Ramstein base. On migration, there are people who want to take away passports from millions of people and expel them from the country — "remigration," they call it. On Israel-Gaza-Palestine, there are people taking extreme positions on both sides within the same party.

It's a party that on most questions has no clear line, which is also reflected in the fact that it has a co-leadership with both presidents holding different opinions. The radical wings are perhaps not winning, but they're not losing either. Those in the AfD who try to drive the party towards the center are not succeeding. They're at around 24–25% in the polls.

Now, the election is only in 2029, and I would say two things. First, the AfD usually does a bit worse than the polls suggest, so perhaps the result is 22%. Second, there is no perspective for a coalition with the CDU by 2029. The AfD's policy program is too incoherent, and most in the CDU simply don't want it. Polls show that 60% of CDU voters say they would no longer vote for the CDU if it entered a coalition with the AfD. If someone from the CDU proposed that, the party would split. So it's not a possibility for 2029, which also means there is no right-wing majority possible. The governing majority will be the CDU with the center-left SPD, and possibly in 2029 with the Greens to get the numbers. You don't get political alternation, and that is a real problem for democracy. It creates further dissatisfaction. That is what worries me.

For Germany, I'm more worried about 2033 than 2029. I don't see a solution to the problem right now. I don't think the AfD is going to go away, but I also think it will never get above 26–27%. So there will never be an AfD government on its own. Perhaps in 2033 there will be an AfD-CDU government, but for this to happen the AfD would need to move further to the center, and the CDU would need to be dominated by its younger generations, who unfortunately are much more right-wing than the older ones.

**Danny:** A naive reading of the situation would say that the CDU is not close enough to the median voter position on migration — that it's not reflecting what voters want. And the naive question would be: why can't they just shift their position? There's plenty of policy space between the CDU's current position and some of what the AfD wants.

**Jo:** The CDU is doing this. There's been a huge shift in Germany and in Europe in general on migration. Trump had two big domestic policy agendas — migration and deregulation. In a sense, Europe has done the same, but even before Trump. Two big things dominate European politics right now: what Brussels calls debureaucratization — massive cutting back of rules, climate legislation, and so on — and a huge change in European policy towards migration. There is no one in government in Europe anymore who is pro-migration, with the exception of Pedro Sánchez in Spain.

In the last year, EU leaders and Parliament have passed reforms on migration and asylum policy that would have been unthinkable two or three years ago. Previously, Olaf Scholz and his coalition with the Greens were blocking a lot of hardening of migration policy in Brussels, and that is gone. Europe's migration policy is really tightening. Everyone is moving towards the Danish model — the Social Democratic Party there and the Labour Party in the UK, but Europe as a whole. We're already seeing it reflected in the statistics: fewer people coming, more money for Frontex, less rights for migrants and asylum seekers, development policy always linked to migration — you get development funds only if the country accepts returns. This is really happening. It's not as extreme as in the US — we don't have ICE in Europe — but we're in full swing.

The question is whether it works politically — whether it has the intended effect of weakening the far right. The Danish example, people would say, did work: the Dansk Folkeparti has been decimated, down to around 5–6%. But there aren't many other examples where it worked. I'm not sure it works in our current environment. Because of the way the media landscape is fragmented — social media and so on — there's a disconnection between the reality of things and the perceived reality. Regardless of how low migration figures are, the AfD will continue to rant about it and lie, and this will resonate with at least around 20% of the population who have a more xenophobic mindset or care more about insider-outsider issues. So I'm not sure we can repeat the Danish success, especially since the Danes went very far with symbolically cruel measures that Germany isn't doing, even if in substance the policy is nearly as harsh.

**Danny:** If we look at Europe more broadly, there's a wonderful line — "everything needs to change so everything can stay the same" — from the *Gattopardo*. Macron has recently said as much, Mario Draghi has been saying the same thing for a very long time. Has anything happened?

**Jo:** Yes, a lot of things are happening in reality. Europe is in a very big shift. The problem as an analyst or journalist is that Europe is the cumulative coming together of many individual, piecemeal decisions. There is no big speech that says "Europe is independent today," because there is no president, no person who has political responsibility for the whole continent. There is no single master plan. But it doesn't mean a lot isn't happening.

As I was saying on migration, politics has changed completely. On defense, massive things are happening — a series of firsts that would have been unthinkable only a couple of years ago. For the first time, the EU has issued common debt to finance defense spending of member states — the so-called SAFE loan, €150 billion distributed to member states in ultra-cheap loans to finance their defense ramp-up. That was completely unthinkable, and it was decided in May 2025 under the impetus of Trump's return to the White House and the questioning of NATO. That we issue common debt to jointly finance a European defense ramp-up is a huge change.

A second first: this money is tied to being spent on European defense producers. 65% of every weapon system bought with this money has to be produced in Europe. Also completely unthinkable a couple of years ago, because previously Europeans saw buying American arms as a kind of premium for American insurance. But since Trump doesn't uphold his end of the bargain anymore, the deal doesn't hold. So Europeans think: we have to spend more, we're not even sure the US security protection will be there, so we might as well spend it on our own industry — it's an impulse to the economy, and it creates capabilities that reduce vulnerabilities and chokepoints towards the US. That Eastern European and Nordic countries all agreed to European preference in arms purchases is a huge change. European preference was always just a French fantasy; now it's becoming reality.

A third first: in December, for the first time, we prolonged sanctions on Russian frozen assets not with unanimity but with a majority decision. That has never been done before. We have understood that clinging to the unanimity principle just blocks and paralyzes Europe, and we have decided that in moments of geopolitical urgency we need to be able to act and be ready to erode that principle.

These are three precedents set just in the last year, all of which would have been unthinkable only a couple of years ago. So yes, things are happening. But there are two concerns. The first: do we have enough time? Europe changes slowly, and people tend to underestimate the cumulative effects of individual decisions. But it always needs a lot of time. Most Europeans say we have to be ready by 2029, but some argue that the build-up of Russia's arms industry is so huge, so beyond what is needed to sustain the war in Ukraine, that it could come earlier. Europe's problem is a timing problem. You see Europe moving towards more capabilities, but the question is whether it's fast enough.

And then on a lot of other things, it's not changing at all.

**Danny:** If someone forced you to deliver the steel-man version — the good-faith version — of the Trump administration's argument against Europe, what would that look like?

**Jo:** Trump was completely right that Europe is not doing enough to defend itself. In that sense he has been successful: European defense spending as a percentage of GDP has doubled since 2015, and we're at the beginning of something larger. By 2030, Germany will be the biggest spender on defense in the world. That's a fact with enormous geopolitical consequences.

But be careful what you wish for. With all these European preference rules and defense procurement changes, Americans are now waking up to this. If you look at the national defense strategy, they're pushing against it, saying it's a problem that only 8% of Germany's procurement goes to US arms. What Trump and Putin have actually created — both of whom agree that Europe as a unified entity is a bad thing, both wanting the EU to fragment so it's easier to play off the parts against each other — is that so far they've actually brought Europeans closer together. We have integrated more, we're doing far more, and we're laying the foundations of an independent European defense, which is something Russia certainly doesn't want, and which the US never really wanted either. US elites always said Europeans don't spend enough, but they were also thinking: it was quite a cheap colony. It wasn't so expensive to make sure Europe always remained dependent on us.

Most of the US foreign policy and national security establishment would say that Trump's policy is backfiring. It has implications not only for defense but for other areas. Think about our dependency on US tech. That was never a problem because we were America's junior partner. We didn't feel it was a vulnerability because we had this deal that we are together no matter what in terms of security. But if that deal doesn't hold, suddenly you start looking at your vulnerabilities in other areas, and that's exactly what is happening. Germany's government is looking for alternatives to Microsoft because they're afraid of it being used as a chokepoint. The French are doing the same. We're at the beginning of a digital sovereignty agenda that would never have existed without Trump.

In ten or twenty years, historians might look back and say this was the beginning of Europe's self-affirmation and long road towards self-determination, with huge implications beyond defense. Or they'll say Europeans didn't manage to get their act together and didn't solve the timing problem. The key year for that question may be next year. 2027 is a super election year in Europe — French elections, Italian elections, Spanish elections, Polish elections. If the far right wins in France — which is perhaps the most important election next year and indeed for the years to come — then this below-the-surface progress in building a common Europe grinds to a halt. The two key elections to watch are France and Poland.

**Danny:** You mentioned the technological dependency of Europe on the US. It seems increasingly clear that developing frontier AI and putting it to good use will be an important — possibly the most important — contributing factor to economic growth and prosperity. While there are European attempts to catch up, the political economy of Europe seems to work against the continent: this is a technology that requires really concentrated investment, whereas the EU decided for political economy reasons to spread its AI gigafactories across the continent. What should Europe do to avoid becoming terminally dependent on the US? Or maybe it just matters less — maybe we just let American capital markets invest a lot of money into developing the technology, and then pick it up once it's mature and ready.

**Jo:** There is an argument for that — that the value AI ultimately creates is in the application of AI, less in the companies developing it today. There's an economic history argument in that direction.

But there's a caveat to the narrative that Europe's economy is really bad and lagging the US. Since 2010, real GDP in Europe lagged the US by about 8%, which is a lot. But if you strip out demographics, Europe lagged the US by only 2.4% — a huge part is just immigration. Foreign-born population in the EU is roughly 9%; in the US it's 15%. So GDP per capita in Europe has grown about 16% since 2010, and in the US about 19%. The gap is not huge.

If you look at what explains that gap, there are four factors. First, Americans work more — roughly 12% more hours than Europeans. Second, Americans have cheaper energy because they've become a net energy exporter — that is a genuine problem for Europe. Demographics and work hours are political preferences. Energy is a problem Europe needs to fix. Third, the American budget deficit is much higher — around 6% of GDP versus the EU average of around 3%. Europeans could also spend more, and that's a big challenge. We're getting there; Germany is making a huge shift in fiscal policy. The fourth factor — only the fourth — is productivity. And this is where the question of technological leadership comes up.

In reality, the gap is much smaller than people think. If you look at which sectors drive higher US productivity, it's the tech sector. The rest of the US economy is often less productive than in Europe, but the tech sector is far more productive. And you can ask: does this really trickle down to the economy as a whole? Much of the productivity gains in the US don't really trickle down because they're often monopolies, so the money goes to shareholders rather than spreading through the economy.

So there's a whole argument that this is less of a problem than we think. I think Europe's prime problem with these technologies is less that we're missing out on growth potential — though we are — and more that we lack capabilities, which makes us geopolitically vulnerable to the US and China. The problem is less about growth and more about sovereignty and capability.

Then the question is how to address that, and as you said, the political economy of Europe is very unfavorable. We don't have EU-wide industrial policy. We don't have EU funds large enough for the kind of subsidization required. We don't have European capital markets sufficiently integrated to allow European companies to get private money in the massive sums needed — the kind of enormous CapEx investment that only works because of American capital markets. We have fragmented markets for the application of AI and technology, with each country having different rules, making it difficult for a startup to become pan-European.

There's a lot Europe needs to do on this front. Using the quote from the *Gattopardo* — in the economic domain, I think it's really true. We see more happening than is apparent from a high-level perspective in defense, but on the economy — full market integration, creating a true European market for capital and digital services — we're not seeing real progress. European industrial policy with both subsidies and buy-European public procurement to create demand — we're seeing some of that, but not enough. This is where the frustration is greatest.

**Danny:** What should we learn from history about falling birth rates?

**Jo:** Falling birth rates are a relatively recent phenomenon in the grand historical narrative. What we Europeans need to learn is to look at Japan — how changing demography has created problems there. What does it mean for an economy to be so old? What does it mean for fiscal transfers and the welfare system? The Japanese answer was ultimately to print a lot of debt and to have very expansive monetary policy. It's not the worst approach, but it creates a lot of problems.

Europe has an advantage over Japan in that despite the migration backlash, we're still more open to migration than they are. Japan is now slowly pivoting on migration, but it's very hard. European societies, despite all their faults, have some capacity to integrate foreigners. It's interesting — look at Meloni, for example. She is completely anti-migration, but at the same time she regularly takes foreigners who already live in Italy illegally and legalizes them. We see that in Greece as well, where a right-wing government is doing the same thing. Within this paradox of a hardening migration policy, you see politicians sometimes doing different things because they have a problem they can't otherwise solve.

**Danny:** What's your theory of history?

**Jo:** There are many theories of history. One I believe in is the law of unintended consequences. For example, did Trump intend, when he invited Justin Trudeau to Mar-a-Lago and completely humiliated him, that he was in reality ensuring Mark Carney would win the Canadian election — someone who at that point had absolutely no shot at winning? Things always play out differently than intended.

The other thing is that most government action is not always part of an overarching strategic framework. As a journalist or analyst, we're always looking at government decision-making trying to find logic, trying to see the red thread, the coherence. A huge mistake to avoid as an analyst is thinking that everything makes sense. It doesn't. Politics is a human affair, and there's a lot of room for misunderstandings, miscommunication, competing interests, competing strategic visions. I have great sympathy for chaos theory in history, to some extent.

But what is interesting about history, and about thinking in terms of historical analogies, is that it is the only discipline that is truly interdisciplinary. To make sense of the world, you have to understand financial markets, macroeconomic developments, societal changes, values and norms shifting, domestic politics, geopolitics — everything comes together. All these factors are different vectors, and there's always an interplay. Sometimes one vector is more powerful than another. History is the study of how in the past all these different vectors came together. By definition, then, you don't have one theory of history — you have many. But that's what makes it exciting.

**Danny:** What should I have asked but didn't?

**Jo:** What is also interesting about the study of history is that as a historian, you study change — that's the subject. And if you look at change over time, you realize that you have these abrupt moments of rapid change, but usually it's a much longer-term process. What we need to understand, especially as Europeans, is that change for Europe comes slowly. We're dissatisfied because we think Europe moves way too slowly. But in a historian's perspective, things are moving nonetheless.

As an analogy: the euro, the common currency, came into existence in 1999. But the starting gun for creating the euro was Richard Nixon killing Bretton Woods in 1971 — the system of fixed exchange rates and the gold standard. Europeans were appalled, saying: the US has become unpredictable, they're unilaterally putting tariffs on us and killing exchange rates, we have to create a common currency if the US is no longer assuring stability. In 1971, everyone started writing think-tank papers about a common currency. But leaders weren't courageous enough — they decided on half measures like the European stability mechanism, technical fixes that never really worked. It took thirty years from the Nixon shock to the euro.

In a sense, I think from the first Trump shock nine years ago to Europe becoming a power that can more or less exist on its own is maybe a thirty-year process as well. Maybe we're ten years into that process. It's slower, it's iterative, it goes back and forth. It doesn't mean at all that an independent Europe will exist in twenty years — it can reverse. But just because it's not here now doesn't mean it won't be there in twenty years.

**Danny:** On that hopeful note, Jo — this was such a pleasure. Thank you so much.

**Jo:** Thank you very much for having me.

---

## Vienna: How the City of Ideas Created the Modern World – with Richard Cockett

2026-01-21 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/vienna-how-the-city-of-ideas-created-the/id1839231559?i=1000746000426) | [Spotify](https://open.spotify.com/episode/1SzpJgINiKw78kn274m5pv?si=a577154f5f1e4875)

Every aspect of modernity was shaped by individuals with intellectual roots in Vienna, argues Richard Cockett, author of 'Vienna: How the City of Ideas Created the Modern World'. Richard, historian and senior editor at The Economist, joins Danny Buerkli to discuss his latest book. They cover the story of the stunningly productive and creative Viennese emigrés, discuss implications for progress today and trace how the intellectual feud between Vienna and Frankfurt still reverberates today.

**Danny Buerkli:** My guest today is Richard Cockett. Richard is a historian, a senior editor at The Economist, and the author of numerous excellent books. His most recent one is called Vienna: How the City of Ideas Created the Modern World. Richard, welcome.

**Richard Cockett:** Thank you very much, Danny. Good to be here.

**Danny:** Richard, what did you learn from your correspondence with Karl Popper?

**Richard:** My correspondence with Karl Popper was very brief and came towards the end of a book I was writing called Thinking the Unthinkable, which was all about Austrian economics and how it influenced the Thatcher governments and the Reagan governments in the United States. I think I was merely asking for permission to quote some letters between him and Hayek, his colleague, and it was sort of a couple of letters. But it was just, yes, I think just before his death, actually.

**Danny:** Oh, wow. So just in the nick of time. Nothing too interesting or revealing?

**Richard:** Nothing too interesting or revealing.

**Danny:** In that book, in Thinking the Unthinkable, you make the point that sometimes outsiders are more astute observers than the insiders. What were you able to see about the Viennese story that others didn't?

**Richard:** I have been told that myself by many Viennese, that only an outsider, only a non-Viennese could have written the book I wrote about Vienna. I think the reason for that—I mean, I think that's true of all places or writers—the best book probably that remains about America is Alexis de Tocqueville's Democracy in America. The best chronicler of modern Italy is the British historian Denis Mack Smith, and the list goes on. And my predecessor writing about Vienna was Carl Schorske. So I'm merely one of a long list, I think.

But I think particularly, the Austrians—or the Viennese, sorry, I should say—have a difficult relationship with modern history, of course, because of the Nazis. You know, Hitler was an Austrian. Nazism as an ideology started in Vienna. Theirs was an outsized contribution to the Holocaust.

So there's a lot of embarrassment and shame mixed with this sort of tawdry excuse, the Anschluss alibi, that Austria in 1938 were the first victims of Nazism because Hitler invaded Austria. And that's always been an out for the sort of general official narrative of Austria, that they were victims rather than perpetrators. So in no other country, I think, is the narrative so mixed—so confused, I should say. So I think I could come along with a much clearer perspective: what was good, what was bad, and how the two sides interacted.

So I think that's what I could bring to the subject. And the other thing, of course, I don't think anyone in Vienna realized the enormous impact all those exiles had on world history. That's what I was most interested in, actually—what happened after 1938. That was my primary interest. So I was working backwards.

If you're a Viennese historian, of course, an Austrian politics writer or whatever, of course you start from Vienna. I started from New York and London and worked backwards. So I think that gave me a very unique perspective on the whole subject, which, of course, the Viennese were extremely grateful for because, again, their relation—many of those exiles, they were Viennese, clearly. But because of the Nazi period, etcetera, many of them didn't want to have anything to do with Vienna ever again. Many of them never came back to the city.

So there, again, there was a complete divorce between the city and its extraordinarily high-achieving diaspora. So in some ways, the book brings the two together again and brings back the legacy of those people who left Vienna to Vienna itself. So I think I'm repeating a lot of what I've been told by Viennese people themselves when I give the numerous book talks I've given about the book in Vienna.

**Danny:** They must have been very excited that someone wrote such a good book, and it is a very good book, I should say. Everyone should read it.

**Richard:** I think they were excited, and it came at a good moment too, when I think the balance between considering the good and the bad is now roughly in equilibrium. So they themselves have a much clearer notion of what went wrong and what went right, and they will celebrate the things they should be celebrating and lamenting the things they should be lamenting. So I think it came at the right time too.

**Danny:** You make the point that Vienna of the early twentieth century and up until the Anschluss was this really phenomenal hotbed of intellectual activity, and you explain the different reasons for what made it so. I'd love to go into some of these, starting with how the conception of the Viennese or the Austrian idea of Bildung—education—was quite different from the German one, interestingly.

**Richard:** Yeah. They were related. I wouldn't say it was wholly different, but Bildung was a German concept, a concept of the Enlightenment, the German Enlightenment, the North German Enlightenment, Protestant Enlightenment of the late eighteenth century, early nineteenth century, associated with Goethe and Schiller and those figures, and musicians too—Beethoven. And it was basically the idea of self-advancement, self-improvement through education, autodidactism, going to university, etcetera. It was a way, if you like, of getting on in the world if you were not part of the aristocracy, of the settled aristocracy.

So it was all wrapped in political liberalism, part of political liberalism—that society could be more fluid, meritocracy, if you like. And the Viennese version of Bildung was a bit more democratic and a bit more meritocratic still than the German. And of course in Vienna, it was inextricably linked with Jewish culture and Jewish self-improvement. So Vienna was a magnet for clever Jewish families from throughout the vast Austro-Hungarian Empire. There were very important, wonderful cities—little Viennas, if you like—dotted throughout the Empire in provinces like long-forgotten provinces like Galicia, which is now modern Western Ukraine.

So there, Kyiv and particularly Lviv, as we call it now, they were hotbeds. Lviv was particularly famous for its mathematicians and lawyers. But generally, if you wanted to get on in the great Austro-Hungarian Empire, you came to Vienna, you went to a gymnasium, and then you went to the famous University of Vienna, which at that point was the most prestigious on the continent and one of the oldest. So that was all part of Bildung, and the Viennese, like their German equivalents, worshipped the gods of Bildung, as they say—Goethe, Schiller, Beethoven. It was a sort of shared culture.

It was very, very strong in Vienna. Austria was a relatively open society—I stress relatively—compared to Germany where the Prussian influence, a very rigid hierarchical system, was stronger. But under the Habsburgs, the Habsburgs were more liberal in the sense that they allowed—they were much more tolerant of different ideas, different outlooks, different dispositions, which meant it was much easier for modernism, cultural artistic modernism, to get a foothold in Vienna than it was, say, in Germany or Berlin. This was before the First World War. After the First World War, of course, you then have the Weimar Republic and Weimar culture.

But before the First World War, Vienna was the leading modernist city. Paris would quibble with that, but I would put the case for Vienna on the continent, and that again drew more people in. So it became a sort of virtuous circle.

**Danny:** How come so much of the important intellectual formation seemed to happen at home rather than in the gymnasiums and universities?

**Richard:** Yeah. I mean, a lot—I mean, the gymnasiums have these amazing reputations as kind of intellectual hothouses. In fact, going through the memoirs of all the people who had to go there, they just talk of the tedium, learning by rote systems in the gymnasiums. I mean, it was a sort of very rigorous education, but entirely uncreative and deadening.

But as you've mentioned, a lot of the most important education went on at home, and this speaks to the autodidactism of the Viennese middle classes and the Jewish—particularly Jewish—thirst for education, for self-improvement, for knowledge. And in my book, I give various snapshots of these Viennese kids in their apartments. And the apartments themselves on the Ringstrasse, the main circular boulevard running around the First District, many of these apartments were like mini universities in themselves in which these kids grew up. So the fathers would often be civil servants, economists, you know, fairly ordinary jobs in the monarchy, but their passions would be biology, zoology, science, electricity, usually something science-related. And the kids were indulged to an extraordinary extent.

So my prime candidate for this is a biologist called Paul Kammerer. And by the time he was twelve or fourteen, he had amassed this extraordinary collection of animals, dead and alive, including crocodiles, in his house, in his apartment, to study. And by the time he was in his teens, he was at the forefront of discovery, of scientific discovery in his chosen field. And he went on to make important contributions to evolutionary biology, particularly a branch of that called epigenetics. It really all started in his bedroom at home.

This was not an uncommon story. And the other thing that was very important was music. I mean, what linked all these people most comprehensively was music, the culture of the salon. There, they would gather for musical evenings, and anyone who aspired to any status in Viennese society would want to host a salon, and these were gatherings at people's apartments, in the drawing rooms, to listen to music, and then to mingle. And there were various famous salons which brought together the most prominent writers, scientists, physicians, anatomists, artists of the day.

And that's where you got this intellectual ferment of crossover, everybody talking to each other and interacting about the latest advances in their fields. And this was incredible. I'll give you one example of this, which is being written about at the moment—the discoveries of these links is ongoing—was the artist Gustav Klimt, who was very well known for great canvases like The Kiss. But he was fascinated by the latest advances, particularly in medicine and particularly in anatomy and biology, and attended lots of lectures on the latest advances in anatomy and biology and incorporated a lot of this, what they were discovering under the microscope at the University of Vienna, into his pictures. All those apparently abstract shapes in famous paintings like The Kiss were in fact detailed copies of what was being published at the time, what they were seeing under the microscope, in terms of spermatozoa and reproductive cells, etcetera, to do with reproduction.

Very, of course, fertile material if you're going to paint pictures like The Kiss, a pair of lovers intertwined. This cross-fertilization was very important. And just one other example: the writer Arthur Schnitzler, best known for his play La Ronde. I mean, he was basically trained as a doctor. His father was a surgeon, a great mentor was a physician.

So he was—and of course he then became a great follower of Freud in psychoanalysis. So he was absolutely steeped in the Viennese milieu of psychology, psychoanalysis, surgery, medicine of the day. I mean, Schnitzler is basically the natural sciences of the day on paper, so possessed was he by all this. So I think it was this remarkable fusion of disciplines, the lack of barriers between them, that made Vienna so remarkable at the time.

Because this was also a time when academia elsewhere was expanding, more colleges were being founded, universities, and at the same time it was all split into these narrow silos which we know and love today, whereby one academic will never speak to another academic. In Vienna it was actually very, very different, in a wonderful way, and it was, if anything, advancing in the opposite direction of making transfer of knowledge as fluid and open as possible. I think that was a huge contribution. And I haven't mentioned the other great locus for this fusion of knowledge: the coffee house. So Vienna, of course, has its famous coffee houses.

All the great names would hold court at one or other coffee house during the day. And so for the price of a cup of melange, you could sit there in the afternoon and hear the great architect of the day or politician, whoever it was, musician, pontificating away on their favorite subject, which of course also encouraged the transference of knowledge and a sort of democratic openness. That's part of Bildung as well.

**Danny:** This is really one of the most striking things, this utter and complete absence of the epistemic boundary policing that we're so used to today. Conceptually, what explains how the Viennese got there? I mean, we can describe that fact.

**Richard:** That fact. Yeah. I mean, I think there's two very obvious explanations. The first is the university education. So the University of Vienna, where most of these people went, only had four faculties at the time.

So it had a law faculty, where you went if you wanted to train in the law, of course. It had a medical and pharmacy faculty, where you went if you wanted to be a doctor, surgeon, or a chemist. It had a theology faculty, which was becoming smaller and smaller, where you went if you wanted to become a priest. And it had a philosophy faculty, the faculty of philosophy. So if you think about that for a minute, that meant that every other subject, if it wasn't law, medicine, or theology, was taught together in the same faculty. So that meant the entire natural sciences, mathematics, science, physics, philosophy of science, etcetera, were taught alongside what we now call ancient history, classical studies, languages, history, economics, philosophy itself, all sorts of philosophy.

So there was this vast range of knowledge in the faculty of philosophy, and students were encouraged to roam throughout all these things because the fundamental purpose of philosophy is to study knowledge, the formation of knowledge, the roots of knowledge. That is what philosophy is about, and that's what they were studying through the lens of all these different subjects. So it's a completely different way of thinking about education than what we're used to today. And again, if you read the memoirs of people like Hayek—Hayek, a philosopher, he won the Nobel Prize for Economics in 1974, he's known as an economist—was brought up at home in an apartment by one of these autodidactic fathers who was so good at biology and zoology he wanted to be a professor of biology. He was a civil servant; he eventually took up a non-stipendiary, non-paying professorship at the University of Vienna.

So the young Hayek grew up steeped in biology, zoology, and basically scientific methodology. And these interests were his main interests for the rest of his life. And at the university, he studied a lot of ancient history, went to lectures on Greek theatre, he studied a lot of law. It was only towards the end that he began to narrow down and look at economics. So he had this vast range of subjects which he felt confident to write on and publish about, and this he did for the rest of his life.

He was very interested in psychology too, as all Viennese at the time were. So when he came over to England or America, the English and Americans found this range of knowledge breathtaking, and they called them polymaths. And the Viennese were often a bit bemused by this. And they said, well, it's just what we learnt at the university. There's nothing particularly special about this.

But what they didn't know is that that education was very special. That was the standard setup at German-speaking universities, these four faculties. So that also occurred in German universities. But I think what also made Vienna special was that there were very few—again, you talk about all our epistemic barriers and boundaries—there were also very few boundaries in Vienna between the university and beyond the university. So many of the great names of that time—Freud, Ludwig von Mises, the founder of the modern Austrian School of Economics—they did not have full-time posts at the university.

Ludwig von Mises, for instance, was the main economist to the Chamber of Commerce for Lower Austria. So he had his office outside the university, but he taught. He was invited in to give lectures to the students, invited by the students. And all these great names held their own salons at home, to which were invited lots of people studying the same subject from the university or not from the university. So you had this very fluid mixture of academics, if you like, people studying the subject outside of academia, and people actually practically involved in the real-world application of all these ideas, all mixing, sharing the same spaces and discussing ideas.

And I think that made Vienna very unique. I mean, the three most famous examples are probably the Wednesday psychoanalytical club of Freud, Ludwig von Mises's fortnightly economic seminar where Hayek and everyone sat around in his office to discuss the economic subject of the day. And there was a famous lawyer, jurist, Hans Kelsen. He had a similar gathering at his salon. Kelsen, again, switched between the university and practical work—for instance, he drafted the 1920 constitution, Austria's First Republic, still in use today.

So again, a mixture of the purely academic and practical applications of those ideas to the Austrian constitution, etcetera. So as I say, there were no boundaries, very fluid movement of ideas and practical applications of those ideas. I think that was a very important aspect of the Viennese experience.

**Danny:** Now, you also talk about this in the book: there's a dark side to this, or everything that you've been describing in a sense is agnostic as to the direction at which this intellectual firepower and all of this creativity is aimed. I think it was Odilo Globocnik, a Nazi party official, who was the first person to come up with the idea of industrial-scale mass murder. And this idea of industrial genocide was in a sense a very Viennese project too.

**Richard:** So all those great names I was just talking about, I mean, they operated in, I guess, what we now call a liberal bubble. So all this intellectual fervor, and they were allowed to do extraordinary things, was protected by the Habsburg Monarchy. That's the irony of all this, that this liberal period was only made possible by the insistence of the autocratic emperor on liberalism. This was gradually changing from 1900 onwards. The reaction against this by what we'd now call ethno-nationalists—people who wanted to reclaim particularly a German culture from what they regarded as the pollution of Jewish and alien cultures—was really gathering steam from 1900 onwards. And certainly a bulwark against that was the emperor, Franz Joseph I, and his insistence on the liberal character of the Austro-Hungarian monarchy at the time—relatively liberal, I should say.

That's true. But once that went, from the 1920s, fascism and Nazism gathered pace much more quickly. But people were less aware of that before 1914. After the First World War in particular, the university quickly became a bastion not of liberal thinking but of conservative thinking, conservative culture, resisting any appointment of women, Jews, and socialists to teaching positions at the university, which meant that an alternative, privately funded intellectual culture flourished in Vienna in the nineteen twenties alongside the official, increasingly conservative culture.

**Danny:** Speaking of partially privately funded culture, what made the Rockefeller Foundation apparently so good at identifying the right people to fund?

**Richard:** Yeah. I mean, the Rockefeller Foundation was amazing. So basically, they stepped in after the First World War. Austria in particular was devastated by the First World War. Vienna was impoverished.

It had mass inflation, famine basically, in Austria from about 1918 to 1921, 1922. It was in a terrible state, so poor that it was not required to pay reparations because everyone knew it had no money to repay reparations. And the Americans stepped in—American aid organizations, for instance, evacuating a lot of the children of Vienna to the countryside or other countries so that they wouldn't starve to death. It was in this atmosphere that the Rockefeller Foundation came looking for talent, not only in Austria but they did this in Germany too and funded projects in Britain as well, for instance at the London School of Economics. But basically what the Rockefeller Foundation was doing was looking for talented individuals frustrated in Austria but who would have a lot to contribute to America.

So they were brilliant at talent-spotting young academics, young intellectuals in all fields—sociology, economics, philosophy, science—and giving them often one-year or two-year travelling scholarships to America or funding posts at American universities. And this was the beginning of that great pipeline of talent that emptied Central Europe to the United States, as well as, to a lesser extent, Britain, which I spent quite a lot of time on in my book. This has continued right up to the present day, although the Viennese now hope that under Trump, only now is it showing signs of going into reverse. That brain drain that started in about 1920 might finally go into reverse as everyone flees Trump, and there are a few signs of that. But this was incredibly important to America.

I mean, amazingly important when you look at how much of modern America in every field—all the achievements of those Central European refugees and exiles contributed to American culture. And I chronicle a lot of that in my book, but there's a lot more on top of that. More often than not, it was the Rockefeller Foundation which started their journey to America. And in the 1930s, this was absolutely vital for many of them because they'd made their contacts in America through the Rockefeller Foundation. So when it was clear that they couldn't go back to Austria or Central Europe because of the rise of the Nazis—and many of these were assimilated Jews—

They had the contacts and often positions in America which they could use or continue from 1938, 1939, the 1940s onwards. And of course, they then advised Rockefeller and other funders on the talent they knew existed in Austria and encouraged those people to join them in America, and posts were found for them in America. So the whole thing snowballed. A remarkably high proportion of Viennese talent got out of Vienna, much more so actually than Germany. I mean, again, to give you one example of that, there was a famous American sociologist called Paul Lazarsfeld who worked in New York for decades after the war.

And he first came to America in the early 1930s, I think late 1920s, on a Rockefeller Foundation scheme. His scheme was extended, etcetera. And then when it became obvious what was going to happen in Austria in the mid-nineteen thirties, he was in a position in America to get a whole load of other people, his friends from Vienna, into America, and he and his patrons got them all jobs, essentially saving them from the Holocaust. And they made extraordinary contributions to American business and culture. Just to give you two examples: one of those people was called Ernst Dichter, and he was really the founder of the modern American advertising industry in terms of applying psychology and psychoanalysis to advertising, marketing, design—an incredibly important figure in mid-century American consumer culture.

Together with his colleague Herta Herzog, who was at one time married to Lazarsfeld—she also invented the focus group, the modern focus group, for instance, and she was known as the Queen of Madison Avenue and again helped to invent the modern advertising industry. She has a walk-on part in Mad Men in the first episode as a little tribute. So, yeah, the Rockefeller Foundation was extraordinarily important at the time. And of course, they had a huge budget relative to the postwar poverty of Central Europe devastated by the First World War.

They had limitless funds to do all this, so it's not surprising that many of the young Viennese availed themselves of the opportunities.

**Danny:** One of the things you also do in the book is contrast the strong Viennese commitment to empiricism and methodological rigor with the Frankfurt School critical theorists—Adorno, Marcuse, and others. And you joke—or maybe not—that the critical theorists only wanted to interpret the world, whereas the Viennese, as you pointed out, because they were engaged in practical endeavors, wanted to change it. But in retrospect, critical theory has been enormously influential, maybe more so in the past couple of decades. So today, who do you assess has actually won that intellectual battle, as it were?

**Richard:** I think this is an underexamined aspect of modern twentieth-century, twenty-first-century intellectual history. And that's partly, of course, because in the West, because the Austrians speak German and the Viennese speak German, a lot of historians and commentators, when they look at Central European diaspora culture, the Viennese were always lumped in with the Germans. I have a stack of books I'm looking at now saying, "saving German-speaking refugees in Europe," "German-speaking culture" and stuff. One of the points of writing the book was to state clearly that this is not the case, that Vienna had a very, very different intellectual culture from, say, Berlin or Germany. And in many ways, interwar culture was a constant dual rivalry between the Viennese—if you like, let's call it the positivist culture—and the idealism of German high culture. This is the Heideggers of the period. The empirical culture of Vienna, let's call it that, was much more akin to the Anglo-American way of thinking, particularly the British way, which is why so many of those Viennese came to Britain and felt very at home in the British intellectual culture.

Popper, Hayek, Gombrich, etcetera. They all came here and felt thoroughly at home here because to an extent, they modeled that culture on the British Enlightenment—Hume, Smith, Mill, etcetera—of the nineteenth century. But you're absolutely right. I'm fascinated by this.

And to a certain extent, as soon as the Frankfurt School was born, they were dueling with the Viennese. So the Viennese philosophers of the Vienna Circle, they were taking on Heidegger and the early versions of the Frankfurt School. And this would continue. To a certain extent, the great works of the Frankfurt School after the Second World War were very much reactions against the empiricism of the Vienna School. So they were daggers drawn. Popper had these great famous debates with the Frankfurt School too.

And so did Lazarsfeld. They all had their run-ins with the dons of the Frankfurt School. So the Frankfurt School were basically—I mean, it's a toss-up, isn't it? Because the Frankfurt School were basically interested in society as a culture. That was their main point.

And I suppose, you know, the best, the most effective modern inheritors of the Frankfurt School in a way are the alt-right of America, because it was Steve Bannon and his acolytes who coined the phrase, "all politics is downstream from culture." And that's basically what the Frankfurt School was saying from the 1960s. So all those students on the barricades in 1968 were carrying their little—well, they carried Mao, but they carried their little Frankfurt School pamphlets saying, you know, the basic argument: yes, we might be drowning in material riches, but our culture is still authoritarian and Nazified, etcetera, and this is not true freedom. Whereas many of the Austrians, particularly the Austrian School of Economics, of course, really bore down and concentrated on political liberalism and economics as a way of raising prosperity, and that was the best way of ensuring the survival of a capitalist, open capitalist—to use Popper's phrase—the open society. So there were absolutely clear dividing lines here.

Now who's winning or who won? I mean, it's a contest, isn't it? Because I would argue that in today's world, until Trump came along, the economic liberal consensus that Hayek started fighting for in 1947 with the founding of his first think tank, the Mont Pelerin Society—and that's an open, liberal world economy protected by rules like those of the World Trade Organization. Most of these institutions were created by scions of the Austrian School or ordoliberals of the German school. That was very much the dominant mode of economic arrangements in the world. They had vanquished communism, they had vanquished socialism, they had vanquished protectionism.

So in that sense, I think they won. Culturally, of course, the Frankfurt School were probably much more significant in terms of being critical of consumer culture, of political notions of the Austrian school. So you had—and they were, in some ways, the two meshed in the late 1960s into schools such as libertarianism, which combined—and this is the irony—which combined the economic liberalism, the economic freedom advocated by the Austrian School of Economics with the New Left, Frankfurt School's politics of liberation, in terms of sexual freedom, sexual orientation. Of course, that all feeds through now to the great trans debates, because the trans activists, very steeped in their New Left Frankfurt School rhetoric, advocate that everyone should have the right, the personal right, to decide their sexual orientation, indeed what sex they want to be, when they should change, etcetera, etcetera, and transgressing societal norms. That is the job of the vanguard of the Frankfurt School and philosophers, is to exactly transgress those societal norms, thus moving society on in a progressive direction.

So you can absolutely see where the Frankfurt School and the New Left have been very influential, but at the same time, so has the Austrian School and the empirical school. I think that's why there's such alarm about Trump: he's the guy who's come along and upended all the assumptions of the economic liberals and the Austrian school, the empiricists, going right back to von Mises when he started inventing the whole thing in 1920. So at the moment, we're at an interesting tipping point because, of course, the populist right, they reject both the New Left and the Frankfurt School, their traditional conservatism back then, and the economic liberalism of the Austrian school. I mean, they reject the science-based empiricism, evidence-led stuff of the Vienna school as well, which is why the populist right is, I think, such a novelty. It's innovative in that sense, and it's so powerful.

So that's how I interpret the grand flows of ideas. But I think the Austrian School found its foothold in Western culture because it stood very clearly for that empirical, liberal, economic, empirical culture that was threatened by fascism and communism, etcetera, in the nineteen thirties. Does that all make sense?

**Danny:** It does. And it strikes me that, one, you've talked about some of the—if you wish to call it that way—sort of failure modes, if you will, of the Frankfurt School. I think there's also an undercovered, maybe, failure mode of the empiricist school where empiricism veers into scientism negatively understood, or cargo cult science, the way that Richard Feynman has described it. And it strikes me that one of the fascinating things about the Viennese scientific culture was how scholarly it was. And it strikes me that today, we could occasionally do with a bit more scholarliness and maybe a bit less emulation of scientific rigor, which then may not end up being particularly rigorous, in fact.

**Richard:** Yes. Popper himself was a great critic, of course, of the Vienna Circle philosophers. So the Vienna Circle was famous, important—this is Wittgenstein, etcetera—for importing empiricism into philosophy and thinking. And they wanted rigorous—they wanted an exactitude in language whereby everyone saw absolute clarity in conversation, etcetera. But as members of the Vienna Circle realized at the time, with the growth of fascism and Nazism, you can't really talk about morality, evil, all these things with any exactitude.

And if you insisted on exactitude of language and ideology, then you couldn't talk. You had no language to talk about all these things that actually seemed to matter very much to most people. I would call that now the language of feeling. And this is where—this was the gap through which mass movements such as Nazism came. And indeed Trump comes—that what they are good at is articulating people's feelings, people's resentments, etcetera.

And it's astonishing to people like me, because it's still astonishing, I'm afraid, I'm saying that Trump can get up and he just makes these statements. I say, well, none of that is true. If you look at the evidence, and we've tested this in the field, and of course, like everyone else does, and of course, it doesn't seem to matter. It doesn't matter if you inhabit the language of feeling, not the empirical. Well, then actually none of that matters. And what's more important, the language of feeling is quite legitimate, it seems.

Empiricism is less important than the language of feeling, of evidence-led argument, etcetera. So those clashes still go on. Those were the clashes that the Vienna School were most involved in with the German school, with the German idealists and the Frankfurt School. My friend Anthony Gottlieb just published a new study of Wittgenstein. It goes back to why Wittgenstein was so fascinating: if you go back to his first work, the Tractatus Logico-Philosophicus, proposition seven, isn't it?

Which there's always—I mean, he's supposed to be the epitome of the rigorous, hard, exact thinker trying to make language exact and order the whole of civilization and existence. But at the end, proposition seven says, you know, well, what we cannot speak about, we must pass over in silence. And it's that proposition, it's that thought, that is in fact the most famous Wittgensteinian thought, not the previous thousand-something propositions on how to make language more exact. And that, I think, is the dilemma, because that proposition seven leaves the door open to any old rabble-rousers to talk about the language of feeling. They began—and that was evident in Wittgenstein's own time in Austria, where you had the beginnings of fascism, Nazism, with the early antisemitic politicians in Vienna, like the famous mayor Karl Lueger, who was the first European politician to weaponize antisemitism and make it into a political culture.

And he had, by all accounts, plenty of Jewish friends. So in some ways, he could say, it's not personal, but I know that many of the people who will vote for me bitterly resent the wealth and privilege amassed by wealthy Jewish industrial families, like the Wittgensteins indeed, and I will articulate their resentments, and they will put me in city hall. That's the way it works.

**Danny:** Now, that is one contemporary parallel that you've picked up now. Right? One reading of your book is the parallel with the assault on critical rationalism, which we are seeing today. There's another parallel. There's another way of reading it or another question to ask, in a sense, which is: where is the Vienna of today?

Of course. And it's an obvious question. One of the obvious answers will be, well, probably somewhere in California and possibly in the San Francisco Bay Area, which, of course, also has benefited from Viennese emigres going there. What would your answer be? If someone was looking for the Vienna of today, where would one find it?

**Richard:** When I'm asked that, I say exactly what you've just said: somewhere near Palo Alto. Yes, I mean obviously, I think Silicon Valley would be a modern type of Vienna where it sucked in talent from all over the world, and you have this concentration, critical mass, lots of people—scientists, designers. Apple is a very good example of that, of this melange, this mix of scientists, designers, artists, mathematicians, algorithm makers, and pure entrepreneurs, meshed together, producing these extraordinary ideas, in this case machines, which have undoubtedly transformed the world at a rush in the past twenty years. So yes, absolutely. For Viennese, interestingly—now I go to Vienna quite a lot. They're very interested. I get a lot of questions like, can we recreate this Viennese culture in the Vienna of today?

Some people have come back. The Viennese diaspora, as I said before, have started returning to try and do that. They sit in California in their traffic jams in LA thinking, why am I—I could be back in nice Vienna in the clear air of the mountains and having a nice calm journey on my very good Autobahn, not being clogged up on the freeways of California, and doing my science there. Some of them have chosen to do that, and I think it would be marvelous if they could, but Silicon Valley has a great critical mass. But the great threat to that, as we all know, is the American president himself, who—because he's a populist rightist—doesn't get the importance of knowledge transfer, of why immigration is so important.

If he paused for a minute and looked at how much of modern America has been built by, first, European immigrants and then much wider—Indian immigrants now, Chinese immigrants, South Korean immigrants—that should give him pause for thought, because if you choke off that flow of talent—Vienna, the creation of Vienna, was all about flow of talent in the Austro-Hungarian Empire. No passports, complete freedom of movement, anyone could come there, and they did. That overwhelmingly is what made that culture of Vienna. And up to not so long ago, Silicon Valley benefited from the same sort of flows, a bit more difficult, but now Trump is, of course, tightening up the vital visas. The H-1B visa—tightening the rules for H-1B—and that's the particular one you need to make America great again.

And so it's a sort of policy of self-sabotage. But of course, going back to which political dispositions or beliefs you prioritize, that particular bit is not a priority because the populist right, they're culture warriors. They're New Lefties in their inverted universe. They're Frankfurt School politicians in the sense that they're culture warriors, and that takes priority over economics and over economic growth and economic activity. So that sort of makes perfect sense.

But I think, not to be too gloomy, I think San Francisco and the corridor there, of course, is so huge and at the moment it's had such an overwhelming preponderance of talent and inventiveness that I think you'd really have to double down to destroy it or at least stop it, perhaps. But no, I would say the same as you. I mean, India could be, with that sort of extraordinary talent. But, again, most Indians go abroad. It hasn't got the knack yet of attracting people to come and work in those hotspots like Bengaluru.

**Danny:** What should I have asked but didn't?

**Richard:** I think you've been pretty comprehensive. I mean, the only thing I would say, again, is—for somebody, one reviewer of my Vienna book said it's both a warning and an inspiration. Those are the two sides. One of the important lessons for me of Vienna culture, this glorious culture of the beginning of the twentieth century into the twenties and thirties, is that very—I was going to say very few—nobody saw the Holocaust coming. Nobody saw the reality of Nazism coming.

From the nineteen twenties, many of these great progressive minds, they were not getting jobs at the university, they were not getting the opportunities they would have got a generation or two earlier. For as Austro-fascism and fascism were taking hold of Central Europe, they would be told sometimes, you know, well, I'm afraid it's antisemitism at the university, you're Jewish. And they often refused to believe it and said, but this is Austria, land of the Austro-Hungarian Empire. This is the country that has done the most for Jewish culture, for Jews, of anywhere in Europe over the past fifty, seventy years. And many of them—so the scales would fall from their eyes far too late.

And it's amazing, in retrospect, how quickly things turned against them—that this culture, this extraordinary pre-First World War culture, capsized within one generation, where the capital city that had produced this glittering, progressive, humane, modernist culture was absorbed into the greater Germany and was run by Nazis only a generation later. So I think there's a lesson in there that you have to keep your eyes open and the two sides have to talk to each other. The liberal empiricists, etcetera, knew nothing of what the other side was thinking. They were very dismissive of it. Whereas the fascists, etcetera, because they had no way into the dominant liberal culture at the time, their only reaction therefore could be resentment, hatred, jealousy.

All these—I go back to the politics of feelings—that fed Austro-fascism, Nazism, and eventually the Holocaust. So it is a cautionary tale as well that if you don't give—provide these sorts of cultures with solid constitutional political guardrails, then they can be threatened very quickly. You have to be very aware that things can change very quickly. So you need that sort of deep institutional grounding. And of course, America is now the test case for all this.

With Trump there, he wants to tear up all the rules. Will the guardrails, to a degree, limit the damage?

**Danny:** We will find out. Finally, which book are you writing next?

**Richard:** Well, actually, I'm going to Vienna in a couple of months. I'm taking up a fellowship at the IWM in Vienna. And having thought quite a lot about the populist right, I'm actually writing a book on the intellectual roots, cultural roots of the populist right, the history of the populist right, how we got here, where I do want to delve more deeply into exactly those contrasts between empiricism and the politics of feeling, where ethno-nationalism came from, etcetera, etcetera. But mine is a world study, so I want to look at not just America, more obviously Europe—you know, Hungary and Italy, etcetera—but also India, Russia in particular, with the breakup of the Soviet Union and elsewhere, and see who copied what from whom.

So that is the next book.

**Danny:** I'll be sure to read that. Thank you very much, Richard.

**Richard:** Thank you very much, Danny. It was great. I really enjoyed our conversation.

**Danny:** Thanks for listening to High Variance. You can subscribe to this podcast on Apple Podcasts, Spotify, or wherever you get your podcasts. If you like this podcast, please give us a rating and leave a review. This makes a big difference, particularly for newer podcasts like this one.

---

## Labor Market Impacts of AI – with Bharat Chandar

2025-12-29 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/labor-market-impacts-of-ai-with-bharat-chandar/id1839231559?i=1000743018739) | [Spotify](https://open.spotify.com/episode/2ubJu0uQeTsuZO22QvurCQ?si=47b7f150ed694c86)

Which effects of AI are we already seeing in the labor market? And what might be coming down the line? Bharat Chandar, postdoc at Stanford and co-author of the 'Canaries in the Coal Mine' paper, joins Danny Buerkli to discuss what we know about the impacts of AI on the labor market and where the jury is still out.

**Danny Buerkli:** My guest today is Bharat Chandar. Bharat's a labor economist. He's a postdoc at the Stanford Digital Economy Lab and got his PhD from Stanford GSB. Bharat is one of the surprisingly few economists studying the labor market impacts of AI, which is why I'm delighted to have you on. Bharat, welcome.

**Bharat Chandar:** Thank you. Excited to chat.

**Danny:** What do we know about the impact of AI on labor markets today?

**Bharat:** Great. So there's still not a ton of research on this topic. I can say that we know that in the United States, overall, we are not seeing major disruption being caused by AI. So if you look across the entire economy, you use kind of our standard government data sets for trying to study these questions. In my own work and work by others, other academics, we basically found that if you kind of zoom out and you look across the economy, the jobs that are more exposed to AI, they're not seeing major disruptions being caused by the technology so far, at least it seems.

They're not seeing substantially different trends in terms of employment or wages and things like that. Now in our paper, Canaries in the Coal Mine with Eric Brynjolfsson and Ray Chen, we try to dig into this a little bit more. There's been a lot of discussion in particular about entry level jobs and how those might be having different kinds of adverse trends. And there was kind of speculation over the summer, especially about whether some of those trends might be being caused by AI. But there wasn't a lot of evidence to try to support one way or another whether that was the case.

So we wanted to bring some data to this. And the way we did that is we have this partnership with a company called ADP, which is the largest payroll software provider in the United States. And with this partnership, we were able to track employment for millions of workers across the United States, across different jobs, pretty much spanning industries across the United States. So it's a very large data set where we could really dig into the specific occupations that we might think are more exposed to AI. But then going further, we could also dig into certain age groups within those occupations that there's been this discussion about adverse impacts for and kind of tracking the data, whether that's actually bearing out in practice.

So when we looked at the data, we found that we were in fact seeing that these jobs that were more exposed to AI, such as software development, customer service, etcetera, there has been a slowdown in entry level hiring. And that's not the case for more senior workers in those roles. And it's also not the case for entry level workers in roles that are not as exposed to AI. One example that we have in the paper is about home health aides. So this is a job that you would think is not very exposed to AI because there's a lot of in person interaction, physical interaction, talking to the patient, etcetera.

So that's a job where in fact we're seeing faster employment growth for young workers than for more experienced workers. So overall, that narrative that these jobs are seeing a slowdown in the entry level labor market does in fact seem to be showing up. Now, there's certainly questions about whether that's being caused by AI. And we tried, to the best of our ability, to test different alternatives that we thought could plausibly explain some of these trends. So one example was tech overhiring.

So there has been this idea that during the pandemic, tech companies hire too many people, and now they're kind of trying to reduce that overload in their workforce. But we got similar results if we took out the tech sector, if we took out any computer jobs, any coding jobs, so not even just software development. There was also this idea that it could be driven by interest rate changes. And I think that's certainly one potential explanation. So the way we looked at that is we looked at how exposed different occupations are to interest rates.

And it turns out that AI exposure is actually negatively correlated with interest rate exposure, and one way to think about that is that there are jobs such as in construction or transportation that are very interest rate exposed, but are not very AI exposed. And so you can kind of cut it between jobs that are more or less interest rate exposed, and you get the same results there. There's an idea that it's driven by outsourcing. You get the same results for jobs that are teleworkable versus not teleworkable. And another thing I'll mention is there was this idea that maybe it could be driven by education disruptions during COVID.

So obviously, schools moved online, but it turns out that you get the same results for college graduates and jobs that don't involve a college degree. And what's striking about that is that for the non-college graduates, you actually see it even at higher age groups. So up to age 40, you're kind of seeing these disruptions. So that's suggesting that it's not necessarily being driven by education either. That's not to say that we proved one way or another that it's being caused by AI, but we tested some of these alternatives and we were seeing the pattern still held.

And I think there's certainly scope for more research here, and we're lucky that we're starting to see that a little bit now. One example of that is a paper by Hosseini and Lichtenger. They're two grad students at Harvard who use data from Revelio Labs. It's essentially like LinkedIn data. And they look at companies that put out a posting that references generative AI and implementing it within the company.

And then they look at companies that don't do that. And they find very similar patterns to us. Belki Klein-Tussink at King's College in London also finds a similar pattern in the UK. Now there's one paper that kinda goes against that, which is this paper by Humlum and Vestergaard, which is interesting. They're researchers in Denmark, and they're looking at companies that adopt versus don't adopt the technology, and they actually find a different pattern.

They don't find any differential trend in employment over there. So I think we're still in the early stages of trying to figure this out. And I think one thing that we can try to do is try to reconcile these different results that are showing up in the literature and what might be driving some of that. I think we want better adoption measures. And then just like, is this being driven by AI?

Are there other explanations that we're not accounting for? What might those be? But I think this trend about this slowdown in entry level employment, and especially in jobs that are more exposed to AI seems pretty robust and has been replicated in different places.

**Danny:** And with the exception of the Humlum and Vestergaard study that you mentioned, everything you've listed seems quite coherent also with other studies that look at the macro effects where we, for instance, the one that Molly Kinder and the Yale Budget Lab did that shows no aggregate effect in the US labor market, which is perfectly compatible, I believe, with what you found, which is that there are some effects in some very specific pockets of the labor market.

**Bharat:** Yeah. This is an important point. So we should mention that if we're looking at 22 to 25 year olds, that's a pretty small share of the workforce. And so we can see pretty concentrated impacts for those workers in jobs that are more exposed to AI. But then when we zoom out across the economy, that's not necessarily gonna appear, especially if we're looking at these more aggregated government statistics. Like, how much of this can we separate from just noise and who's getting sampled versus an actual impact?

Once we're aggregating it up to that level, we might not be able to kinda tease those apart.

**Danny:** What was your prior going into this?

**Bharat:** That's a good question. So I wrote a paper in May or so where I was doing this kind of exercise of zooming out across the economy and using the government data to try to compare employment in more versus less exposed jobs. And I got results that were very similar to what you're mentioning from the Molly Kinder paper, where I was not finding these differential trends in employment across more or less exposed occupations. But I think that didn't get the question about the entry level workers, which is where a lot of the narrative was and a lot of the discussion was. And we just didn't have the sample size or the reliability in the data to try to get at that question as well.

So that's why we wanted to try to take a larger dataset where we could try to get at some of these questions around entry level employment, especially in AI exposed jobs. And so going in, I guess I didn't have much information to try to form an opinion about this. There were discussions in the media, different narratives about whether this was or was not being driven by AI, but I don't think there was good data on, one way or another, about how employment was actually changing in these jobs that we could reliably trust to get an estimate of that. And so that's where we wanted to bring in the data to get a sense of that. So I would say before we wrote the paper, I was just very uncertain about what was happening with that specific group.

But I do think that I had some amount of confidence that overall across the economy, we weren't seeing major changes.

**Danny:** If you were to deliver a steelman critique as it were of the Canaries in the Coal Mine paper, what would that look like?

**Bharat:** I think the steelman critique is just that we don't have an experiment where some companies—I mean, it's actually pretty tricky. So you can't even necessarily say we need an experiment where some companies use AI and some companies don't. That might get you some of the way there where you can see these companies that are implementing it, like how is it changing their hiring? But I think the problem with that is that there's also anticipation effects. So it could be that maybe I haven't fully fleshed out my AI strategy at the moment within the company, but I'm anticipating that I'm going to be doing that in the future.

And I don't want to hire a bunch of people right now who in a year or two, I might not necessarily want to retain because of these changes in AI that I'm anticipating that I'm going to implement that are going to make it so that I don't necessarily need to have those people on the payroll anymore. So that's another thing that makes this tricky. Like, I do think it would help a lot to have better adoption and there have been some innovations on that front. So whether it's via job postings, I think earnings call records and how people are talking about AI adoption in those, I think is also a good way to go about this. Other adoption measures that might be based at the firm or whatever it is.

And then there's also survey-based estimates. So that's what Humlum and Vestergaard do, where they ask employees who are using AI. So I do think that there are innovations happening on that front about measuring firm adoption better. And I think that could go a long way towards addressing some of this, but there are other factors involved too that won't necessarily be solved by that, including this anticipation issue.

**Danny:** Now your paper has an evocative title, Canaries in the Coal Mine. If the effect we see on young workers is real, which other effects would you expect to see in the future?

**Bharat:** Well, I think there's a lot of uncertainty about what we might expect going forward in the future. So it could be that right now we're in an adjustment period where firms are getting used to the technology and they're making a lot of investments to figure out how to use it properly. And once they do that, they'll potentially reverse course and start hiring these young workers again, especially since they might have greater capacity for adjusting to these technologies in the short run and knowing how to use it and things like that. So that's one possibility. Another possibility is that the capabilities, which are accelerating very quickly, the AI technologies today are very different than they were in 2022, and they're much more powerful now.

That could mean that we could see greater impacts on more senior workers. So the effects might not just be concentrated on entry level workers going forward. So there are a variety of scenarios that could play out here, and uncertainty is mostly around both adoption and also model capabilities and what kind of tasks that they'll be able to perform going forward. So that makes it difficult to assess, are these trends that are going to continue? So far, we haven't seen any reversal.

I recently updated the data through November, and we didn't see any reversal. So at least insofar as this is more than three years out since ChatGPT was released, and those trends are just kind of continuing. So insofar as that tells you anything, but in one year, in two years, are we gonna see the same trends? I think there's a lot of uncertainty around that.

**Danny:** Now if we think about the aggregate effects, the aggregate potential future effects of AI, if you had to place yourself somewhere between Anton Korinek, "things will go wild," and Daron Acemoglu, "maybe not so much to see here," where would you place yourself?

**Bharat:** I would place myself somewhere in the middle of that range. I do think that economists probably are underestimating the growth in capabilities over time, and there are reasons that might be the case. Maybe they don't fit as well within our existing models, and I think there's also hesitancy about making predictions about future capabilities. That's not a typical thing that I think economists do very often. So that might be contributing to that.

But I do think that we're seeing these massive improvements in model capabilities. I use AI very differently today than I used it a few years ago. So I think we should be cognizant of that. Now, I do think that the economists have some good points around bottlenecks in the economy, areas where it's going to be more difficult to roll out these technologies. And there could be a variety of reasons that that's the case.

There are regulatory reasons or cultural reasons within the firm that might prevent them from using these technologies. There could be slowdowns driven by challenges with implementation. So what are the data requirements that you need? What are the security guarantees? What are the privacy guarantees?

I do think that there are certain sectors of the economy where we're going to not see as fast of an improvement. And of course, there's all the industries that require physical or in person interaction and things like that. So I do think we want to take that seriously. Maybe robotics will change that and there will be a much larger scope of industries and occupations that might be exposed going forward. So I do think I'm sort of in the middle here where I take seriously this idea that there could be serious bottlenecks, but I also am cognizant that there are massive improvements happening to capabilities.

**Danny:** And how do you think through how improvements in capabilities translate into diffusion and then some form of labor market impact and then presumably some form of aggregate impact? Because presumably, diffusion is also on some level a function of capabilities themselves. It's not just a purely exogenous variable.

**Bharat:** I think the first person I saw make this point is Jonathan Maslisch, who is a growth economist who also studies some of these questions. It's a great point that adoption is also a function of capabilities. So the more things that these models are able to do, the more possibility there is for adoption. And maybe that's one reason that we're seeing very fast adoption with AI. One is the general purpose technology, but also there are increasing applications that are being developed.

Multimodal is much further along than it was before. Agentic coding is much better than the technologies that I think we had initially with the release of AI. It's being integrated into codebases better. So that's just one example. I do think that we could see the potential for much more applications that are directly relevant to work.

And I think one example of that is the GDP val set of results that are coming out of OpenAI. So they're trying to look at real world tasks and how well the models can perform in completing those tasks. And so they look at different occupations and what are the most common things that people do in these occupations? What are typical workflows? And can the model produce something that kind of an unbiased observer would prefer relative to a human output at that same task?

And what was striking about that is that the main failure point for the models in performing those tasks was actually in just the output that they were producing. So if you wanted a spreadsheet, the models weren't very good at creating a spreadsheet or a slide deck or whatever it was. And it turns out that I was just looking at this. I think if you look at the more recent models, they kind of took this to heart in the development, and they tried to improve the output that was coming out of these models so that they better match requirements for these tasks. And if you look at the ChatGPT 5.2 or Gemini three or the most recent version of Claude, they're doing a much better job of producing output that is what these tasks are looking for, whether that's a spreadsheet, etcetera.

And so I think as people become more cognizant of what the actual requirements of the work are, and I think we're moving in that direction and people are taking this very seriously, we could see a broader range of applications going forward, which could also drive adoption.

**Danny:** What you're pointing towards is this idea that we keep shifting our model of what we think the binding constraint is in terms of these models doing really economically useful work. For a while it was maybe things like factual recall, and then it was things like, well, the ability to produce a slide deck or spreadsheet with actual formulas that work. Some people think it's something like continuous learning. And this may be true, and it may be one version of the world is we at some point, we aggregate enough of these capabilities and that unlocks everything. But historically, it would seem that every time we figure out one thing, we realize that there's yet another thing that we would need.

Where do you come down in that debate?

**Bharat:** Yeah. I think that there's actually something that I'm trying to figure out whether it's a fundamental bottleneck or not. So I think right now, the way that we interact with AI, especially with the agents, is we're increasingly, at least I am, we're increasingly developing these more complicated prompts to specify what it is that we want the model to produce. And especially for longer horizon, more complicated tasks, we need greater specificity for what we want to convey to the model about what it should be producing. It can't necessarily come up with that on its own.

It needs guidance on what it is that we want the model to do. And then it's amazing at the execution. But right now, we still need to specify what it is we want. And I think in practice, the way this usually works is we specify something, we put in a prompt that's maybe not as detailed as it needs to be. We get some output, we observe the output, we iterate on it, we improve aspects that it didn't do as good of a job of implementing what it is that we wanted in our mind, and that could be because we didn't explain it well enough.

And then we kind of do that iterative process. And that process is mostly not necessarily about the execution itself, but more about us trying to communicate what it is that we want to the model. And I think the question is, is that going to be the way that things are going to be forever? So is there just an inherent bottleneck that's caused by us needing to specify what our preferences are to the model so that it can do the execution well. And unless we can exactly explain what it is that we want, it can produce whatever it's capable of producing, but it needs to be something that's useful to us.

And that requires a lot of manual intervention on our part, which is just literally us expressing our preferences, or even clarifying what our preferences are. When I write my initial prompt, I may not actually have a good idea of what it is that I want the model to produce. And part of that iteration is me understanding better what it is that I want. And it's not immediately obvious to me how to solve that problem. Like maybe the models get a better sense of who I am as a person and what my preferences are, but those are also moving over time.

So I am often confused by this, and I don't know to what extent we're going to be able to reduce this bottleneck, which is us just communicating with the models and specifying what we want. Annie Liang—she's a professor at Northwestern—has a recent interesting paper about this where she considers theoretically this idea that the models may not have a good sense of what it is that we want them to produce. And she gives the example about matching in the marriage market or the dating market. So I may have certain attributes about a partner that I want to communicate to the model.

And then I want it to go out, go on the dating app and then find people who match those preferences. But the exact preferences over the partner that I want are super complicated. They're not easy to communicate well. I may need some time to iterate on that and refine what it is that I want. And so if I were just sitting here and I was telling the agent, oh, this is what I'm looking for in a partner, it would not do a very good job of actually picking out that person because there's a lot that I'm not communicating to the agent that are extremely relevant to who I would actually want to match with.

And so that, I think that's a good example of where this bottleneck is coming from and why it's going to be, I think, pretty tricky to solve. But I'm not on the model development side and maybe they're thinking very carefully about this and how to solve that problem. But at least for me, kind of backing up, I'm not exactly sure how to solve that bottleneck.

**Danny:** Would that be—the example you gave, would that be because your preferences are too high-dimensional and therefore too difficult to express explicitly?

**Bharat:** Exactly. Yeah. So it's specifically about the dimensionality. In her work, that is the key force that makes it difficult to offload the preferences to the model and just let it do the execution. So for simpler problems where I can easily communicate what it is I want, and I just need the model to go solve this very simple problem, that is much more possible now and going forward.

But if it's something where I need to communicate something that's quite a bit more difficult, then that could make it more challenging to kind of offload the execution and actually get something that I'm looking for.

**Danny:** The counterexample would be, well, human matchmakers exist, but the counter to that in turn would be, well, yes, but they were a fairly small market, it would seem. So maybe that is an inherent limitation.

**Bharat:** Well, the matchmaker exists, but you still go through hours and hours of dating to figure out, is this the person that I want to spend my life with? So

**Danny:** I was thinking of human matchmakers. Right? There are marriage markets where where humans do the matchmaking. Need not be marriages.

**Bharat:** So that's correct. They can help you find potential candidates for you to then invest the time in kind of evaluating whether this is the person that I wanna match with. But I still have to invest all that time, and the matchmaker doesn't exactly know what I want.

**Danny:** Fair enough. What you're also getting at is this fundamental question of automation versus augmentation, and there is an argument that exists that says, well, with all these evals, specifically GDP val and others, we're in fact incentivizing the wrong thing. We're in a sense incentivizing labs to build automation capabilities, whereas we might actually prefer them to build augmentation capabilities. The counterargument to that is to say, well, LLM capabilities seem highly, highly correlated. The model that's best at math is also likely to be the best at legal advice, etcetera.

And that in turn implies that we may not actually be able to differentiate between increasing augmenting capabilities versus increasing automating capabilities because they're ultimately one and the same? How do you think about that problem?

**Bharat:** Right. I think this is a great question, and it's something that I have been thinking a lot about. So can you direct the technologies to be more augmentative than automated? And I think the main point that kind of gives me pause here, or at least uncertainty about to what extent this is possible, is exactly what you're talking about. So the LLM capabilities are extremely correlated.

Models that are very good at math are also very good at other things. And so an implication of that—and I think Tom Cunningham, who is now at NBER, is someone who has been very provocative in my thinking on this front.

**Danny:** And I owe this point to him, I should say.

**Bharat:** Yes. Yes. So he, I think, makes a great point that if the model capabilities are extremely correlated in this way, it's going to be more difficult to direct them in whatever direction we wanted to go that we think is more socially valuable. So indirectly, an implication of that is it might be more difficult to direct it in an augmentative direction instead of an automated direction. Now, I'm actually working on an essay right now that I'm hoping to release pretty soon, but I do think that there's one case where we could certainly be investing more to make the technology more augmentative kind of by design, which is developing the technology in a way that makes it more conducive to learning.

So when we think about augmentation, we're basically talking about areas where the technology could increase human capabilities. So we're able to do more stuff with the technology than we were able to do before. And I think a great example of where we can see augmentation and an increase in these capabilities is by improving how well the technology enables us to learn. I think the history of the twentieth century kind of gives a great example of that. If you look at the beginning of the twentieth century, essentially no one in the United States, something like 10% had a high school degree.

And the United States was actually cutting edge on this front. They were much further ahead in terms of universal education than other advanced economies. But even then, only 10% of people had a high school degree, and these were either people who were very rich or extremely academically talented. And that changed very quickly. So by the 1940s or even the 1960s, the share had just like shot up.

It had gone up to like 70%, 80%, 90%. So there was a very rapid improvement in education in the United States, the universal access to this education. And I think that had an enormous effect on the economy, both in the United States and elsewhere. So what it led to is over the, especially the first half, the first seventy, eighty years of the twentieth century, we simultaneously saw massive productivity growth and also saw reducing inequality because this expansion of educational opportunities allowed many more people to pursue higher forms of work that were better paid, etcetera. And I think that we could see a similar transformation with AI technology.

So in the past, the way that we increased our capabilities is we spent more time in school, and that's starting to have actually some negative impacts. So a good example of this is Ben Jones from Northwestern has a few papers where he's documenting this increasing burden of knowledge. So if you look at the first age at which major inventors have their first breakthrough innovation, that's increasing over time. So earlier in the twentieth century, it was something like by age 32 on average, you reach your first major invention. But today that's something closer to like age 40.

And the reason that's the case is because I need a lot more knowledge and I need to learn a lot more. I need to spend a lot more time in school to get at the frontier of a field than was the case in the past, because we're kind of building on the shoulders of giants in our profession. But if there is a way that instead of requiring people to spend more time in school to reach their frontier, we actually improved the learning technology. So the rate at which people can get to the frontier, so they don't have to spend twenty five years in school like I have, but instead we can compress that timeline by personalized learning, different innovations that could be enabled by AI. That could be, I think, pretty transformative.

And that would, one, both be augmentative, so it would increase human capabilities. And also number two, improve productivity growth because the humans themselves would be able to make much more use of their time and labor productivity, etcetera. So I think it could require a lot of investments in improving the infrastructure and the technology around education. But in my view, that's a great direction for us to go in to make the technology more augmentative. And I also think it's actually pretty robust.

So in pretty much any scenario, AI outcomes, it's essentially always good to make people more capable by improving learning. So whether it's just maybe in the future, we won't have any work. Right? And even if we improve this learning technology so that we're all super smart and we can all learn anything in, like, three months or something. Even in that world, our leisure will be better if we're much more knowledgeable and smarter.

Or in the world where there are these severe bottlenecks or adoption frictions that prevent AI from being as transformative as it is, if we do improve learning and we improve human capabilities a lot, that is actually kind of setting a baseline on the productivity improvements that we could see. So I do think that this is kind of a robust solution. It's not solving all the problems, but I think it's going a long way towards making this technology something that is better for humans.

**Danny:** So I suppose your claim is not that that would prevent the technology from overshooting as it were and veering into automation territory. Your claim is that it would be optimal in the sense that under all conceivable possible futures, we would still prefer to have that learning accelerated versus not.

**Bharat:** Yes. Yes. I think that's right.

**Danny:** Speaking of learning, one thing I think you've also thought a fair bit about is critical thinking, which seems increasingly important, possibly the scarce input and the binding constraint when it comes to working with AI. How should we strengthen critical thinking skills?

**Bharat:** Right. I do think that this kind of goes hand in hand with what I was talking about with learning as well. So part of the objective of improving learning capabilities is maybe making it so that people have more incentive or making it easier for people to develop those skills in such a way that's maybe not as demanding as the way we currently try to both cultivate and also evaluate those skills. So it could be that AI could improve that process as well. But I do think one point I want to mention is that we should think about kind of these equilibrium or incentive consequences of AI in developing those critical thinking skills.

So imagine that we're in a world where because of AI, knowledge work is significantly displaced or like not as important as it was in the past. Right? So if that's the case, my returns to really investing in my critical thinking skills when I'm in school will be lower because monetary payoff is just not as high as it used to be. Like over the past fifty years or so, the returns to a college education grew. So the college wage gap is much larger now than it was in the seventies. And that's because of the direction of technology.

But if that reverses and this is a different technology, that's not skill biased, that's not increasing the college wage gap, that's not making it more lucrative to go to school for longer, then that may not be the case going forward. And we could see less investment by people at universities or at the K-twelve level than we've seen right now. So it's essentially kind of your basic incentive trade-off, where if we see these compressions in inequality in the labor market, that could interestingly enough have perverse incentives in terms of my investment in education when I'm in school. Now that may not necessarily be the case. I do think that's kind of an open question whether AI will increase or decrease the demand for critical thinking skills.

And it gets back to what I was talking about before in terms of specifying what it is that I want to the model and how persistent of a bottleneck that will be. If increasingly this becomes a world where we're managing these teams of agents to act on our behalf, then that's increasing the importance of correctly specifying what it is that I want. And that's not a trivial problem. That's like the task of a manager or an executive who is running a company or is managing a team of people. And that requires a lot of critical thinking presumably.

And so I think it's not obvious whether in the long run, like this is going to increase or decrease critical thinking skills, but I do think that we need to kind of sort through these incentives and what the implications might be for education. And maybe that also encourages us to think about how to develop the technology for learning in a different way.

**Danny:** Speaking of management, you've worked closely with Nicholas Bloom. How do you think LLMs will impact management quality in firms?

**Bharat:** Management quality. So I do think that there's been some discussion about whether it might lead to a flatter structure of firms. And I think the rationale for that is each person essentially becomes a manager where they're managing a team of agents to act on their behalf. And so maybe we don't need as much of a hierarchical structure within the firm where I'm directing employees under me to execute, right, because they themselves are just directing the agents to execute on their behalf. And so that is a potential implication of that.

Like, it could be that management skills become both increasingly more important, but also number two, increasingly more widespread and demanded by companies because everyone is essentially gonna be a manager where they're offloading this execution to AI entities. So it could be that that becomes like a more important task even for entry level workers who find their first job. And in turn, that could lead to flatter work structures. So that's kind of one way that I'm thinking about it.

**Danny:** Now if we think about the overall potential effect on the labor market, the standard response to some of these sort of unemployment scenarios is this is the lump of labor fallacy. We've been there before. There isn't—we've talked about this. There isn't a finite amount of work. This is a partial equilibrium in the dynamic equilibrium.

People will find new things to do and the conversation is kinda silly. Let's not even go there. Now it's possible that that will hold true, but it seems to me, you and many others also take quite seriously the possibility that, in fact, this time, it may be different. What would make it so that it will be different this time?

**Bharat:** Right. I think there are two dimensions to this, one in the shorter run and one in the longer run. Over the longer run, I think the way that I'm thinking about it is how quickly can the model capabilities improve so that they end up automating just an enormous share of tasks. And so there's just not a lot left that humans can do at a greater capacity than AI can. And again, I mean, there are a variety of bottlenecks that people have posited that could prevent something like that from happening, including what I was talking about before in terms of specifying what your preferences are.

But that I think is one scenario that could make it very different. It's dependent on how the capabilities evolve over time, how quickly we see improvements in physical tasks and robotics and things like that. So that's one thing that I would want to think about is in the past when new technologies displaced labor, there were new forms of work that were created as a result of that. And humans, because of our malleability, were able to kind of move instead into those forms of work with growing labor demand and shift away from jobs that were decreasing in demand. But if those new tasks that are created are also being done by AI, that could lead to a different outcome.

So over the long run, I think that's kind of the question for me is, are these bottlenecks gonna prevent something like that from happening widespread across the economy? Or are we going to see that these new capabilities are also being done by AI, and that's gonna make it more difficult for many people to find work? In the shorter run, I think the big question that I'm trying to think through right now, and I think this is more immediate, is the areas where we're going to potentially see job displacement, are the people in those occupations going to be able to adjust by finding alternative forms of employment? So that's what we've seen in the past. People have faced displacement.

Today, the unemployment rate is under 5% despite all the enormous technological change we've seen over the past a hundred years or so, because as certain sectors that are being automated face decreasing labor demand, there are other new forms of work that are being created that people can move into instead. Now, the question is, are the people who are being displaced going to be capable of moving into those new forms of work that are being created, the new labor demand that's being created? Or are their skills too far afield from those new pockets of potential demand that they won't be able to make that transition well. So I think a really important question here, and this is a project that I'm working on right now, is kind of thinking through these equilibrium implications, these implications of allowing people to adjust to the scenarios that they're facing. So are there pockets of the economy where there are going to be occupations that face displacement, where the people in those occupations are going to have a hard time finding work?

And do we want to target interventions towards those specific groups of workers? And at the same time, are there pockets of the economy where occupations that are potentially facing displacement are actually not going to have that much trouble finding new forms of work? So I think number one, we want to kind of test in the historical record when you've seen displacement like this in the past. The people who were displaced, what did they do instead? And how easy was it for them to find alternative work?

And then do some forecasting going forward and some scenario planning for what are the occupations that we think might face more displacement and how easy will it be for them to find alternative work in areas that are gonna face growing labor demand.

**Danny:** How should we think about demographic change and migration as part of this? And this may be more relevant for certain European economies than other places, but on some level, if you wanted, you could think of some of these technologies as a substitute to migration and also as a substitute to the workers that you're losing because of demographic change with all the attendant political implications, of course.

**Bharat:** Right. I think this is a fascinating point, and I think it leads to some of the differences in response to this technology that we're seeing between different countries. So I do think that this is quite salient for places like East Asia, where they're seeing this kind of demographic crunch with much lower birth rates. And I don't think that labor displacement is necessarily top of mind for them in relation to this issue—they're just facing a labor supply crunch where they don't have enough workers, and that's going to be a growing problem going forward. And so they would love it if there could be AI that could essentially mimic a huge influx of workers entering the economy.

Now that said, I do think that this could be quite problematic, certainly in the short to medium run. And I think the reason that's the case is because if you look at places like the United States, a disproportionate share of innovation is being done by immigrants. And I think if we make it more difficult for immigrants to enter the country, that could also potentially slow down or that could slow down the rate of progress in developing these technologies or the development of those technologies to happen in other places that are potentially more receptive to immigrants. So I do want to be cognizant of that—I mean, even now in these AI companies, a lot of development is happening via the work of immigrants as kind of the place where they all congregate from around the world, the brightest minds who are working on these problems. And so I definitely think that that's kind of a first order issue in the short to medium run.

**Danny:** What will you work on next?

**Bharat:** So there are kind of three big things on my docket right now. One is measuring the international impacts of AI. And so, that's work I'm doing with Belki Klein-Tussink at King's College in London. So we're trying to expand this canary style analysis to countries around the world, and we're using this data from Revelio, which has employment information for different countries. And we want to first clarify in what places we're seeing these impacts on entry level workers and which places that we're not seeing that, how the trends kind of vary by country, both developing countries versus more developed countries, Western Europe versus other parts of the world, where there might be more of these adoption frictions versus less of these adoption frictions.

And then we also want to take seriously this idea about measuring adoption at the firm. So using these job postings to measure that and then get a better sense of, is this being driven by firms that are actually adopting AI, is there something else going on potentially? Maybe anticipation effects or other things that are influencing employment outcomes that are not being driven by AI. So we wanna extend that analysis internationally. The other thing that I'm working on, like I was just mentioning, is modeling these equilibrium implications and doing the scenario planning around how will workers adjust to potential displacement, and identifying the pockets where we might want to target interventions towards certain occupations that are going to have a hard time adjusting.

And I was actually just talking to my boss, Erik Brynjolfsson, as well about thinking through some of these Baumol cost disease type ideas. So certain sectors of the economy are going to represent a growing share of importance going forward, whether that's health care. So how do we incorporate those types of ideas as well in terms of where we're going to see a greater need for employment going forward? So that's one idea around this kind of the modeling and calibration and the simulation of these labor impacts and where we're going to see these impacts and have difficulty adjusting. And then number three is thinking more about these education ideas.

So, I mean, even just to start, like how is AI impacting education? There was a nice piece in The Economist about this recently that was kind of surveying the research in this space. But I do think we need to know a lot more about how AI is impacting school, how it's impacting curricula, how it's impacting students, as well as in terms of what careers they want to pursue. I think we're still in the very early days of understanding some of these things. And so I think we both need to just collect a lot more data about it.

And then I think we want to think more seriously about, can we design tools that build on top of these models and improve the rate of learning? I think if we could do that, that could be a huge benefit to society as a whole.

**Danny:** Bharat, thank you very much.

**Bharat:** Yeah. Thank you. This is a great discussion. Thank you for the questions.

**Danny:** Thanks for listening to High Variance. You can subscribe to this podcast on Apple Podcasts, Spotify or wherever you get your podcasts. If you like this podcast, please give us a rating and leave a review. This makes a big difference particularly for newer podcasts like this one.

---

## Improv Wisdom – with Patricia Ryan Madson

2025-12-10 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/improv-wisdom-with-patricia-ryan-madson/id1839231559?i=1000740540830) | [Spotify](https://open.spotify.com/episode/2Ap4Y4NWcYbmO2hH3lKdqY?si=ae550d18906e4ef1)

Patricia Ryan Madson, professor emerita at Stanford and author of 'Improv Wisdom', joins Danny Buerkli to talk about how she got into improv, how she starts a class, how status works, Keith Johnstone's dark side, and the four A's of improv: attention, acceptance, appreciation, and action.

**Danny Buerkli:** Patricia Ryan Madson is professor emerita from Stanford where she began teaching in 1977. She's been teaching drama since '62, in fact, and is the author of the really wonderful book, Improv Wisdom. Patricia, welcome.

**Patricia Ryan Madson:** Thank you, Danny. It's an honor to be here, really.

**Danny:** Patricia, what do you teach when you teach improv?

**Patricia:** Well, I think what I'm attempting to do is to cultivate some human qualities. The first thing I say when I start a class is the most important thing you need to know about improv is it's not about you. Of course, it's about me. I'm here too. So what I'm trying to teach is shifting your attention from our normal self-centered point of view onto your partner, what's going on in the world, to being more involved.

It's a shift of perception, I guess, and then experience. And I think I'm hoping to try to teach some social skills of how to work with others and how to trust. I call it trust reality, trust the way it is. You don't have to like the way it is, but it's a sense of confidence, not in the self but in things as they are, reality or the way it's working. Does that make sense?

**Danny:** Yes. Absolutely. And we'll go into sort of some of that in just a minute. But before we do that, you've been a teacher for a very long time. And I think in college, as you were taking drama classes, the person running the program at the time came to you and said, I think you've got what it takes to be a really good teacher.

Would you consider being a teacher? Which is extraordinary. So two questions. What did he see, and what makes you such a good teacher?

**Patricia:** Wow. Oh, what a grand question. I feel that teaching is a way that we get to impart whatever life wisdom we've discovered in ways to help others uncover important things about life. I, for many years, was interested in studying philosophy. I love to think about how we think about things and what we value.

And so I think I learned to teach by teaching and then discovered it was the greatest job in the world because I had summers off. The calendar was wonderful and that there was a chance that I could keep learning about life and others through the process of teaching. I guess I always felt lucky to have Raymond Hodges who kind of threw me in, saying, "I think you'd make a good teacher." And then starting at Saint Catherine's School in Richmond, Virginia, it became natural and I moved from a preparatory, a private preparatory school into teaching at various universities and got really lucky to get a job at Stanford.

And I think I got the job at Stanford because while I had credentials to teach acting, I also knew something about how to teach voice and speech, all of that projection, the techniques. As I taught through the lens of teaching acting, I discovered, especially at Stanford, that students were good at following a script, but they were not so good when I would say, okay, what do you think or feel? How would you react in that situation? They weren't connecting with their own feelings and thoughts and humanity.

So I needed some kind of method to help unlock all of that, the creativity and all of the imagination that's in all of us. Even though the Stanford students wanted the right answer, I wanted their human response. So into my life popped Keith Johnstone, the wonderful British guru who has become a famous spokesperson through his book, Impro. And Keith came into my life in 1980—I think I figured it out—right at 1980 because he had just published Impro in 1979 and I met him at Esalen Institute where he was teaching alongside my tai chi teacher at the time. I was all interested in Eastern thought and religions.

And Johnstone's work then kind of came into my life and I began using his games and exercises in my drama classes. And it was wonderful. It helped the students to begin to flourish. And because one of the things in improv is there aren't any wrong answers. And so that's a new point of view for certainly the kind of student that Stanford would attract.

So what does that mean? There are no wrong answers. It means that we learn to set aside the natural capacity we have for judging and evaluating something in advance to opening to it and seeing what happens if I go with this, if I follow whatever is going on rather than evaluating it. Stanford students hate mistakes. And most of us probably do.

But a favorite quote of mine is a mistake is an event followed by an evaluation. Cats don't make mistakes. And so when we get into talking about how improvisation is helpful as a life skill, mistakes take on a new meaning. It's another event that we can work with in some way. Sometimes, I mean, there are many evidences in the world of art where an artist made a mistake and it turned into their greatest work.

**Danny:** When you teach, how do you start a class?

**Patricia:** Well, I have to improvise too. So when I come into the class, I see who's there and what's going on. I look at the temperature of the room, what the students seem to be doing, and I usually have a game or an exercise in my pocket. If I need to improvise, I need to get the body involved. So I get everybody up standing in a circle and we might walk around and I might say, raise your hand and clap your hand to another person and say, I'm glad you're here because I'm interested in them noticing who else is there rather than doing something that shows themselves. So a lot of the games are about becoming aware of what's happening in the room, doing something that creates a new interaction, and preferably one that's positive.

That's why I'm glad you're here makes me look you in the eye, clap you on the hands, and notice. It also kind of relieves a little bit of tension. Okay, we're in an improv class and what are we going to do? There are a lot of opening games, throwing the ball back and forth, an imaginary ball, where the emphasis is not on being creative in some interesting word but the emphasis is on how to receive. Notice what is coming at you, how to receive it and then acknowledge and then find someone else and send the ball.

So improv is an interactive subject that lets us playfully work with other people in ways that are different than most classes. I'm not going to give a lecture with a slideshow and a PowerPoint on the six ways to improvise. We're going to stand up and make some noise, laugh a little bit, and then my whole class evolves out of how that works, how we're doing, and maybe there are some people that are still a little bit jet-lagged from the day, or we need to pair off and play some games together. Would you like to tell a story right now, word at a time with me?

**Danny:** Sure.

**Patricia:** Okay. So, we're just gonna see what kind of short story comes up. I'll start. Whenever he goes

**Danny:** to

**Patricia:** the

**Danny:** shop.

**Patricia:** He

**Danny:** buys a bar of soap.

**Patricia:** Ta da. Whenever he goes to the store, buys a bar of soap. Terrific.

**Danny:** What an excellent story.

**Patricia:** But it is a story. For most people, that's a surprise. You've studied improv at Stanford some years ago. You're aware that, without any particular plan, we can create together. More important than some kind of brilliance is the dynamic that goes on between two people or more when we're doing something together without an agreed-upon plan. I think one of the gifts that Johnstone gave us—I think he's one of the first writers I know that talks about the value of yes—and developing a positive response to whatever is going on.

**Danny:** There's a wonderful warm-up exercise where you walk around the room, you point at things, and you shout the wrong word.

**Patricia:** Shout the wrong name.

**Danny:** Yeah. You shout the wrong name. You point at a table and you shout chair. And two things. It's surprisingly difficult.

And somehow, the room looks different afterwards. Yes. Why is that?

**Patricia:** I think that disassociation drill shifts something in our perception, in our motor activity, and the room always does feel and look different. It's a great exercise if you're, say, in your office and you're working on something and you're a little bit stuck and you don't quite know what to do right now and you might think I'll go and have a cup of tea, or I suggest you get up and you play that game. And if you can push yourself to go as fast as possible as you trip: mother-in-law, bricks, dynamite, chairs, silver, tree, mock, my gosh, four, fireplace. In fact, our listeners might want to pause this podcast and walk—it works best if you're also moving around as

**Danny:** Yes.

**Patricia:** As you put things. It plays with your mind.

**Danny:** I highly recommend it.

**Patricia:** I recommend it.

**Danny:** Why is it that when we're watching improv, which is really satisfying, somehow it's very clear that watching simpler scenes is more satisfying than incredibly elaborate, complex, outlandish things?

**Patricia:** I think it's a misunderstanding about creativity or innovation. So the notion that innovation sort of by definition is something that nobody's ever thought of. It's way outside the box—that's the common parlance for that. And it's a misunderstanding because it's actually easy to get something—fried mermaids is my example of something that nobody's thinking about. You put together two weird things and that's creative.

Johnstone's view and what I want to prescribe is it's more valuable and interesting, we'll see, to look sort of inside the box or what is right in your visual mental field, look with sort of fresh eyes. Open yourself to something ordinary. A scene about a table as the opening prompt is likely to be more interesting than fried mermaids because that's already a joke and it's finished. So I do believe that we're all incredibly creative if we let go of the notion that the product of that creativity needs to be something that is way out in left field.

**Danny:** What is a good scene?

**Patricia:** A good scene is where two humans are reacting and working with each other to liberate the imagination of their partner. So being more interested in how I can take whatever offer you give to me and make it work, make it come alive. I think human interaction is what's interesting. And Johnstone says that the true test is whether in the course of the scene, one or the other of the characters is changed. They don't just hold on to their position.

So often, beginning improvisers will start a scene just with some kind of an argument, some way that I don't like this or you need to stop putting trash in my yard or something that is combative. And much more likely to produce an interesting scene would be an offer that is kind, or I noticed you got a new haircut. That's looking good on you. Or did you do that for a special event? By engaging your partner and often in a positive way, you're much more likely to get a scene that's interesting. These days, with so much on television—thousands and thousands of movies to watch—I've been noticing that so much of a film these days is just about conflict and violence.

And I'm so much more interested in watching stories that are about humans who, like me, make mistakes and are having interesting adventures or trials and whatnot. I think improv teaches you to look for the story that can come out of something other than just aggression or conflict.

**Danny:** And how do you find the end in a scene?

**Patricia:** When someone is changed and recognizes that in some way. We don't—it's hard because we don't like to be changed. We like to kind of hold our position and feel strong about whatever we think or whatever. But to shift—we're seeing so much now in the way that our governments are not working together because that capacity to interact with each other and change one another isn't functioning like it was meant to in some democracies or at other times. I would like to get all of our legislators in a room and do some improv games where they have to say yes and accept the ideas of their partner and then work with them.

Because almost any idea can work if we decide to work with it.

**Danny:** You mentioned Keith Johnstone, and I have his book right next to me here. This is the book that got me into improv. So he was clearly an important influence on you. Another important influence in improv in general, but I think also on you, was Viola Spolin.

**Patricia:** Yes.

**Danny:** Can you say more about her and how her way of thinking about improv is different from Keith Johnstone's?

**Patricia:** I met Viola Spolin many years ago when I was first teaching in my early twenties. I went to a conference in Chicago, a theater association conference, and took her workshop. And Viola Spolin had a kind of fresh, positive mind. A lot of her work is about perception, about noticing and receiving what's happening. I remember one of her games where we were to sit and imagine that we were watching a ping-pong game.

So we would take in—we would act like a person watching a ping-pong game, and it would make our eyes go back and forth. So these games trained you to see and embody different states of things. She was just a great teacher of perception. Everything begins with attention. In fact, what we're paying attention to becomes our life, if you think about it.

Because if I'm sitting in a business meeting and a PowerPoint is going on, speakers giving me interesting remarks, but what I'm paying attention to is a series of problems that I have with my cat. And we're often off in our heads with stories and fears and plans and whatnot. But if I notice what I'm noticing, then I can shift back—perhaps what I need to be noticing is the speaker's content or there's something interesting and important about why I'm here. I think the study of improv and Viola's work and her first book on improvisation in the theater—I think it's called Improvisation for the Theater.

**Danny:** For the theater. Yes.

**Patricia:** Yes.

**Danny:** Okay.

**Patricia:** She gave us games for paying attention to life and helping us to embody it. Her work didn't include the work on status that Johnstone is famous for. Johnstone is a fascinating character and someone with a little bit more of a dark side to his persona and to his teaching. Whereas when I think of Viola Spolin, I think of light and bright colors and positive energy. And Keith sometimes is really interested in—

Yeah. The dark side's the only way I can talk about it. It's not—some of his work with masks allows characters and content to come forward that you wouldn't find otherwise.

**Danny:** This is so interesting that you bring it up because I was gonna ask you about the mask chapter in his book. The book is very accessible to a lay, non-theater audience. And then comes the last chapter, which is the chapter about masks, and it is very different indeed—dark and weird.

**Patricia:** Yeah. Dark and weird. In fact, when I teach an improv class and I have his book as one of the required texts, I actually say, you don't need to read the chapter on masks. That is optional. Some people find it problematic in some way.

So of course, when I say you don't need to read it, students will immediately go to that chapter. Isn't that weird? But with masks, his theory—it's kind of not supernatural, a little woo-woo. His belief, and I've experienced this, is that a mask can sort of liberate by its use, liberate a different kind of spirit. So we get into another level of acting with his theories of mask work.

And I've seen him teach masks in various classes, and it happens. When you put on a mask, it gives your body-mind permission to go somewhere. This happens all the time at Halloween where we put on a costume, and if you put on the costume of a pirate with a sword at your side, somehow all of a sudden the body begins to take on some of the spirit of those roles, and that's helpful in acting. All acting is kind of mask work.

**Danny:** Can you say more about his work on status?

**Patricia:** Well, Johnstone points out that everything we do is filled with status and is making some kind of a statement. There are four status possibilities—say what they are and then try to explain. I can be raising my status, which is actually what I'm doing right now by lecturing on status. I'm the person who knows something about status. So I'm telling you about Johnstone's status.

I can be raising my status with my behavior or speech or body. Or I can be lowering my status and we sometimes do that if we're self-deprecating or we say I often start a conversation on a podcast and you know I'm really—I'm nervous about this. I'm hoping that it'll be useful to you. When I begin talking about myself in an honest way, I'm lowering my status. You can perhaps even feel that in my voice.

Or I can be pointing out that, Danny's work with government, you have been doing some really important work. Your recent paper on AI, I think, has some brilliant insights and helps us possibly in the future deal with what's coming our way with that new technology. You're a great writer, Danny. So you can see I'm raising his status. Or I could say, I read that paper and there's another author who's also looking at that work that has a different point of view. So I can kind of put him down a little bit by my comments, etc. And so Johnstone points out that there is no such thing as a neutral interaction.

We're always in some way raising our status, lowering it, raising the others, or lowering it. And so if we understand that this dynamic is always in play, it can help us understand certain situations. A good example is we maybe have a coworker or a boss who always seems to be only raising their own status or lowering others. And so when I see that that's their nature, if you will, then it's not so personal when he puts me down or lowers my status. And it may be that to get along with that coworker, what he needs is to have his status raised.

I'm gonna look for every opportunity to make sure that I notice his value and compliment him on what he's doing, listen, etcetera. Listening to someone else is a way of raising their status. Anyways, these status games, once you see them as games, it's fun to play. And then we begin to see how these dynamics are at work in our everyday life. And we can use that knowledge for better relationships.

I had a coworker once who was one of those who had to have high status, and I used to end up kind of confrontational. I had to try to raise my status and go back and forth. When I realized what was going on, I started shifting that and raised his status and listened better. And our whole relationship just flowered. We have control of our own reactions or how they're perceived.

Also, a lot of the status work shows how our physical behavior, whether we are standing or sitting tall with our shoulders back, which tends to give the effect of a higher status, or whether our shoulders have softened and whether our head is bowed or we use our hands and wiggle them a lot. So we can send off signals of dominance or submission. Keith points out that dominance and submission is what this is about, but he doesn't like to use those words because they're loaded in a sense. Certainly, just Jane Goodall came to mind because her work with the animal kingdom certainly shows that status is alive and well in the primates' world and we're part of that.

**Danny:** Certainly not exclusive to us. Thinking of people who have influenced you or have been your teachers, as it were, what have you learned from Peter Brook?

**Patricia:** Oh, thank you for bringing that up. Brook was one of my heroes. I think he had a brilliant imagination as a director and someone who understood the whole meaning of live theater and how it can be an art form of a very high quality. He knew actors and he knew the stage and he had a vision and a kindness that I remember. I did a thesis on him and spent some time with his company in Paris watching.

He's brilliant. At the time, I wasn't involved in improv, but I was very much involved in traditional theater. And he was my superhero of someone who was working on a very almost spiritual level, I would say. Brook was one of the masters. I think he passed just recently in the last years.

Someone who has influenced me, another teacher that I mentioned at the end of Improv Wisdom, I say I'm standing on the shoulders of two giants, Keith Johnstone, a theater person, and then the other is psychologist David K. Reynolds, whose work I recommend. His work on something called constructive living, a set of philosophical ideas based on Japanese psychotherapies. Something called Morita therapy, developed by Shoma Morita, and Naikan. Naikan is a Japanese word that means looking inside. Naikan is a form of meditation created by a Japanese gentleman named Yoshimoto Isshin.

And Naikan is about noticing, really noticing what we're receiving from others, from the world, at any point in time. Right now, I'm receiving, starting with your attention and perhaps the attention of some listeners who've kindly taken the time to tune in thus far. There's a technology that is making this possible. Actually, when we think about it, we probably had a meal today and that food came from somewhere, was grown by farmers and brought to a market and then brought into our houses and consumed and cared for. Naikan is about shifting the attention from yourself onto how this self is able to live and function in the world thanks to the specific work and efforts of others.

And it's a practice that I also had the privilege of experiencing. I went to Japan and sat in the monastery for over a week and looked at the details of my life in relationship to what I've received, starting with my mother. And Naikan is about three questions: what have I received? What have I given back or done for them? And the third question is what trouble and bother have I caused them?

Which is not a question we normally ask because we know a lot about the upside of that. The trouble that people cause us very often fills our mind with what's wrong with so-and-so.

**Danny:** Easy to answer.

**Patricia:** Yep. Yep. I know a lot about that. But we almost never examine who am I troubling through the way I'm doing something or whatnot. And it's helpful and humbling to really ask and try to answer that question.

A lot of constructive living is about again shifting the attention off of yourself onto the abundant world that is providing and giving to us even if it isn't always what I choose or want. It's a big difference. For example, if I'm in a restaurant and having a meal and the waitress that's coming back and forth I notice is sort of surly and has kind of an attitude and maybe she didn't refill my coffee as fast as I'd like. It's quite possible that what I focus on is, notice, she's not very nice. But what I miss is the other side of that is that this lady who's maybe having a bad attitude has brought the food that allows me now to eat and sustain my life.

A lot of times the things that come to us, we judge in some way, but we miss the fundamental truth that our lives depend on the work and the kindness and the efforts of others. And so Naikan is about enumerating those things, keeping a list every day, noticing what it is that is sustaining me now. It's an antidote to the critical mind, which we all have. One of the maxims in the book is wake up to the gifts. And it's about that topic.

I'm often asked what's my favorite of my maxims or is there any one of those that I find that in the world that we live in now, that maxim of waking up to the gifts and noticing is more important than ever. There's a great book by Rob Walker called The Art of Noticing. I highly recommend the book. It's a series of exercises about teaching your mind to pay attention to different things in different ways. So if we work with—I think improv is a great subject to practice working with your own attention.

What are you noticing? What are you interested in? How can you contribute? Improv Wisdom has 13 maxims but I'm finding as I get older a lot of people don't have time for 13. So I've condensed the 13 maxims into four maybe more easily memorable topics and they're all with As, the four As.

Then the first—we're talking about—is attention. We have to notice what is going on or where we are or what's happening. In fact, often at this point, I would say close your eyes. Let's do that. Let's close our eyes and listen to me for just a moment.

And so my question is, what's around you? Can you create in your mind the vision, the picture of the world that you're in right now? What are you sitting on or standing near or what's around you? See if you can get that picture. Okay, after a few moments then I'll say, okay, now let's open our eyes, look around, and what hadn't you noticed before?

I'm often surprised—well, I thought I was pretty much aware of my surroundings. I sit in this area a lot of the time but there's a card here that I hadn't realized. Anyway, attention is the first thing that we have a control over and that we can shift it around and notice what we're noticing and then place that attention where we want it to be if it's off in some kind of dream state. So attention, then acceptance. And that's where improv—acceptance doesn't mean liking, but it means opening to and letting whatever that reality is that you're working with, letting it in without pushing it away, or blocking.

In improv, blocking is the cardinal sin. We don't want to push it away. We want to say what's going on here.

**Danny:** This is the yes and bit that many will be familiar with.

**Patricia:** Yes. Yeah. The yes and. It's adding. Then the third A is—

I'm sorry. Attention, acceptance, appreciation. That's where the waking up to the gifts come in. Once I open to something, what have we got here? What can I appreciate about this?

Maybe you're in the dentist chair and he finds, oh, you're gonna possibly have to have a root canal. That's not going to be fun. But you might then appreciate, aren't I lucky I've got the dentist that has found this and he has the credentials and all of the work that it's going to take to make this better. And aren't I lucky to have insurance that's going to cover this? Or what can I appreciate even about a negative situation?

Okay. Attention, acceptance, appreciation, then action. That's where what we do matters. So if we live in an improv world where we open to things, we're noticing, we're gonna adopt a yes-and kind of mind. What's my role here in making things better or adding to or helping to raise the status of the work of my partner.

So there you got it. Small nutshell.

**Danny:** These are great. And they rhyme with—Rob Poynton, another improv teacher, has a saying. I think he says, improv can be boiled down to three things: let go, notice more, use everything, which rhymes very nicely with the four As.

**Patricia:** Great. Yes. He's wonderful. And those three things of letting go, which is important to get to the acceptance part, and notice more and use everything. I love use everything, especially now.

**Danny:** In '82, I believe, you spent a year traveling around Nepal, Indonesia, India, Thailand, Japan, possibly other places to study performance pedagogy. How did that change your pedagogy?

**Patricia:** Wow. My goodness. That delights me. Yes. That trip around the world right at my fortieth—yes.

Right at my fortieth birthday. What I learned was for me an answer to Einstein's question, is the universe friendly? I found that the universe was friendly, which meant that wherever I went, if I had some problem or need, there was always someone who showed up to help me or to point me in the right direction or to lend me a sweater or something. Because the trip was fascinating in today's standards because it was without a phone. No one had phones in nineteen eighty, eighty-one, eighty-two.

And my way of getting around in the world would be I had a ticket that took me around the world and I could stop in different places. And I could keep going in the line of direction. So I couldn't backtrack, but I could go up and down and forward and I could stay as long as I wanted. So I had different experiences and adventures along the way and I always found that help was there if I opened to it. And then I always looked for ways in which I could in some way give something or be useful.

Sometimes it was maybe a foot rub or to tell a story. Part of the trip I was in India and I went to a drama school there. And so I taught some acting exercises and some voice things in exchange for their kindness to let me visit the school. And I spent time in a Buddhist monastery in Nepal and then crazy adventures in Thailand on the beaches and things. So I moved through the world with the help of other people.

And it's this theme of shifting—about it being all about me. And I think today, if I tried to make that trip, me and my phone would be figuring out where to go next or what Yelp said about this bed and breakfast, or I would call a Waymo or something. That trip around the world was wonderful because I was still able to interact with people in the world, move from place to place, and recognize that humanity—we're all in some ways wanting the same things. We want friendship and kindness and a place to stay and human needs are very similar around the world, even for those with whom we disagree. Wish we could get that remembrance more clear today for all of us.

**Danny:** There's an obvious convergence of interests between a lot of people who do improv and people who are interested in—I'm not sure what the right term is—mysticism, Eastern religion, different ways of putting it. I think you've talked about improv as being the Tao, the way, rather than—can you say more about that convergence and at least why these two schools of thought and practices tend to be compatible?

**Patricia:** You're right that there's, certainly for me, some kind of alignment with a lot of the Eastern religions—Buddhism, Taoism, Hinduism—which are all worthy of study and interest. They often feature the notion of acceptance and openness. There's some sort of philosophical convergences with the principles in improv that make us go forward. And it's kind of fun to look at those, just the four that I just talked about: attention. Almost all of Eastern thought, I think, has some form of meditation or of working with the mind to develop—today the term is mindfulness, which is very much kind of in fashion. Bringing your attention again into the present moment, being with whatever is going on there, accepting it.

So some of those notions of acceptance and being positive seem to align. It was interesting that you kind of picked up on that.

**Danny:** You're in California. You mentioned you met Keith Johnstone at Esalen, which is on the coast, of course. How has the fact that you've spent the last decades in California, of all places, sort of shaped how you do improv?

**Patricia:** First comes to mind is this notion of the Eastern religions. When I moved to California, really the religions that I'd studied were Western religions, Christianity and Judaism. And getting to California, there was this flux of different Eastern religions available. There were monasteries and temples and a whole bunch of different kinds of Buddhism that were available, and so I was just like a hungry philosopher trying out different classes and courses—part of the Zen Center in San Francisco and another Mahayana Buddhist organization over in Berkeley. And so I think being in California, we're right on the coast and we're closer geographically to the Eastern countries where many of these religions and philosophies grew up.

So I was happy to have a place to study and meet other like-minded people.

**Danny:** Something that came up actually earlier but that just popped back into my head is when we sort of go back to the classroom and how you're teaching, there's a thing called side coaching.

**Patricia:** Mhmm.

**Danny:** Can you talk about that and why is it so effective?

**Patricia:** Fascinating. Because Johnstone is a master of side coaching. I've been in classes with him and then there'll be two students on stage doing something and he'll throw an idea to them which changes the direction of what they're saying or doing. He'll side-coach them to go somewhere different. It's very effective as a teaching method.

I have to admit I am not good at that and I don't side-coach because I'm not looking for a particular result. I'm more likely to let the scene kinda keep going wherever it's going or not. And then afterwards, we'll say, what was working? And after the fact, we'll look at what we did or didn't do and perhaps point out now, if you'd accepted that offer, it might have taken a new direction. So I'm not myself very good at side coaching.

Johnstone was brilliant. And I often would not have thought of the coaching note that he gives. He has insight about the structure of scenes in theater that he really was a master at. And to this day, I don't perform improv. I'm not working with any of the groups as an onstage player.

I do enjoy going to Bay Area Theater Sports or some of the groups around town, but it's much too scary for me to go on stage. Oh, boy. That's hard stuff. I can't do that. All I'm useful for is pointing out that the things that make improv work are also useful things in life and we could all be more of a yes-sayer and a better listener and be more grateful.

**Danny:** Well, and an excellent teacher of those who do go on stage, I would add.

**Patricia:** Thank you. Yeah.

**Danny:** Final question. What should I have asked that I didn't ask?

**Patricia:** Sorta, what's up now with me? Alright?

**Danny:** What is up now with you?

**Patricia:** Well, I just got a new haircut yesterday. I've got short hair for the first time in my life. And I'm gonna be teaching one last time for Stanford for the continuing studies program for the winter quarter. They have a winter quarter starting in January and I've been teaching a class for now almost thirty years called Everyday Spontaneity: Improvising Your Life, which is using the improv games that we've been talking about, and it's open to anyone. I'm going back into the classroom one last time, and I will turn 83 years old on December 3rd. So, wow.

I have to compliment you—as a host, you did more homework in preparing for this than anyone has ever done. I'm flattered and delighted by the questions you asked. It was fun to be able to revisit some things about my past. So thank you, Danny.

**Danny:** Thank you, Patricia. This was such a pleasure. Thanks so much.

**Patricia:** Have a great day.

---

## Scenario Planning – with Jamais Cascio

2025-11-26 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/scenario-planning-with-jamais-cascio/id1839231559?i=1000738468842) | [Spotify](https://open.spotify.com/episode/6SBlmNgxDhz3hlgcwgI8qT?si=a6b71f13e5994eed)

Jamais Cascio, futurist, scenario expert, and author of Navigating the Age of Chaos, joins Danny Buerkli for a deep dive into scenario planning. They discuss how the discipline has evolved since the days of Herman Kahn at RAND and Pierre Wack at Shell, whether the military or the private sector do it better, why geoengineering might lead to predictable trouble (and why we might do it anyway), and whether today's AI is more or less weird than Jamais once imagined. Jamais also reflects on his time working with Ken Waltz and shares the story behind his BANI framework, which captures how many now perceive the world: brittle, anxious, nonlinear, and incomprehensible.


**Danny Buerkli**: My guest today is Jamais Cascio. Jamais is a futurist and scenarios expert. He also has a new book out called "Navigating the Age of Chaos". We'll get to that in just a minute. 

The core problem is that the future is uncertain, and yet we must act – as you also write in your book. And that's why scenario planning holds so much promise. Jamais, welcome.

**Jamais Cascio**: Thank you very much. Glad to be here.

**Danny**: Let's clarify some terms first before we get going. What's the difference between scenario planning, forecasting, anticipation, and strategic foresight, and how do these things fit together?

**Jamais**: Well, these are not mutually exclusive terms. Scenario planning is a type of foresight practice which can be used for strategic planning, and they're they all fall under the umbrella of anticipation. That is, they're all about thinking about what might be the consequences of present day decisions. 

Most people, when they think about the future, do so in a either in a linear way, you know, think about the future, and this is what will happen, or they think about it in a binary way. 

Like, here's a good outcome. Here's a bad outcome. With scenario planning or scenario forecasting, scenario thinking, what that asks us to do is to think about multiple plausible futures that are, if at all possible, broadly equal in positive and negative. So you aren't steering towards "this is my delightful outcome" and "this is my tragic outcome" so much as "here are three, four, five different ways things could turn out". 

Now, the the stumbling block for a lot of people is that you're going to be wrong. Your scenarios are going to miss the mark on a lot of in a lot of ways.

The goal then is to come up with scenarios that are illustrative of different possible futures such that they give you a a new sensitivity to change such that that when something is happening that is that parallels one of your scenarios, you recognize it early. The goal of a scenario is not to tell you the future, but to give you the tools to see it coming.

**Danny**: What's the intellectual heritage and the lineage of scenario planning? Because one of the interesting features of scenario planning is it seems so obvious today that it's hard to imagine there was a world where we did not think in scenarios, and yet that was true at some point.

**Jamais**: Well, certainly to a degree. Certainly, people have always or have for centuries, if not millennia, thought about different possible outcomes. What came about in the nineteen sixties with a gentleman by the name of Herman Kahn, who wrote a book called "On Thermonuclear War", among several others, was essentially formalizing a process. So, Herman Kahn, in in the early sixties, was a researcher at the RAND Corporation in Santa Monica, California, and his job was to think about what he called the unthinkable, to think about the consequences of an actual nuclear war. And in doing so, he developed a process to think about different ways the future could turn out, but to do so in a way that very formalized, so that all of the scenarios made sure they covered the same types of issue areas, that all of them were on roughly the same timeline.

And then a gentleman named, Pierre Wack, who worked for Royal Dutch Shell, took that same line of thinking and turned it into a business process. So, the late sixties, into the early seventies, Royal Dutch Shell started doing a regular, I think, every five year scenario process where they would work out different stories of what the future could hold for Shell around everything from, you know, oil production to climate to demographic global demographics. And tell, for them, they they've formalized it into four stories, and they used a something that people nowadays call the four box method. Essentially, the more technical term is the strategic divergences or, diversion catalyst futures, where you look at two large axes of uncertainty. Okay?

So two big issues that when you cross them, you get four big four large scenario types. So, for example, if you're talking about the future of of computing, you might have one axis be what the the economy might look like over the next ten years, the economy outside of information technology, what it might look like, and the other axis being how fast information technology hardware gets better. That those are not the best ones you can do just off top of my head. And so you come up with a scenario in which there is a strong economy and rapidly improving IT, one that's a strong economy with slowly improving IT, and then a weak economy, strong IT, weak economy, weak IT. And those four scenarios give you pathways to tell a story, pathways to figure out what would the future look like, how would we get from where we are now to that future?

Or as the technique evolved, they started looking future back is the term that gets used in the industry. You start out by looking at the what your imagined future is like, and you start thinking backwards from there. Like, okay. What caused that and what caused that and leading back to where you are now? The goal with all of these is to come up with different operating environments.

That is to stories about what the future may hold that you that will be the environment in which you will be making decisions.

**Danny**: How has the practice evolved since the time of Herman Kahn and Pierre Wack?

**Jamais**: Well, let's see. In the late seventies, an academic named, Jim Dator at the University of Hawaii – they actually have a futures studies department in their political science department. Actually, you can get a a graduate degree in future studies from Hawaii. It's kinda great. But Jim Dator came up with a – he'd been doing a lot of work both for governments and for businesses and for academia, but he realized that there were four different archetypes of what scenarios could look like, that no matter what you're talking about, good I strong IT, weak economy, whatever your your specific characteristics, the underlying them were four broad archetypes.

The the first is what he called the growth archetype. That is, it's a scenario in which whatever you're looking at is increasing. Now most people interpret that to mean a good scenario. Like, things get better, things get we get more of the economy growth or whatever. Dator noted that that that doesn't always mean a good scenario.

You could be looking at scenarios of, flu propagation, infection rates. And if that's growing, that's not a a good scenario. But so the the focus is the growth. It's a scenario that's built around whatever you're looking at getting bigger or accumulating in in number. The in contrast to that, the second is what we call the collapse scenario.

Again, here's one that most people interpret to mean that things are horrible. And in in most cases, that's that's the case. But it could also mean a collapse of opposition, a collapse of well, a collapse of infection rates. It's it's whatever you again, whatever you're looking at, whatever is the focus of the scenario falling apart. 

The third is what he called discipline, usually gets called constraint these days, and it's a scenario in which there are limitations placed on the focal point. And they could be regulatory. They could be environmental. Because something happens in over the course of the scenario that puts, hard limits on what kind of kinds of outcomes you can get. 

And then the fourth is transformation, and that's a scenario where essentially all bets are off. Things happen that you don't expect that may not even seem plausible.

And underlying all of this, Dator had an has an aphorism, because he's still around, has an aphorism that is any useful scenario of the future must initially sound ridiculous. That is if you if the scenarios you come up with, whether they are growth constraint, collapse, transformation, whatever they are, if the first reaction isn't an, oh, no way. If the first reaction is in some way being shocked, then it's probably not going far enough. If the scenario sounds too plausible, it's probably far too cautious. 

And so around and in the mid eighties, the whole bunch of folks from from Royal Dutch Shell spun off into their own company, a scenario planning company called Global Business Network, in which they basically had as their business practice, both providing scenario services but also training others to be scenario forecasters.

And I actually worked with them from '95 through '98, and they had this real philosophy of this should be part of every consultant's toolkit. Every business analyst, every strategic planner, they should all have have scenario thinking, scenario planning in their set of tools, and they did it. It's actually kind of remarkable to look back at the eighties and nineties, the early parts of the February, and see how little how rarely scenarios were brought up as a strategic planning method. But over the course of the late nineties into into the present day, scenario thinking has become extremely commonplace. And it's it's really odd to see a revolution succeed.

I mean, because they they really did try to do something what they felt was revolutionary. It really changed the way that business people and government people and military people thought about the future, and they did it. And so at this point, scenarios, whether they are the the Royal Dutch Shell style four box axes of uncertainty or the Jim Ditter scenario archetypes or just some other kind of generative process, a kind of a way of just sort of pulling the different scenarios out the possibilities out of what you're thinking in your head. All of these are very not just commonplace, but very useful to people, whether we're talking business or government or military or community organizations.

**Danny**: How did you get into scenarios work in the first place?

**Jamais**: I'm an easily distracted generalist. That is, I'm interested in everything and, oh, shiny. It's something I don't know whether I have ADHD or never been actually, officially diagnosed with it, but it is my education comes from a study of history with a focus on revolutionary movements, anthropology, which is a stud, with a focus on, human evolution, political science with a focus on international politics, then doing a lot of work with computers, primarily in the network administration sense, and doing a lot of writing in comic book type stuff. I mean, so I have this really both a varied footprint of interests and a desire to to learn something new. And my wife got a job at Global Business Network doing some some administrative work and actually set up their first web server.

And so I got started to go to GBN events and met these people and thought, oh my god. This is exactly what I want to do. So I figured out a way to get myself a job there, and that was in '95. And it's just been that my life's been going gangbusters as a scenario thinker since.

My life has been going well as a scenario thinker since. It is an interestingly applicable skill because I from that basic training and then the elaborations upon it that I did, I worked in the I worked in Hollywood for a few years, working with, screenwriters to help them build out their science fiction stories. I did I actually created a couple of role playing game books for Steve Jackson Games, a pretty well known game company based in Texas.

They were doing a series they called Transhuman Space, which was a very realistic look in what the year 2100 could look like. And so I wrote a couple of the settings books for that, one about what the developing world looks like, one about what popular culture and religion looks like in 2100 in a world of AIs and genetically engineered animals and the like. 

So and then you can start to see that the this practice has given me just an incredible playground to imagine different possible futures. And that's so that's what I've been doing, and I work with I I've given talks and done done work with everything from major, Fortune 10 companies. 

One of the biggest automakers in Europe actually use some of my scenarios for their strategic planning to governments, various whether you're talking ministry of education in The UK, the Department of Defense, or the Ministry of Defense in Singapore, you know, of all.

I've done work with them on their the the ways they think about their future, and I've just a lot of writing and ultimately accumulating coming finishing up in this book, Navigating the Age of Chaos.

**Danny**: As an anthropologist, you're familiar with the the idea of of manifestation. And it seems there's an interesting concern occasionally around scenario writing, scenario planning that and I think one of the criticisms leveled at Herman Kahn was, that by describing, by thinking the unthinkable, he might bring about the undesirable future of nuclear holocaust. 

How do you think about the this worry of manifesting undesirable futures just by describing them? Or maybe put differently, have you ever refused to develop a scenario because you're worried that making it so vivid might actually somehow make it real?

**Jamais**: Generally speaking, I'm not concerned in the esoteric sense of being able to manifest a reality. And, frankly, if I if I were able to do that, it'd be coming up with a much better reality than what we have. I actually am a strong believer that you can't think about what you would do if you don't have an idea of what's coming. That is, if you are in a if you're in a crisis or if you think a crisis may be on on its way, thinking through the different possible consequences is I believe that is a really important it's a very important task. It is something that allows you to recognize the surprising elements.

So when you we all have an embedded view of the future. It's just the nature of how we have to work in reality is we have to think about the future. But we do so in a very haphazard way. Like I said at the outset, it's it tends to be very either linear or binary, and it is we tend to expect that the world will be more or less the way it is, except some things might change. But we already have a sense of what those things that that would change might be.

So phones will get will get better or whatever. Computers will get will get smarter, or the environment will continue to get worse. And we usually unconsciously build out our expectations for what we might be doing in six months or a year or ten years based on those internalized expectations of normality. 

But always, there are surprising and unexpected developments along the way. Sometimes they're small. Sometimes they're massive. Sometimes they're, like, they're pandemic level massive. It's important not to imagine exactly what will happen, but think about the possibility. 

What if something really disruptive happens? What is something that makes it impossible for us as a company to have all of our workers in one place at the same time, which is something that many companies in California already think about in terms of earthquakes.

So if you're a company and because highway roads have been damaged, but there's still but there's electricity and Internet and you wanna keep working, how do you keep going? Will you build out plans for remote work, or you build out plans for remote for remote access to your clientele, your customers, or your patients. And a lot of companies in California have done that. 

And it turns out the companies with the with the best remote work, remote whether remote employee or remote client structures in place, plans in place, are the ones that you know, around earthquakes were the ones that were best suited to shift around COVID. Because earthquakes and pandemic are very different in many ways, but the same underlying it's a crisis that has made it so that we can't get our people in here, you know, all in one place for an extended period, what can we do?

Those plans actually still could still apply. And so that's my response to the manifestation idea is that they didn't manifest the they didn't manifest the pandemic. They were thinking about earthquakes. But the planning that they had done, the thinking about possible outcomes, possible consequences, and building up the resilience around how do we adapt if some this something really disastrous happens. Mhmm.

That was very important that allowed many of these organizations to survive and even thrive.

**Danny**: There's sort of a a positive version of this, which I believe comes from Adam Kahane, who was also at Global Business Network, possibly with you. And I think he calls it transformative scenario planning, which sort of deliberately using thinking about the future to then shape reality or give he works in conflict facilitation. Right? Give conflict parties some sense of how a different future might be possible.

**Jamais**: Well, right. Yeah. That that's actually a really useful thing. I have worked with Adam before, and we've talked a lot about the work he did in South Africa and thinking about the what happens post apartheid. And he I think he raises a really important point that in many ways, we can't build out a future we can't see.

That is, if we have convinced ourselves or have let ourselves believe that no good outcome is possible, then when opportunities arise to create a good outcome, we may miss them. We may not be able to see that as a possible path forward. 

Whereas, if you think in terms of scenarios and there are scenarios where you succeed at this, in dealing with this big problem. Now there may be side effects and other consequences to deal with, but you you generally succeed at this problem. You now see that as a as a possible path forward.

And even though, as we said before, the real future isn't going to be identical to that, You will recognize where there are opportunities to make positive changes in ways that you didn't that you might not have seen before simply because you didn't believe it was possible. There is that old cliche of it. If I didn't see it, I wouldn't believe it. Well, with scenarios in the future thinking in many ways, it's if I didn't believe it, I wouldn't see it.

**Danny**: A critique that comes up often of the scenario method is that sort of it stops that insight, but then it doesn't translate that into strategy or, you know, what what should we do now. Merely improving insight doesn't just give us better action. Right? What what's needed to make that happen to bridge that gap?

**Jamais**: So that's actually part of I would do a lot of work with a group called called Institute for the Future, which actually spun out from the Rand Corporation fifty three years ago. And, so they're one of the few futures foresight companies, foresight organizations that have outlived their own scenarios. They have a pretty good track record, but they have as their guiding principle this trinity of Foresight, Insight, Action that by developing foresight, it allows us to to then build insights into what our problems are, what steps we could take. Action allows us to to let is our movement, which then generates a need for more foresight. So it becomes a circular process.

But they also recognize that they are not experts in auto making or health care or environmental defense or politics or all these different issues that that, IFTF has to work with. So whatever action steps and what are strategies the IFTF people come up with are not going to be as useful or as insightful as the strategies that are that are come that people within the organization, the client organizations come up with. 

So what they argue from early on is that the decision makers have to be part of the process. The people who ultimately make the strategy, make the decisions, have to be part of the underlying scenario and foresight creation process. Because if it's just brought to them afterwards, they'll say, thank you, and put it on a shelf.

I in doing this for as long as I've done I've done this, I've seen that so many times where the people who are who within the organization who are responsible for the foresight don't have the authority then to turn around and implement the strategy. All they can do is bring the foresight and their insights and suggestions to the decision makers who invariably have their own agendas. And some of them will will make use. 

Like I said, the major European automaker, the folks in charge who were who were a level above the folks that we worked with, did take those those recommendations seriously. But by and large, if you don't have the buy in from the outset and the participation from the outset of the people who have the responsibility to make decisions and to carry out strategy, it's really hard to to make that leap from insight to action.

**Danny**: So it seems that maybe something that occasionally gets misunderstood about this type of work is that what you're describing is is an experiential good. It's not the quality is not, as you say, in the brilliant output, though, course, the output may well be brilliant. It can be, but much of the value is in taking people through the process. And I think Pierre Wack also talks about this. 

Unless you manage to sort of change the mental model in the head of the decision maker, it's very unlikely to actually generate any meaningful.

**Jamais**: Yeah. As you say, the the foresight work, the writing, the videos, whatever the artifacts you come up with, they may be brilliant. But if no one is deriving an insight from them, they are an exercise in creative fiction. And creative fiction can be great. We're all big fans of creative fiction. But the the the business case for creative fiction is fairly limited outside of a few companies.

**Danny**: What's different in how the military and the intelligence world use scenarios versus the civilian world? Because naively, it would seem that that is a domain where this type of method, this type of work is taken way more seriously, probably on average, compared to the private sector.

**Jamais**: It's interesting. My experience with having worked with military and intelligence, and of course, with the the civilian business world, is that the military and intelligence scenario thinkers, foresight thinkers are very broad minded, very, very much willing to look at seemingly disparate, seemingly irrelevant drivers that might actually have a long term consequence. They're willing to look at the big, big picture and see how it all become is interwoven. 

Whereas my experience with business leaders, even the ones that have bought into the idea of doing foresight and scenario thinking, so often focus solely on their industry or solely on the economy without really thinking through, well, how does the how does climate change affect us? How does the potential for conflict with between The US and China? How does that affect us? And so the military and intelligence services tend to be better at doing the work. 

However, they haven't they are definitely in the position of not having the authority to make the decisions. All they can do is bring the analysis and insight to their to the political officials. And very often, those political officials don't wanna hear the hear it.

In 2000, the organization I worked with had someone come in who was a mid level CIA person complaining about how can we get our higher ups to listen to us? They they all have a preset idea of what the future's gonna hold in mind. How can we communicate this? Because we're seeing all sorts of stuff they're just not listening to, not paying attention to, which was frustrating and had a lot of sympathy for them. And, of course, it takes on a very different very different tone when he thinks that it's just about a year before nine eleven. 

Similarly, I did a a big project in 02/2012, I think, with the the CIA Center for Climate Change and National Security. There was actually a CIA Center for Climate Change and National Security. We did a whole role playing game, essentially, of a tensions between The US and China over geoengineering. And we all brought in a bunch of people, people who were who had worked in government, people who were currently in government, people who were and intelligence, people who were media folks, people who were just thinkers and scenario thinkers and the like.

And we all had different roles within this larger I think about a 100 people involved total. And it was fascinating. It was great. We actually came up with some really interesting results. And the next year, congress was taken over by the Republicans and shut them down, stopped funding them because it was climate change, and so, therefore, it was a liberal hoax.

So the great tragedy is that the folks in with government connections who do this kind of work tend to be really good at it, but they have a very limited amount of success because of who they have to report to. And the people they have to report to tend not to be those who like to think in terms of different possible futures. Very often, politicians have a future, and if you push against that future, you are a you're a traitor. 

Infamously, during the Gulf War, in so early February, Condoleezza Rice, who was the national security administrator for the for, George Bush George W. Bush, said something along the lines of, it's counterproductive to think about what happens if you fail when you're trying to succeed.

And my response is, no. That's exactly the time you need to be thinking about what happens if you fail, because you know that you are not certain to succeed. And so you need to be able to think about what do I do if. At least that's my perspective. It wasn't hers.

And she was in a high position in government, and I'm not. So what do I know? But that that's the dilemma. So you have businesses who, when they implement, are usually usually able to make that connection between insight and action possible or better at least better able to than folks in the government, but they're less likely to have good insights. Because in many cases, especially those who are not working with scenario service organizations and consultants, you have you have very limited narrow views of the future.

And it's something that we've that a lot of us in this field talk about when we talk to our the clientele, the audience, is the the cone of possibility. And to think where we are today, we have a a range of different possible outcomes, and the further out we look, the wider the range is. 

Well, most people have a fairly narrow cone of possibility, a cone of what they can imagine as possible, and the job with of the scenario of the foresight people is to get you to widen that cone, which means including some stuff that seems ridiculous. But as Jim Dator says, any useful forecast must initially seem ridiculous.

**Danny**: You mentioned geoengineering. I know you've done quite a bit of work on geoengineering. You've published a book on geoengineering. Should we just put the sulfur back in the shipping fuel?

**Jamais**: No. No. First of all, that's that's not getting getting it high enough in the atmosphere to do anything more than acid rain. So the the thing about the the stratospheric sulfate injection types of geoengineering, you know, these those plans, that those part particles get put very high up into the stratosphere such that they get caught up in the global wind circulation, global atmospheric circulation, and spread around. 

And that we know from we know from history that that that kind of material in the stratosphere can have a cooling effect on worldwide, and that's from volcanoes. Because big volcanoes, like in most classically Mount Pinatubo in '91, pump a lot of crap into the air, and a good chunk of it hits the stratosphere. That makes it possible. And so as a result of Pinatubo, we saw a couple of years of global cooling of about a degree. We, actually cooled down about a degree from where we were prior to Pinatubo, and that lasted for a couple of years. 

The problem with any of that, any kind of thermal regulation geoengineering or solar solar radiation management geoengineering, is that it only attacks one part of the problem.

Temperature is a big problem. Heat is a big problem, but the accumulation of carbon in the atmosphere has its own set of of other issues, including, in particular, the increased carbon levels in the oceans. And ocean carbonization has led to leads to coral bleaching and the death of coral reefs, just as of one small example or one big example, but just one of many. 

And because even as you put stuff into the stratosphere to block out 3% of incoming sunlight or whatever the number is, if you're still pumping c o two and other greenhouse gases into the atmosphere, it's still it's trapping more and more heat. And so you have to put more and more stuff into the stratosphere to to keep up with that.

So you do this for for fifty years, and you build up all enough stuff that you're blocking now five or even 10% of incoming sunlight, and you have your carbon levels on the and greenhouse gas levels on the atmosphere are much higher than they are today, and then something happens. There's a war. There is a some kind of global event that makes the geoengineering system break down and stop being able to put stuff into the atmosphere, and it dissipates. You know, the stuff that we have we have to keep pumping. We have to keep pumping the sulfates or the whatever, you know, whatever per particulates we put into the stratosphere. 

And if we stop doing that, it dissipates fairly quickly, and all of that accumulated greenhouse gas in the atmosphere remains. And all of that heat that is accumulated that's being compensated for by blocking sunlight remains. And what we see is a temperature spike. A temperature spike of, you know, something like 10 degrees is the is the typical estimate for for one of these processes. Something that is outrageously unsustainable, frankly unsurvivable in large parts of the world. 

And so it is something where you have two two pathways ahead, broadly speaking, if you start geoengineering, thermal geoengineering, and one is to keep doing and keep doing and cross your fingers that it never stops, and the other is to use that as a tourniquet. Use it as an emergency procedure to give you the time to pull as much greenhouse gas out of the atmosphere as possible, because you'll need to get actually get down to a level below pre industrial levels, pre industrial greenhouse gas levels in order to compensate for the accumulate the accumulated heat. 

And so it is what thermal geoengineering does is it basically stops it stops the world from getting too hot to act. So in the so my analogy of a tourniquet is very much on point. The goal of a tourniquet is to stop you from bleeding out and dying while we get you to a doctor who can do something more than that.

**Danny**: And there's very much an expiration date on it.

**Jamais**: Exactly.

**Danny**: Measure measured in minutes, maybe a little more, but not much.

**Jamais**: But yeah. Right. Exactly. And with thermal geoengineering, it may be measured by a little bit more than minutes, but it's still a finite process. And you need to be to use that time to work rapidly and comprehensively.

We should be doing that now anyway. The heat, however, is a big issue. And I was actually on a project fairly recently with a with an organization that had a lot of money that they wanted to provide for an NGO. They had a lot of money they wanted to provide for climate related adaptation. And my suggestion was easy, air conditioning, because most a good part of the world is not built around having air conditioned buildings.

And those parts of the world, when they get hit by large and persistent heat waves, see thousands, if not tens of thousands of people die. There are massive waves of of death, heat related death, across Europe nearly every summer. And parts of the world that are built around air conditioning, and there are all sorts of problems with air conditioning. 

There with the environmental issues around energy consumption, around the the cooling the cooling material, but people dying from it is not one of them, at least dying in the short term. The parts of the world that don't have the infrastructure around air conditioning, they don't it doesn't have to be a big Freon or chemical based cooling system. It could be radical changes in architecture, because there's really interesting architectural designs that allow for cooling. 

But something has to be done, and that will save a lot of people fairly quickly and fairly inexpensively. But it's very hard to get a big organization that's accustomed to thinking on this massive scale about something so prosaic. And that's one of the big dilemmas about doing any kind of foresight work, any kind of futures work, is convincing people that you're working with that the changes that they see are both plausible and, in some cases, desirable, or they must be avoidable or they must be avoided. And so often people are stuck in their their predetermined visions of the future that it's very difficult to get them to see a different perspective, which is isn't a problem unique to foresight work.

**Danny**: There's an interesting asymmetry there in what you say when we think about geoengineering because the AC solution requires, at least in Europe, regulatory approval. The reason why in many European countries there is no AC is for regulatory reasons. On the other hand, certain geoengineering interventions have this frankly fascinating property of being permissionless. Technically, a single sufficiently sophisticated actor could just do it.

**Jamais**: Yep. That's one of the things about about thermal geoengineering is that it's something that a moderately sized nation state could start to do. And and one of the handful of centibillionaires that are out there could do it without without question. 

And then it becomes an issue of it's something that a single nation, a single actor can do, but it has a global impact. And the global impact isn't just bringing down temperatures.

It includes things like changing rainfall patterns, changing wind patterns, changing the the calendar for when when plants can do their first leaf after winter. All of these major changes that then have their own cascading follow on effects that quite often are not pleasant. 

So one thing that gets brought up fairly often is one of the earlier studies around the possible consequences of thermo geoengineering was one that showed, or at least gave gave strong evidence, that monsoonal rain patterns would shift in South Asia such that it would eliminate the traditional growing season for South India. And so that's the most populous country on the planet suddenly going without food or going without a large amount of its food. 

Even if the overall climate is cooler, the the lower level impacts can be so dramatic and unexpected that we don't we don't like that I don't have a good policy in place for how to deal with that, and it's almost certainly going to lead to conflict.

I mentioned that CIA center, simulation event that I participated in a decade ago, more. The outcome of that simulation was both the American team and the Chinese team preparing to preparing to engage in nuclear retaliation, expecting the other side to launch first, but both of them being ready to engage in nuclear war over the actions taken during geoengineering to that have adverse effects on each side. Now in many cases, the adverse effects in that simulation were imaginary. That is not based on on a particular scientific study.

But broadly, they were the kinds of things we should expect to see, changes in drought and rain, changes in growing seasons, changing in the types of plants you can grow. It's people often when we talk about geoengineering, it very often gets brought up in the context of it's a quick fix, but it's not a fix, and it's not quick. 

It's something that you end having to do for decades, and the consequences can be terrible. It really it's one of those and I I've talked with a lot of scientists who work on geoengineering, who have done who do that research. And pretty much to a person, they are adamantly opposed to the deployment of geoengineering except in the most desperate situations.

They do not want to see this used because they know how many consequences and bad consequences there could be, but they recognize that there may be a point where not doing it has even worse consequences.

**Danny**: Speaking of weird things, in 2008, in an interview, you were asked about "advanced AI", and your answer was, "we're waiting for a breakthrough, but it's going to be weird"". We arguably have advanced AI. Is what we got more weird or less weird than you anticipated?

**Jamais**: Weird in a different way. I think my assumption was in that in that 2008 interview was that we were talking actually actual self aware sapient AI, singularity path AI. And what we have now, I tend to at least with the language model based systems that are the ones that can easily pass a Turing test and seem intelligent, they're fancy autocomplete. They are not self aware. They're not sapient.

They just do a really good job of pretending. But it is weird because of the impacts that the one, the impacts that the existing systems have had on behavior, especially among younger people. Two, the impacts that they've had on the behavior of corporate executives. And three, the almost violent demand by so many tech companies to embed language model AI into everything. No.

I do not need to have my word processing program give me writing tips based on what their LLM thinks about what what I should say next. I do not need that in my email. They're all we just see this in so many companies embedding semi useful, at best, AI systems into their products and making it the only product path available. And we see what happens when too many people start using it. There there are already multiple studies showing a decline in critical thinking for people who use LLM based AI on a regular basis.

There are the way that these systems have been designed to be essentially be obsequious to tell you what you wanna hear and to tell you that you're great for even asking that question, That has all sorts of terrible psychological and sociological impacts. 

And so, yes, the it's weirder now that we have this kind of AI. It's just not weird in the way I I was expecting. I was expecting something along the lines of much more rapid leaps in capacity to do things, completely inexplicable designs for, energy systems or the things that we really do require super intelligent AI to come up with. 

But, also, at the same time, I think what I was also arguing was that as much as we may think of it we think of it as a weird future, we should avoid thinking of it as a salvation future.

So whether we're we're talking they finally get out language model AI to do something decent or they actually do invent superintelligent AI, I have talked with so many people who think the singularity is a sustainability tactic or strategy. That is who think of, don't worry about this present day problem because once we have the singularity, once we have super AI, they'll fix it.

**Danny**: The the deus ex machina will fix it.

**Jamais**: Exactly. Exactly. And that's why you had Cory Doctorow and Charlie Strauss calling it "the rapture of the nerds" way back when. You know, it it there are some amazing parallels between rapture eschatology and the harder core Marxist revolutionary eschatology, whereas Marx himself wrote, all that is solid melts into air, and once we get to that point where of a true communist, utopia and singularity eschatology. 

It's all essentially coming up with the big system to take to that makes us not need that makes us not need to have responsibility anymore because the system is taking care of us.

And I have to say that that is a really appealing future. The the Richard Brautigan machines of loving grace, if you've ever heard that poem. Yes. All watched over by machines of loving grace.

**Danny**: Incidentally, also the title of Dario Amodei's, one of the cofounders of Anthropic, essay about the future capabilities of AI.

**Jamais**: Right. Or to to be less literary about it, but perhaps more enjoyable, the Ian Banks culture novel science fiction novels where you have these massively intel superintelligent, hyperintelligent AIs in control of everything, allowing humans to just be, to go out and explore and have fun and play games and have a lot of sex and just all the things that would make life wonderful. 

I sometimes when I've talked about this kind of kind of possible future, I've talked about it as being like a permanent burning man where there's no need to work because it's all part of the environment. You you get what you need. It's art and drugs and games and sex.

And for a lot of us, that's a really appealing vision of the future, but that's not where the current companies in charge of designing AI seem to wanna go.

**Danny**: Strange that. You you coined the acronym BANI, brittle anxious nonlinear incomprehensible, as a sort of counterproposal to VUCA. Right? Volatile, uncertain, complex, ambiguous

**Jamais**: It's not a counterproposal. It's an evolution. VUCA was a term developed by the US Army War College in 1989 as a way of giving a broad umbrella description of the state of the world at the end of the Cold War. So you think about late eighty nine, early nineties, you had the collapse of the Berlin Wall, the the collapse of the Soviet Union, the emergence of the Internet in a truly substantive way. And then within a few years, the worldwide the emergence of the worldwide web, you had radical changes in demography.

It just seemed like the world was really volatile and very uncertain and complex and ambiguous, and that was for the military and government planners. That was a really useful term. When nine eleven happened, the use of VUCA among the among officials might have broadened out to being it becoming a term that's really commonplace for business analysts, business consultants, in part because they love that military language, but and in part because it's actually a really it was a really handy acronym. 

Because all of the stuff that we saw happening was very volatile, but ultimately understandable. That is, yes, the consequences of nine eleven in terms of global power and ethics, all of these issues around who who's in charge and what military what kind of military gets used, a military action happens.

That was all very uncertain and complex. But once you had an explanation, once you could see all the pieces, you could figure it out. It was not baffling, but it was VUCA was certainly in the February a really useful way of looking at the world. 

But as things got weirder, as politics started to change and technology started to get strange, and you started to see some really radical shifts in behavior. By the mid twenty tens or so, VUCA became less and less useful, both because it was everywhere.

I sometimes would say in talk, we eat VUCA for breakfast. It's just it's ubiquitous. We swim through it, in part because it was insufficient. And so in a talk I gave for Institute for the Future for their fifty fifty year anniversary in 2018, which makes them '57, not '53. Anyway, for the fifty year anniversary in 2018, I gave a talk about global anarchy and chaos, and I offered up the term bonnie, brittle, anxious, nonlinear, incomprehensible, partially as a a provocation.

And I didn't really expect it to go to get big, and in part because it felt more satisfying to me based on what I was seeing in the world. I had been doing a lot of writing around resilience and brittle systemic brittleness. I have been doing writing doing a lot of research around, anxiety and especially among younger people and just seeing that being a a global phenomenon. Nonlinearity, I I had done a lot of writing around climate change, and one of the things that had stuck with me is is something called climate hysteresis, which is a lag between cause and effect, a lag between changes in levels of greenhouse gas and changes in temperature and precipitation of about five to ten years. So I would start and I immediately started thinking of it not as a geophysical issue, because it's interesting as geophysics, but as a political issue.

Because if you get people around the world to radically change their behavior, to stop eating beef, to stop driving, or at least as much, to stop taking air travel, at least as much, to stop using, trans oceanic shipping that runs on diesel, all of these things that would would lead to radical changes in our lives that a lot of people would see as not for the better, at least in the short term, and then nothing improves as a result. Temperatures continue to rise. 

Things continue to behave just as they did before we made all these massive changes. Thinking that as a political issue, not just as a geophysical issue, was when I how I started to see it as this non that kind of nonlinearity of a disproportion of action and result, a disproportion of power, that's really critical to thinking about how the next few decades are gonna transpire. And then ultimately, incomprehensibility, not in the sense of we don't understand what is happening, but why.

We've seen we have seen a measurable increase in situations that make a lot of people ask, what the fuck? And it's things that in when I wrote the book, Navigating the Age of Chaos, which is a folk look at BANI, I talk about the incomprehensible world as meaning unthinkable, things that we just don't want to wrap our heads around and absurd and and baffling. And I include things like the anti vaxx movement. It's it's absurd, and yet there's something happening that needs to be understood that why is that become such a powerful political force? Unthinkable.

And talking there, actually, I this I wrote in detail about geoengineering there because that really does fall into that unthinkable category. It's something that the people who study it don't want to be thinking about, and the people who aren't studying it but have heard of it just sort of somebody else will think about it. Let's not think about the consequences because it it has a particular utility, particular goal, and let's just focus on that. And so when COVID hit in 2010 or 2020 or 2019, 2020, few folks who had seen my original presentation, which was met with polite applause and not much more than that, except for people from Brazil. They saw and said, that exactly describes our world.

In 2020, when COVID hit, I was it was suggested to me that I make that talk available. So I rewrote it as an essay, put it on on Medium, and didn't think much more about it until I started to get email and started to get pings popping up on Google, you know, Google alerts about my name. And, suddenly, BANI was everywhere or at least was increasingly used, especially by people in the in the global South, in the developing world, because what they told me is that it was a much more accurate articulation of their of their experiences than VUCA ever was. That you had so folks in in Brazil I started doing remote talks in Brazil and Mexico. Rajarata University in Sri Lanka actually held a Bani symposium that summer in the 2020, and I gave a remote keynote for that.

And it at this point, there are now hundreds of thousands of uses, if not millions of uses that you can find on Google, of people using BANI primarily, but not exclusively, in the developing world. It's really popular in India among business and political leaders in India and Indonesia, but I'm starting to see it now pop up in Western Europe. 

The minister the of education for in Spain just a couple of months ago did a talk around the BANI world, for his team. I get a lot of pings from Russia and Ukraine, and I have talked with a lot of people in in The United States and in Europe and Japan, and it's starting to become a more prevalent idea in the in the post industrial world too. It not because it is technically better, but because it is it feels more accurate.

And that's something that really has become an important part of the the Bonnie conversation is that it is not just a depiction of what we think about the world. It's a depiction of how the world makes us feel. And for quite a few people, that's actually a really powerful tool to be able to talk about what they've experienced, to talk about their their lives with a language that is both informative and evocative. That is, it helps to give them a name for what they're experiencing. You know?

Yes. These systems are brittle. I can see I I see how brittleness fits into this into what I'm seeing. But, also, it's not just informative. It is it's evocative in that it helps to legitimize what they're what they're feeling.

That's something that I I've had quite a few people react to me talk to talk to me about is that they felt they were alone, or they felt they were overreacting. Looking out at the the chaos of the world and having a very extreme what the hell is going on, but they must be overreacting to it. Because, surely, this kind of stuff has happened before. I mean, there was the whole nineteen sixties in The US. I mean, there were and the the revolutions in in Central Europe, the the Hungary uprising, the Czech uprising, surely we've gone through this before, but we haven't.

Not to this degree and with this level, this breadth, because it's not just the poll that politics have gotten weird. It's that politics and our technologies, especially around AI, but not exclusively, our climate, our ways of communicating with each other, our abilities to to our the impacts we have on others and the way that everyone in the world can have an impact on us goes far beyond anything we've ever experienced. 

And so BANI became a language the language that people wanted to use so many people wanted to use. I knew I needed to write a book, but in talking with Bob Johansson, who's a former president of Institute for the Future and has been doing this kind of foresight work longer than I have, he strongly suggested they have a positive BANI, something that uses the same acronym, the B-A-N-I, but flips it and says, okay.

They're not fixes for the for the BANI, you know, the negative BANI, the original recipe BANI, forces so much as they are adaptations. What can we do to cope? So instead for a world that's becoming increasingly brittle, we need to think about being bendable. That is building up our resilience, our adapt our adaptive ability, our capacity to both bend, you know, bend and not break when the problem hits, but remain rooted. For a world that's increasingly anxious, we need to be attentive, which is the a version of empathetic.

We need need to increase our ability to have empathy, to listen to what other people are saying, to recognize that they're having they're undergoing a crisis just as much as we are, and to have and to be kind, frankly. My wife has a T shirt, and I may have mentioned this to you before, but my wife has a T shirt that she wears every now and then that says, in in a world where you can be anything, be kind. And that really does resonate for me as being the underlying message of the a in Positive BANI. The n, in a world where of that's increasingly nonlinear, you gotta be neuroflexible or neuroplastic, improvisational. That is you need to be able to recognize when situations are not what you thought they were, and you need to be able to adjust on your feet.

Improvisational training, improv acting training is actually really useful for leaders, it turns out, because it puts them it trains them in how to listen to what the situation is in ways that it you didn't expect it to be, that break away from your assumptions, but still allows you to have a plan have an idea of what where to go. 

It is a embodiment of the the definition of fluid intelligence that I that I read a while back is that it's how you know what to do when you don't know what to do. That is your capacity for recognizing the stat the the state of a situation and how it differs from what you expected and to change your change your actions accordingly. 

And then I, for an incomprehensible world, what you need is interconnectivity, interconnected or inclusive. And, actually, inclusive, I used initially, but there's such a weird pushback in the political and business world around the DEI words that we decided that it's better to have that as the secondary definition, interconnected.

But, basically, what it says is that you broaden your circle of of people that you get information that you exchange information with. Your your circle of people who have different perspective, different backgrounds, different are in different locations, different cultures, different life experiences, because very often the things that make no sense to you may make sense to somebody else because they are seeing it from a completely different angle. 

And so by broadening your your circle, expanding out who you're connected with, you have a better chance of getting that getting access to that new perspective, those differing perspectives, and to be able to provide your own. 

So the the book, navigating the age of chaos, the first first chunk is the original body going into detail about what these four terms mean and how they have manifest in the past and issues that may be coming in the in the over the next few decades that are manifestations of it to come, things that the way that the Western Antarctic ice shelf is very brittle and what that may mean. And I use geoengineering in in the incomprehensible, unthinkable part as an example of the the kinds of actions that people are talking about taking in the face of climate that just, on their surface, are totally crazy.

The second half or the second part is positive BANI. And here, we take a slightly different approach. We go into the the mechanisms of what each of these what resilience and empathy and and improvisation inclusion is what that what they include, but also give multiple concrete examples of people who have, in their lives, experienced something along the lines of brittleness or anxiety. And what they have done and their life experiences. 

So as an example, for the improvisation chapter, I talk about, Stanislav Petrov, who was a Soviet air defense officer in 1983, had a was on it was he was on duty about a week after the Soviet Union had shot down Korean Airlines jet double o seven over Kamchatka Peninsula, basically, shot down an airliner.

And so the global tensions were dramatic. There was the US was engaging in a military exercise in east in in Europe. And on the night in question, the computers lit up saying that there are five missiles incoming from the United States. 

And Petrov, who had written the book for air defense officers about what you do when you get an alert, how to send it up the chain of command and the the language you have to use and all that, felt conflicted because it just something felt wrong about this. And yet he knew that he needed he he should send it up the chain.

And because the Soviet Union at the time had a massive retaliation strategy, nuclear strategy. He knew that sending up the train had a very high likelihood of leading to the Soviet Union shooting all of its missiles back in The United States, leading to the US to shoot all of its missiles back in the Soviet Union, and that's all folks. And so he said, okay. This is just weird. I why would the US only launch five missiles?

That just doesn't seem right. I'm gonna sit on this. And he sat on it for longer than it would take for the missiles to hit, and nothing happened. And then the confirmation came through that it was a satellite error. It basically saw, it interpreted the sunlight reflecting off a layer of clouds as the the infrared signals of launches, and and so he broke he broke his rules.

He did not follow the expected pattern. He improvised because the situation felt different than what he expected. And so that, to me, that was an example, because I love that story, because it really does drive home just how critical that the need for being aware of change can be, aware of aware of how the reality differs from your expectations. 

And on top of this, a a an academic psychologist, Judith Moskowitz, I think she's at University of Chicago, had come up with a, with methods for dealing with a chaotic world, psychological practices. And some of them are very familiar, mindfulness and gratitude journals, the kinds of little things that seem massively out of, out of proportion with talking about collapsing ice shelves and climate engineering, but they all come down to how do we cope?

How do we, as individuals and as leaders, develop the skills that will allow us develop the capacity to think about how we respond to these these large issues. Because it's so easy, and so many of experience of us have experienced this of feeling over so overwhelmed that we just shut down. And so the positive BANI is really about not shutting down. Yeah. Being able to to listen, learn, and adapt.

And then after that, I do a deep chapter on scenarios and storytelling as a tool for thinking about a world. Bob was a primary author on a chapter on leadership, how to be an effective leader in a BANI environment. And so that book just came out on the October 28, and it is very exciting to have it out there in the world. I I don't know what the reaction is gonna be. So far, it's been reasonably positive, but I think it will be useful for people, not as a an instruction manual so much as a a way to shift their paradigm, to shift the way they they they look at the world.

We talk lot about how BANI is a lens. Negative BANI or original BANI and positive BANI, both. They're a lens to see the landscape ahead. And it does doesn't tell you where to go, but gives you greater clarity for figuring out your path.

**Danny**: And it's a great book. Everyone should read it. 

There's one thing analytically sort of chewing on and left with, which is if you listen or read, for instance, Pierre Wack's work, but also others, scenario planning, which often can come across as sort of creative endeavor, which it is, actually rests on really solid analytical foundations. Right? There's we try to figure out what are the predetermined variables that are already baked in that we can see today, things that are baked in certain elements of the future.

And Wack also said, I think most of the errors in futures don't come from wrong reasoning, but from poorly observed facts. So from not looking at the present or the past properly. N

ow if this world that you're describing, this sort of this age of chaos is true, what good are analytical methods? If there is no sort of cause and effect relationship to be discerned, what is the point?

**Jamais**: The cause and effect relationships are are still there even if they're hard to spot, even if they are have become so complex, so nonlinear that they're they're very difficult, functionally impossible to discern. 

But it goes back to something that you and I were talking about just a little while ago, and that is it allows you to see paths that might otherwise not be apparent. So I I often refer to the work that I do, the foresight work, as anticipatory history. 

Because, actually, one of my like I said, one of my undergrad degrees is in in a is in history. And the the same processes used by historiographers today and historians today to look at the past, to figure out, you know, why did the the world take the path that it did?

What led to these developments? Those are the exact same kinds of analysis that we use in scenario and foresight work to look at what might come in the years to come, what might happen in the years to come, what are the driving forces that are shaping that have shaped the world, and how might they shape the world going forward, and how might they change? 

We know that with our force I work, with the scenarios, we are not going to be accurate. We're not going to give you a pinpoint prediction of exactly what will happen. What we are able to do, though, is to give you insights into what might happen so that you become sensitive to those changes, become aware of that possibility, and you start to look for them.

The analogy and there's a lot of analogy and metaphor in foresight work, but the analogy that I often use is scenarios in foresight. They're a vaccination. If you think about the way a vaccination works, it sensitizes your body to a pathogen that you may not ever encounter. But if you do, your body recognizes it, recognizes it early and is is ready for it. 

So, similarly, what the scenarios what foresight does is that it what foresight scenarios do, what foresight does is to give you a sensitivity to the kinds of changes that you might see so that if they do happen, if they do or something very similar to them happens, you're ready.

You've thought about it. And so to the degree that the folks like Pierre Wack really tried to focus on getting it right. I think that was probably appropriate for the era, but that's not where we are now. 

I don't think that getting it right does anything more than make you feel very frustrated as a futurist because nobody listens anyway. You I actually have had the experience of getting it right on some things that people who heard me talking about what I called the participatory panopticon.

And I think a big problem part of the problem was that it was such a mouthful. But, basically, what happens in a world where everyone is carrying around networked always on recording devices, cameras. And I wrote about this in 2005 2003 to 2005.

**Danny**: That's pretty good. Pretty early.

**Jamais**: Three years before the iPhone came out, where the the only camera phones could take really tiny three twenty by four eighty pictures. Okay. But thinking about the consequences of in terms of what it does to relationships, what it does to politics, what it does to policing. And writing about this and giving talks about that, the people who heard it were very positive, but it didn't really make a difference other than I can I can look back and say, I told you so? That's it wouldn't be on my headstone and say, I told you?

I told you this would happen. So what you know, the value that I see is not so much around the act the value that I see to foresight work, to scenarios, is not so much around the accuracy as around the provocation. Mhmm. Does it get the listener? Does it get the reader?

Does it get the thinker or the the strategic thinker to see something different, to have that epiphany, that moment of, oh, wow. I never thought of it that way. I can totally see that happening. And suddenly, their mind starts to calculate or figure out all the different implications that might occur because of that change that they never saw before. And so scenarios are epiphany engines.

They're basically ways to get you to see the world differently such that when the real future does happen, you are better situated to be able to spot it and act on it.

**Danny**: Second to last question. You worked with, Kenneth Waltz when you were at Berkeley. What did you learn from him?

**Jamais**: To think about the world in a systemic way. So Ken Waltz was a massively important political science theoretician. He he died about a decade ago, a little over a decade ago. He wrote a wrote a book called "Man, the State, and War" in 1959.

He was freshly back from the Korean War. He wrote a book in '79 called "Theory of International Politics"". Both of these books are still in use at the college level for, one zero one level courses on on politics because they so clearly articulate a world where we think not in terms of the decisions made by individual actors or the transient power levels of nation states, but in terms of systemic balances and concerns around safety and security. He is a he is one of the few people who think who thought that nuclear proliferation was a good thing. Now I don't 100%

**Danny**: – He wrote the article "More May Be Better"". Right? –

**Jamais**: "More May Be Better"". Yes. So he he wrote a piece that said, look look at the evidence of what has happened in the as a result of the acquisition of nuclear weapons, countries that have fought multiple wars with each other have stopped fighting those wars. They may still be rattling sabers at each other. They may still be very angry at each other, but they haven't actually fought with each other at anything other than maybe some border skirmishes between India and Pakistan or India and China.

They're not engaging in that kind of war because they're terrified of the possibility of being obliterated or, more importantly, of obliterating, functionally obliterating the world. His way of thinking, looking at how systems function, balances not of of individual power, but broader systems of security really does really has influenced my way of thinking about what the future could hold. Really, thinking about the great person, the great man theory of history, not to think about what one person would an Elon or a Zuckerberg or whatever might do, But, really, what does it mean to have that kind of of financial and political power? What does it mean to have that kind of technological influence at that broader level?

**Danny**: Final question. What should I have asked that I didn't?

**Jamais**: I don't actually know what I would just say there. The snarky answer would be where you can purchase the book.

**Danny**: Where can you purchase the book?

**Jamais**: At any of your standard, US based online booksellers or physical bookstores for those of you who have them. You can order it internationally. I've discovered from the from the preorder listings that at least a third of our preorders have come from international buyers. So, this is it's certainly available to to be purchased from overseas, literally and figuratively. 

So I really would hope that anyone who does pick up the book and and gives it a chance, I hope they really think about what it means and what they can how they can use use it.

That one of the things that we've included in the book at the very end is a set of, discussion questions. Really, hearing questions that you might ask in a classroom or in a business meeting about how to take the BANI insights, BANI and Positive BANI insights, and operationalize them or turn them into tools for understanding the world. 

And so I would highly I would encourage people to get it not just because I wrote it, but because I think action is a useful idea. And I am extraordinarily grateful and humbled by the fact that so many people around the world, people from very different cultures and backgrounds to my own, so many people have found it an insightful and powerful tool, and I hope that, everyone feels that way.

**Danny**: Let me repeat. The book is called navigating the age of chaos. It's a great read. Jamais, thank you so much.

**Jamais**: Thank you very much, Danny.

---

## Building an Open LLM – with Antoine Bosselut

2025-11-05 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/building-an-open-llm-with-antoine-bosselut/id1839231559?i=1000735342970) | [Spotify](https://open.spotify.com/episode/0DQ85a9eRRRvbqks1zGMQH?si=9c8411ab3b08484f)

Antoine Bosselut, Assistant Professor at the Swiss Federal Institute of Technology (EPFL) in Lausanne, joins Danny Buerkli to explain how he and his team built Apertus, the 'open' LLM. Antoine and Danny discuss why taxpayers should fund this work, which constraints bite hardest when creating an LLM outside one of the large labs, and which public investments may be needed now.


**Danny Buerkli**: My guest today is Antoine Bosselut. Antoine's an assistant professor at EPFL in Lausanne, Switzerland. He's also one of the creators of Apertus, an open LLM released in September. Antoine, welcome.

**Antoine Bosselut**: Hey. Thanks for having me. It's great to be here.

**Danny**: Thanks so much for doing this. To start, explain briefly what Apertus is and what distinguishes it from, first, other LLMs and then other, quote-unquote, open LLMs.

**Antoine**: Yeah. So, I mean, as you mentioned, Apertus is an LLM — or a large language model. So, you know, essentially a very large model based on neural networks and deep learning, which has been trained, or pretrained as we say, on an insane amount of data — in our case, close to 15 trillion tokens. So it's essentially a large language model whose scale rivals some of its more, I would say, well-known counterparts such as the Qwen, the Qwen-2 models.

What makes it different, I guess, in this context compared to what we often think about with large language models — which is the ChatGPTs and the Claudes and the Geminis of the world — is that Apertus is more like Llama-3 and Qwen-2, in that it's an open model where the weights are actually released online, which allows others to pull those weights from some online repository like Hugging Face or Azure or AWS and essentially run it locally, train on top of it to potentially provide it new capabilities. And it's essentially just a much more flexible interface to a model than, you know, just the chat interfaces that are often used with the frontier organization ones. This naturally has drawbacks, though, in that these frontier organizations have an entire ecosystem built on top of the language model. There, the language model is more like an engine inside a very large car that is cruising forward. In our case, you know, we just have the large language model, so it’s only the engine in that more classic sense.

And then in terms of what makes Apertus stand out on its own relative to models such as Llama and Qwen is that it's fully open. We wanted to be able to provide a scientific artifact in addition to a base model that could be used for downstream applications. And so to do that, we wanted to be able to release all of the data that we trained the model on — all 15 trillion tokens. We wanted to be able to release intermediate checkpoints. And so at Apertus’ scale, it's actually the largest model where all of these additional artifacts — such as pretraining data, checkpoints, evaluation suites — are all available to go along with the model weights.

And there are very few other models that have that same level of transparency in their releases, and none of them are at the scale that Apertus is at.

**Danny** Buerkli: Brilliant. And one obvious question is: why bother? Apart from the fact that this is a cool and presumably helpful research artifact because we may want to understand LLMs better — and in order to do that we have to have a model we can interrogate. Arguably, the more open and compliant and all the rest of it it is, the better. And also maybe for the reasons that we may not want to rely on the largesse of large multinational companies to provide these models.

We may want to, you know, be able to, for research purposes, have one of these ourselves. That’s great. And that may already be justification enough. But apart from that, why bother? And why bother with public money?

**Antoine**: There's multiple questions to break down there. In terms of why bother, I'd say there are lots of good reasons for this. Two, I guess, that I'll lean on here are: first, having access to this type of resource from a research perspective just enables us to really expand the number of studies that can be performed on these models. There's essentially not all that much research you can do with the frontier models — at least for us as outsiders — beyond talking to it and seeing what it answers and trying to get something back from that. With open-weight models, you can do a lot more because you have access to those weights.

You can adapt them. You can provide different stimuli and see how the internals of the model — the mechanisms inside the model — change. You can try to discover circuits that are responsible for particular behaviors. So there's a whole host of things that you can do, but you're always limited by the fact that you don't know what this model was trained on. What does it mean for a model to get this much performance on a benchmark if that benchmark may have been part of the training data?

There’s just this gap between good science — science that actually allows us to extract insights — and the foundation of the experiments that we're running, which for open-weight models often lacks that specification. With a fully open model, we can actually do true science on these systems because we're able to audit the entire training procedure of the model up until that point and say, “Okay, there's a flaw in my experiment because I’m testing for something that — even though I didn’t know it at the time — the data had in it when the model was pretrained.” So I would say that in itself is a great value of such a system. And we're not the only ones pushing models forward along that principle. AI2 had the OLMo models. EleutherAI has released a suite of models in previous years. So there are a few organizations operating under these fully open principles.

The other reason is an effect of wanting it to be fully compliant and also useful for businesses. We really wanted to release under this Apache 2.0 license, which allows commercial use. And we also wanted to release all of the data that we had trained on. And that creates some sticky legal conundrums — in that if other people are going to use this model to make money, and you’ve pretrained on a whole bunch of things that make money for other people, that second group might want to come after you legally.

So we made a lot of decisions in choosing the data to train on that essentially make us a lot more compliant with regulations in Europe — including the EU AI Act and GDPR — in terms of what sorts of data can make it into the mix. And this is quite interesting because when we were designing this project, there were organizations that came up to us when we asked them, “What’s the issue with LLMs for you today?” It didn’t have anything to do with performance or capabilities. They're confident they can make something work there.

But they were afraid of the legal exposure that such a model could create for their company if there was any sort of mistake or safety issue. And so for them, knowing that there was this much stronger data-compliance standard was quite important, and it’s one of the things that is very attractive to many of the companies we’ve spoken to.

Now you asked: why use public money to do this? Well, the answer to that is quite simple: it's super important to do this — particularly to create more responsible foundations for models and to have them be useful for sovereign innovation ecosystems as well. But there's not actually all that much money in “responsible AI” at the moment to be made. And so it's not necessarily an attractive bet for a private company to release a model with all these open artifacts — only to have it potentially be used a bit less because it’s not as capable due to the performance gap.

But to me, this is the responsibility of the public sector: to make the investment in the foundational technologies that can then enable entire innovation ecosystems on top of it. It's too much money to do this for just a startup — especially one in Europe. But as far as the public sector goes, it’s a very important thing to do in order to enable the next stage of building the application layer on such a model.

**Danny**: When speaking about money, can we put an order of magnitude on this? I imagine the biggest cost block would be pretraining and post-training runs plus salaries, but the GPU time is presumably the most expensive thing. How much did the whole project cost?

**Antoine**: I mean, there's two things there: how much does it cost us, and how much would it have cost other people? When it comes to cluster economics, there are different ways of defining these costs. So I think all-in — if we count up the GPU hours used — it comes out to somewhere around 10 million GPU hours to do all the experimentation, the final pretraining run, the post-training, all of that.

That’s the majority of the cost. Then, of course, there's the salaries of all the folks that worked on it along the way, but we can probably say that all of those salaries put together come out to around, let's say, 3 million as a number. So then you're left with the compute cost. Now you've got these 10 million hours. How do you decide what an hour is worth?

If you go on Google Cloud or AWS, you're going to be able to buy a GPU-hour. For spot pricing, you'll get it for less, but you can't reliably train on spot pricing, so you'd need the full value of that compute. And you're not going to use it for a year, so you won't get the committed-use discounts. So you're going to be paying something like $5–6 per GPU-hour. That would be $50–60 million to train on that cloud.

If you go for a bare-bones service, you can often get something like $2–2.50 per GPU-hour. So ~$25 million. All of these things are essentially out of reach for a private company hitting the scene.

However, we do have a supercomputer in Switzerland that has around 11,000 GH200 GPUs. The base cost on that — for energy, cooling, and salaries of running it — is a lot less. And we can probably value the compute for this project at closer to 5–7 million francs.

And so: what would have taken a private company $50 million to run is completely out of scope. But we can rely on public investments that have already been made in infrastructure, and that allow us to get a much better price point. So not only is it the responsibility of the public sector, but it also makes more sense economically.

**Danny**: You mentioned the immediate utility to research, the immediate potential utility to firms who worry about legal liability and other issues. I wonder how much there is also the optionality value — because building these systems presumably involves a lot of experiential expertise that you can only gain by actually doing it. There’s presumably a limit to how much expertise you can amass by only thinking about how you would build an LLM if you were to do it. There’s expertise that comes from actually doing it. And that translates into future optionality value. How would you weigh present-day value versus the value this generates into the future if we were to continue on this path?

**Antoine**: Yeah. That is an incredible question. There is so much to unpack. I'll start with a little anecdote. When we kicked off this project, I had a few people tell me, “This is not what academics do. Why are you trying to do this? You're not going to publish your typical research papers.”

And to be honest, I didn’t particularly care. It seemed like a really exciting thing to do and something we needed to do in order to stay competitive in the AI race in Switzerland.

But something incredible happened along the way: we found research problems that aren’t really well codified or aren’t necessarily talked about. And one thing that came out of this work is that there were something like 20–25 research papers that did get written. So we did achieve the academic mission in the classical sense — we published research, and we trained a lot of students in a very important technology over the course of this project.

And the reason is that the expertise you gain from actually doing the project is really different from what you gain by reading papers by people who do these projects. Particularly in the LLM space, where frankly a lot of what is published lacks depth simply because there is quite a bit of value to the know-how.

This creates a massive future opportunity for the people involved and for the communities they belong to. They gain experience in a technology that is quite complex — where the number of people who can actively contribute is in the thousands or tens of thousands, not millions. And there's a lot of value to that.

When they leave the Apertus ecosystem or the Swiss AI Initiative — the wrapper around it — they go out and join startups, join tech companies, or start their own. And in essence, there’s a whole new class of folks going out into the world taking what they've learned and designing innovation tools around those problems. That’s where I think a lot of the value comes from.

Apertus is a model. We've released it. In a few years, it’ll be completely outdated. But the experience and understanding gained by the people who participate in these projects is a really valuable resource — particularly for a small country like Switzerland.

**Danny** Buerkli: That's excellent. And it confirms that around the launch, a couple of things were not necessarily well understood. That point was not well understood — that there are significant positive externalities beyond the artifact itself. And the other thing that may not have been well understood is what the artifact actually is. It’s the engine — or the fuel rod — rather than the entire power plant. And there's a big difference between those two things.

You mentioned earlier that in a couple of years this model will be obsolete in terms of its performance. I wonder: did we just happen to be at this one fortuitous point in time where the compute resources available in the supercomputing center were just right to pull off a model that was open-source state-of-the-art?

But is that now gone? Because we don’t know how scaling will continue. But already now, if you compare it to Grok-4 — where some numbers are available from Epoch — Grok used something like 60× the power and 70× the FLOPs. And that’s today. Not asking what that gap will look like next year or in two years.

Is this a repeatable exercise, or was that it?

**Antoine**: My answer to the first question is yes, and to the second one is no. It is a repeatable exercise, and this is not just it. But it requires a commitment to grow in scale at the same time.

The story behind this is quite fascinating. Around the time GPT-3 came out in 2020, the folks at the Swiss Supercomputing Center were coming up on an infrastructure investment cycle. And they made this very wise — but risky — decision: “Okay, there seems to be something to these LLMs, and we need to provide capacity for this type of research.”

And so they bought 11,000 GPUs with that infrastructure investment — at a price people today would vomit at, given how aggressively amazing it was. The biggest cost of running the cluster, I believe, is the depreciation on a daily basis — nothing else.

But that’s what enabled the Apertus project: having access to such infrastructure.

Now, as I mentioned, it would be difficult to do that a second time around, because now everybody knows the value of a GPU. NVIDIA is not giving the same discounts. But one of the points of the Apertus project — and others in the Swiss AI Initiative — is to show what can be done when there are resources available for research and development.

And yes, we can point to xAI and OpenAI and Microsoft and Amazon's very large clusters. But something Europe needs to understand is that it doesn't have companies of that scale building their own data centers. And so if Europe wants an innovation ecosystem that can compete with what big players are doing in the US and China, the only place where the investment in data centers can come from is the public sector.

These data centers are expensive — but not for an entire continent.

So the big question is: can we make the investments into the necessary infrastructure to enable this development, which in turn would spur an innovation ecosystem around it?

There have been efforts — including the construction of AI gigafactories coming online in the next few years. These are awesome initiatives, and we should continue and expand them.

**Danny** Buerkli: So what you're saying is: absent that, there’s not going to be another training run because we're tapped out as it stands today.

**Antoine**: Well, I would say that at the same time, the frontier is plateauing a tiny bit right now. And in fact, there’s less gain to be made just from scaling up compute on pretraining runs. People are conjecturing that we can scale up post-training to the level of compute of pretraining.

But in terms of pretraining: yes, we could have larger models, larger architectures. But data is going to become a choke point at some point — though you can repeat data to get a bit more bang for your buck.

We can still attack these limits on the open-model side without too much issue. It's the next generation I’m more curious about — where synthetic data really takes off as the primary data source for pretraining, and where post-training compute is scaled up massively.

Right now, I think we still have enough juice for something like Apertus-2 or even Apertus-3. After that, it will require larger infrastructures.

Really, the blocker now is how much experimentation we can do along the way. For the large run itself: on 4,000 GPUs it took around two months to train Apertus. We could double the scale, and it would take four months. This is an insane amount of compute for academics — but possible when you have a supercomputer.

The question becomes: how much experimentation can you do ahead of time on this cluster — a shared national resource — and how many scaling laws can you run?

That's where you're more limited than the large tech companies.

Another thing to remember: inference is now two-thirds or even 80% of the cost of a model. These huge clusters are used not just for training but for serving. We don't serve the model afterwards — we put it out in the open, and others take care of that problem using private cloud infrastructures.

**Danny**: Right. And from what I understand, it is much easier to serve the model through public cloud infrastructure than it is to run a training
run. 

One thing I don't understand: when we think of compute as publicly funded infrastructure — you mentioned depreciation — and I'm not quite sure how to think about this. We know how to build roads, bridges, railways. But most of that infrastructure doesn’t depreciate as fast as an H200 does. I don’t know the exact lifespan, but presumably it's measured in low-single-digit years.

Which implies recurring investment. Not water pipes you lay and then use for 50 years. It requires recurring investment. What is the correct way to think about that?

**Antoine**: Yeah. Caveat that I'm not an accountant by any means. But yes — GPUs depreciate faster. Not because they become useless, but because they become outdated. You may have four years of useful life, and depending on NVIDIA’s cycle, you get three new generations, all with way more theoretical FLOPs and different networking capabilities.

The difference between GPUs today — like the new B300s — and four years ago — A100s with 40 GB — is night and day. And I can only imagine where we’ll be four years from now. So yes, depreciation is faster.

But that doesn’t mean the chip is no longer useful. You can use them for teaching purposes. Students don’t need top-line GPUs to learn GPU programming, multi-node scaling, etc. You can build local clusters in places that never have access to compute. There are many educational purposes after depreciation.

But yes, in private companies depreciation is even more aggressive — two years — because they know they need to buy the next generation to stay competitive.

**Danny**: When thinking about the model’s performance, what are the binding constraints? One interesting thing about what you've done — because you've trained on “fully compliant” data — is that we can see what the performance penalty might be relative to training on not-so-legal data, which is presumably what most commercial providers do. But there may be other binding constraints: talent, team size, compute. What are those constraints, and how big would the performance uplift be if you had a bigger team, or trained on all available data?

**Antoine**: Yeah. There are many binding constraints. The question is the impact of each.

Regarding compliant data: we’ve been able to study this. One example: public datasets are usually constructed from Common Crawl. You take snapshots of the web, combine them, dedupe, filter by quality, etc. One thing you can do at the start is look at the URLs and their robots.txt files.

Common Crawl does this automatically — if a website blocks crawlers, it's not included. But only at the time of the crawl. If in 2021 a site didn’t block crawlers but in 2025 it does, the 2021 snapshot still contains that content.

We took a stricter approach: we retroactively removed all websites from these datasets that — as of January 2025 — had opted out of crawling. When we measure the impact of this, it's actually quite minimal. It only removes about 8–8.5% of the data. Performance-wise, we don't measure a big gap.

But there are datasets that are not public — like large dumps of pirated textbooks. You can train on that data. We did not — because we didn’t want to release that publicly or steal IP. We know that training on pirated textbooks can give you a substantial boost — 5–10% on benchmarks like MMLU.

Whether that’s because the textbooks contain useful knowledge or because MMLU overlaps with textbooks is another question. But the point is: training on private data does give measurable uplift. So that’s a binding constraint.

The question is whether that translates to UX challenges — that is, whether users would notice. If you had a similar interface, similar post-training, similar layers on top — I’m not completely sure users would notice that the model was worse.

**Danny**: You mentioned earlier that not everything inside companies is published, and what is published may be anemic. If you hold the data constant, the compute constant, the team constant — but had access to the expertise inside the three large commercial labs — what kind of uplift would that give? What’s your guess?

**Antoine**: If the data is constant, compute is constant, and the only difference is having 30 OpenAI- or Anthropic-level engineers? I would expect a substantial difference.

But it’s not often a matter of team size. The more you scale up the team, the more integration nightmares you have. On Apertus, hundreds of people contributed, but only a small, carefully curated set of changes made it into the final mixture — because any new thing can cause problems.

If you have 30 people from OpenAI or Anthropic, you don't need to try as many formulas because they know what works. That expertise is literally valued in billion-dollar packages.

But that doesn't mean you can't do meaningful work with a different profile. As a public institution, we can’t pay billion-dollar salaries. And we have a mission to train people — so many components are student-led. And the students at EPFL, ETH Zürich, and others are brilliant.

**Danny**: What was something you discovered while building the model that surprised you?

**Antoine**: Oh gosh — everything was surprising. One thing that surprised me was how robust and stable a lot of this is. There’s an incredible amount of work we were able to build on — pretraining libraries, data-cleaning libraries, all developed by others. It derisks the enterprise a lot.

I have massive respect for the first person to train a model at GPT-3 scale in 2020. I can't imagine what that was like. Luckily for us, we came later — and lots of decisions have already been documented by open developers like EleutherAI, LLM360, AI2.

Our big contributions were data compliance and large-scale multilinguality. And we were able to dedicate research resources to those because other parts had been handled by the community. That was a pleasant surprise — the community effort.

This is one of the promises of why open modeling can keep up with the frontier. Once you're at the frontier, every design decision has to be tested. But in the open ecosystem, you can build on others’ work.

Another nice thing is that it’s a smaller ecosystem — so what's published is less noisy and high-quality compared to the general research landscape where 90% of papers only work under narrow conditions.

So I was surprised by the quality of open artifacts and how much they derisk the enterprise.

**Danny**: In terms of public investments — what would you ideally want to see from the political system, not just in Switzerland?

**Antoine**: There's what I’d wish for and what’s possible. But I’ll say the ideal:

We need to be able to keep designing and training fully open models that are — if not at the same level as the frontier — very close. And as the frontier expands, we need the capacity to train open models at that level.

You’re forced to think five years ahead — the timescale for infrastructure investment. And we are running out of human data. Synthetic data becomes super promising.

Whoever has the best synthetic data in the years to come is probably the entity that can train the best models. And the best synthetic data comes from the biggest models.

You don’t need efficient models for generating synthetic data — you can sacrifice efficiency for quality. That still pushes toward training absolutely massive models — 10× larger than the biggest today. Think trillions of parameters.

To train those, you will need very large-scale infrastructure.

Whether that’s possible politically is another question. But it's what’s needed to avoid surrendering capability to a handful of private companies.

This requires substantial investment, will, and coordination. But it's not impossible.

**Danny**: It seems an oddly consequential variable is whether we can geographically distribute these clusters for political-economy reasons. If we can, a lot becomes easier. If we can’t, and they must be colocated, it's harder.

**Antoine**: This is a very important question. Politically, distributing clusters is more digestible — and that’s the current European model with gigafactories of 10–25k GPUs each.

But from a training-paradigm perspective, it's limiting. I don't think we could train a model 100× the size of Apertus by fully using 10 supercomputing centers across different countries with today’s paradigm.

So the question becomes: can research advance enough to create viable solutions for multi-node, cross-data-center training?

Within a single cluster, communication is the bottleneck. Cross-node communication is slow. Cross-data-center communication would be far slower. You’d need a new training paradigm.

One approach is training slightly different models in each data center and merging them. But that doesn’t scale the model size.

There are smaller-scale attempts at decentralized training, but I don't think they adapt well to this case.

Absent advances, the only way to scale models is by having a single very large data center. And that's politically hard. And environmentally hard. You need massive power, cooling, renewable energy. Only a few places could support a 5-gigawatt data center without ecological disruption.

**Danny**: What should I have asked but didn’t?

**Antoine**: I think we hit most points. But another reason to have these models is the sovereign aspect. I don't think every country needs its own model from scratch — but they do need representation in how models are designed and trained. The best way to be represented is to be a player.

Surrendering all AI development to a few big players because they have an advantage is shortsighted. The winners of the next computing revolution are typically the winners of the previous one.

The biggest players in LLMs today — Google, Microsoft, Meta — invested heavily in cloud in the early 2010s. Those that didn’t invest were behind for GenAI. And the winners of cloud were the winners of the web.

If countries want to build innovation ecosystems for future computing revolutions, they need to invest in the one happening now — or they surrender future growth, taxation, and sovereign capacity to companies that already exist.

Open models are closing the gap with frontier models. Public institutions can make large-scale investments to close it further.

And we’re on the verge of a paradigm shift: the ChatGPT era is ending; reasoning models and agentic AI are coming. Now is a great time to get into the driver’s seat.

**Danny**: Brilliant. With that, Antoine, thank you so much.

**Antoine**: Thanks for having me. It was really fun to talk about this stuff.

---

## The Art of Facilitation – with Vishal Jodhani

2025-10-22 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/the-art-of-facilitation-with-vishal-jodhani/id1839231559?i=1000732925280) | [Spotify](https://open.spotify.com/episode/5genN8liN0FTPaQgjGHveQ?si=18a76ae5a9744f6e)

Vishal Jodhani, a master facilitator, joins Danny Buerkli to talk about what makes facilitation work. They discuss what makes for a good question, how to know the difference between productive chaos and unproductive confusion, and what is underappreciated about the Berlin club scene.


**Danny Buerkli**: My guest is Vishal Jodhani. Vish is a facilitator, one of the best I've ever seen in action, and someone who thinks carefully about how people gather and work together. Vish grew up in Mumbai and has lived in Brussels, Rotterdam, and Adelaide, and a host of other places before settling in Berlin, where he has been living for the past decade with his wife and children. Vish, welcome.

**Vishal Jodhani**: Thank you, Danny. That's a generous introduction, and I'm glad to be here.

**Danny**: Vish, what do you do when you facilitate a group?

**Vishal**: For me, when I think of the word facilitator, having lived in France, I think of the root of the word which is facile, to make easy. And that's what I often think. I'm like, how can I make things easier for this group? And if that's bringing in a question, helping them focus on the bigger picture, creating space for everyone to be seen and heard, feel a sense of belonging, It essentially is that is how do I make it easier for them to move towards a shared goal even if they haven't named one yet.

**Danny**: And what are people not seeing when you facilitate a meeting or a group?

**Vishal**: They're not seeing all that goes into the design of that before they enter the room. People often ask me to come and run a session, let's say, for half a day, And the session itself is probably 20% of the work. There's 70% that happens before, probably 10% that happens after. And there is so much of the work in speaking to everyone individually, really uncovering the deeper needs and what is not being said, all the logistics, the right space, the right setting for this group, all the inner work to be able to host myself before hosting others, the invitation and how that is crafted. And a lot of the work is essentially crafting the right questions before you come into the group.

So yeah, I think it's the pre work, both outer and inner, that goes unseen.

**Danny**: What makes for a good question?

**Vishal**: For me it is, is this the right question for this group at this time? So is this question one that helps us dive deeper? So for example, what is not being said here right now? Or what needs to be surfaced that's not being surfaced? Is it a question that needs to open up possibility?

How might we find a new way to reimagine the future of? Insert your own blank there. Or is it a question that needs us to focus right now? We've had a lot of conversations. What is the one thing we have to make sure we get right?

It's just knowing a question is our best friend when facilitating and have we invited the right question in at the right time of the gathering. Even the opening question, right, when people arrive, we often call it an inclusion exercise or a check-in moment, and the right check-in question, say, the right question for the right occasion, to move the group along the journey because the day is nothing but a series of well crafted questions put together. I ran a workshop on influence today, and my opening question, we stood in a circle, was who is one person who's been really influential in your life? And that somehow set the tone for the rest of the workshop we had together. So it's choosing the question on the journey.

But maybe also in addition, it's a question that's clear, that's concise, and that really frames what it is we want to talk about.

**Danny**: What's your theory or model of facilitation? What makes it work?

**Vishal**: Danny, I love these questions. It's a chance for me to step beyond my day to day doing, and it's almost a behind the scenes view of what might be happening. For me, it's about taking a group on a journey. It's understanding where this group is trying to go together. That's how I always design it.

I often draw a mountain with a flag and say, Okay, this is the goal. This is where we are going. And by the end of this time together, it's also what I ask every client stakeholder at the start when they work with me is by the end of the session, if this was wildly successful, what is the group leaving with? What do we want them to think or know? What do we want them to feel?

What do we want them to do? And just spending time with that question on why are we gathering? What is our purpose, our intention, and where do we want to get to the end? That's step one for me. And then it's figuring out, okay, where are they starting?

What are we working with? What is what are the hopes, the fears? What are the elephants in the room that we want to work with? So I get a sense of where people are now and where they're going. And facilitation then is how do you design an intentional journey to take them from here to there?

And along the way, might be creating spaces for people to just derive, spaces or questions for them to express, spaces to listen, spaces to ideate and diverge, and spaces to focus and converge. And holding a mentor of mine often says is, Prepare like Helen and let it go. And a lot of it is also improvising. Once you know this is where we go. And we might have something, but what the group comes up with might totally shift the direction.

So it is working with the energy in the room and trusting the collective wisdom of the group versus just having a set agenda for myself. So it is that. It is having some structure and holding it lightly and inviting people along the journey to I always do that. I always invite people to like shape this as we go. Yeah, and going with a clear intention, a few good questions, and to not do it alone.

That's another big one is to hold space together in some way. See, I don't think of it as a solo sport. It's a team sport.

**Danny**: What are the non obvious limits to facilitation?

**Vishal**: There are so many obvious limits that I'm thinking of. You know, I often think of these principles that we talk about in facilitation, which is whoever shows up are the right people, what happens is the only thing that could have happened, and it starts when it starts, and it's over when it's over. So sometimes I think of the obvious limits of just holding an intention and knowing that we might end up we might to trust the process is something that you often hear with facilitators, trust the process. And it's just trusting that we end up somewhere, and where we end up, there is some meaning and data in. What is a non obvious limit?

What comes up for you, Danny, when you think of it, a non obvious limit in your experience?

**Danny**: I'm thinking of limitations that may be apparent to you as an experienced practitioner, but that may not be apparent to others who have less mileage in the field.

**Vishal**: I would say when I was new and naive, it was having a plan and following through with it, making sure I just cover the agenda that I had in mind and for that to drive it. And I noticed that there's only a certain part that I can prepare, but I'm facilitation often is working with the unknown. Just knowing who shows up in the room, who doesn't show up, or who suddenly shows up, that will have a big impact. It's not being able to control how people show up. It's not knowing what their own context and baggage is, what the interpersonal dynamics are at play.

So there is a lot of unknowns that you just walk into, and it's being open to entrusting that there's a big percentage of reality in the room that you have no control over. And can you still hold structure and just acknowledge your own limitations? Like, I might not have all of these answers, and can I still have enough trust and stay centered within the chaos that emerges for the group? Because at the end, what we're trying to do is create a solid, safe enough container for the right conversations to happen. So maybe that's what's in my control is the container and the growth is how do we expand the strength and the capacity of this container to hold the chaos that might emerge.

So maybe that's the big learning is facilitation is a chaotic process. And there's the mix of chaos and order, and figuring out what that thin line is that we gotta walk in the process.

**Danny**: And how do you know the difference between the productive creative chaos and unproductive confusion?

**Vishal**: That is a massive role of the facilitator, right, is to keep things on track. So once we have a sense of what the journey is, how people bring along the journey, And it happens. I've had a participant who suddenly said something which feels like it's a total derailing moment for this conversation. And then again, it's a choice to be made as a facilitator. Do we go off road for a bit and explore maybe there is something here in this creative chaos or in this off comment?

And if we are clear on intention, and if we feel it really is off, then as a very tactical thing, I would just get a little parking lot and say, hey, it's a great point. It doesn't serve us right now. So why don't we in today's world, we call it a bike rack, not a parking lot to be more sustainable. But how do we do that? And when that happens, would sometimes ask like, hey, Danny, that's an interesting comment.

I'd just like to check-in with you on how do you feel it relates to the topic at hand. And maybe there is a connection there that I, in my limited context of the organizing of the content, haven't seen myself. But it's it's almost inviting them back into that question. Mhmm. And more than anything else for me, it's it's designing for divergence and designing for convergence.

So I would even say, like, hey. For the next hour, we're exploring all the possibilities, and I'm encouraging creative chaos. So right now, we don't have to shoot an idea down, and once we come back, we'll have a filter. So based on our ability to execute, how this would move us towards our goals, and truly dreaming bigger, which ones of these meet the criteria? And now let's converge.

But it's giving people because every group has the dreamers and the doers. But do we create the right structure for both of those to happen?

**Danny**: Something that's really been bothering me is we broadly know how to facilitate good meetings. It's not secret, obscure, arcane knowledge. And yet, it does not seem to have caught on. What's the constraint here?

**Vishal**: When I work with groups and I often I was running a facilitation training for a group, and I interviewed every single person one by one and asking them their hopes for the workshop and their fears. And a lot of them were joining because they had high stakes meetings, multi stakeholder engagement opportunities. When I asked them what's holding them back from truly stepping in, it was the fear of the unknown. And I see that in the work that I do is people want control and certainty, and the focus is on content. It's a bunch of slides that I can present and that engage people in Q and A.

But even the Q and A might bring up uncertainty. But to truly open a conversation to surface the collective wisdom of the group with power imbalance in the room, with differing perspectives, it is that fear of conflict, the fear of the unknown, a lack of self confidence or trust to be able to manage chaos or conflict when it happens. And that gets in our way is when I talk about chaos and order, it's the leaning towards too much order and leaving the space for new things to emerge. Or maybe it is going on the other side of just being too chaotic. Hey, everyone's here.

What do we need to talk about? And that's why for me is that art of having some structure but holding it lightly. And another big thing that we talk about is often learning how to host myself before hosting others. And that, for me, is the non obvious work or the work that goes unseen is how much self management it takes before going in. All our inner voices to work through of what if I fail.

Am I enough? Do I know enough about I have that. I I work with different industries, companies with totally different contexts, I don't know much about the context or the people attending know a lot more about it than I do. And it's for me to remind myself my role here is to hold process, not to hold content, and that my the group has the answers. How do I help them surface it?

So the analogy I use often is that the participant is bringing all the ingredients. I'm just the oil in the frying pan. That's helping bring it together.

**Danny**: Who do you consider your teachers in this?

**Vishal**: One of my biggest teachers has been Toke Moller, who is one of the cofounders of the Art of Hosting or one of the, yeah, sort of the senior practitioners. And I felt the first time I came across The Art of Hosting as an organization and so many other teachers, whether Monica and Mary Alice and all these elders, when I attended my first training, I was like, Ugh, this just speaks to me. I get it the way they explained it, the way they held space. A lot of it was just role modeling, good conversation. I remember living in Brussels being invited to a co creation session and I never had an invitation to a co creation session.

And I, in my early 20s, entered this room, which was not set up as a classroom, not set up as a boardroom. It was just a circle and everyone stood around and it was such quality conversation, such deep connection and a really fruitful engagement. Even if it was just two hours in the evening on coming up with a new collaborative space, how I left feeling was so different to what I had felt in other meetings. And that was my invitation to dive deeper into the work and how all these teachers have brought in some of the simplicity of the art of crafting questions, on hosting yourself. It was more a master class in the being and the doing versus a bunch of tools that were there.

And what inspires me about these teachers has just been the way they do it themselves and having experienced it and knowing this is what the impact is on the group. And I've worked alongside them as I grew in the field and even then just witnessing how lit up people's eyes are in the room and how they feel the sense of belonging, a sense of ownership at the end. Yeah, it's tough to it's those intangibles.

**Danny**: Broadly, most or maybe all facilitation methods bake in some loaded assumptions about the world, like that less hierarchy in a room will lead you to better results rather than more hierarchy in the room. Right? And that can be intention with what the group, the organization, the company you're working with tries to achieve in the world. When do these contradictions start to matter?

**Vishal**: The contradictions of hierarchy and group?

**Danny**: One way to think of facilitation is that it's not just a neutral tool. It's not just a hammer, but it's, in fact, it's an opinionated hammer. And the the hammer has opinions about the world and about how the world maybe should look like or what the world should not look like. And you can say every method brings with it certain assumptions. It's not as if facilitation is particularly special.

But it does strike me that a lot of facilitation methods do bring with them very strong assumptions about what a good world looks like, let's say.

And it's not always obvious that everyone who would like to sort of enjoy the fruits of this method actually agrees with the underlying ethos of it.

**Vishal**: It's true. For example, facilitation, we often I love using a circle because a circle shows equality. There's no hierarchy. Everyone can see everyone, but a lot of the organizations that I work with have an inbuilt sense of hierarchy in that. What I notice is if I'm running, let's say, a top team off-site, and I'm facilitating very senior leaders in an organization for two days, even then what I'm working towards is being brought in as an external, having the privilege to say, I am doing that by allowing others to be present in a more equal way.

And I'm neutralizing the group in some way while acknowledging there is hierarchy in the room. So I would sometimes get a CEO to open and welcome in frame, saying this is why we are Galleria. And then to say, and I would love to participate myself, I'm stepping back as a participant and letting someone else hold us for the next few days. And I always have a talk with the senior most person in the room on what their role is, how they frame, how they show up, what their posture is for those two days, and what they experience is a setting where everyone was seen and heard and all opinions mattered. And then you're right.

They go back into the rest of the world in their senior roles and maybe into a default mode of hierarchical operating. But what I feel shifts, no matter what the world is, just that feeling of how might I make others feel included? How might I make others feel seen? How might I allow the collective intelligence of my teams to surface? And in that sense, it is not an eitheror of I need to give leadership and direction versus I need to invite participation.

But how might it shift from an either or to an end? So maybe just expanding their range and our range of both of that is sometimes direction and power is helpful. And there is time and space for that. And there are moments to invite people's voices in because in the long run, we do want to explore that power of the collective at some level. I haven't come across someone resisting it, but it's been more traditional organizations where it is I tell them and they do it.

Then I ask them, but why is that change not working? And then they realize that they forgot to take people along the way. But often it is helping role model an experience that they want to then bring to their teams.

**Danny**: What's something you believe about facilitation that most other people don't?

**Vishal**: That's a beautiful question. For me, my biggest belief has always been about hosting myself before hosting others. And I know there are some fellow practitioners and teachers who believe in that. But I so strongly believe in this is, am I clear on my intention, my authenticity, my vulnerabilities, my challenges? And what do I need to do to truly hold space for myself before I can hold up for a group?

And I have seen people who've just spent all the time designing and working and then showing up. And maybe that's what the difference is. A lot of people feel you can park that part of you aside and who you are can be left at the door to serve the group and what it needs. And I strongly believe how you show up impacts the rest of the group. It's that analogy we often use about the security announcement on an aircraft, when the oxygen mask makes you sort of put it on yourself before putting it on others.

Maybe it's that.

**Danny**: When should you be a chill host?

**Vishal**: Priya Parker often talks about this, and she's got a chapter in the book called The Art of Gathering on don't be a chill host. I notice sometimes I am a chill host, and it's if I have spent a lot of I spend so much of my time crafting conversations, designing gatherings in a work setting, that when I sometimes host in a more private setting, I allow myself to be a chill host. I invite people in with an intention, and I struggle because I struggle to put that hat off. Like I've invited a bunch of people to this gathering at home and there is intention behind the gathering. And do I step in and do more or do I step back and do less?

And over time, I have learned to create the right setting and then step back a bit because I felt I got so into hosting that I would end a gathering and everyone would have such a good time and I feel I didn't get to enjoy it myself because I was just so busy hosting. And that's what I that's what my practice right now is is to trust and surrender. I've called the right people. I've put in that 70% work before the gathering, and I want to enjoy this gathering myself. So it is bringing a certain amount of chill, but it's it's a practice.

**Danny**: Who does better coffee and why? Berlin or Adelaide?

**Vishal**: When I moved to Berlin eleven years ago, I really missed coffee from Adelaide. The Australian coffee scene was amazing. And there was just one place I'd go to to get proper coffee, and it was Aussies who were serving it as baristas. And I feel now over time, Berlin has an incredible coffee scene that I'm so grateful for. And maybe Berlin does better coffee because we've got the world here, and everyone's having their unique flavor to it somewhere.

That's working for me.

**Danny**: If we stay on Berlin, what is underappreciated about the Berlin club scene?

**Vishal**: I am so grateful to the Berlin club scene. What is underappreciated is the fact that you can spend twelve hours at a party and almost you'd never get asked the question, So what do you do? I have a friend who I know from a club in Berlin, and we met and we connected, and I've been with them to different music festivals. And we caught up for dinner last month. And at the end of the day, he's like, hey, Bishop, I've known you for three years.

I still have no idea what you do. And I was so grateful for that. I'm like, we have a friendship over three years that is beyond our roles and titles, which is just based on an appreciation of life and shared values and interests, and that is underappreciated. That question would be asked in London or New York and it would be such an easy icebreaker. Berlin doesn't care about that.

It is such an equalizer. So it is that it's not being boxed in. It is allowing you to show up in your authenticity, in your rawness, and and all of you is just welcome. And how that gives perm how we give permission to each other to then show up fully.

**Danny**: How does that practically operate?

**Vishal**: How does what practically operate?

**Danny**: This norm that you've just described. Everyone's automatically clued into that. How?

**Vishal**: There's a lot of depth in the Berlin club scene, in my humble opinion. I'm no expert. But Berlin for me celebrates authenticity and breaking barriers. And there is so much attention just on the being and the presence. There's a lot of respect for our private personal space, a lot of respect for our personal space and being interested in the humans that are there, you know, through conversation or no conversation.

But yet, there isn't small talk. I don't know if it's a German thing or a Berlin thing, but because these are places of depth and you are there to really to, I don't know, like, shake off the your concerns from your day to day, if it is to do some inner work on the dance floor, if it is really to explore and discover who you are and give each other space to do the same, somehow that would be a buzzkill. Right? I'm not not not here to talk about my taxes or the weather or what I do for work. There is more to me and by it also takes so much effort to get into some of these clubs that once you're there, you wanna be in the temporary alternate universe that has been co created with so much effort and intention for you that you don't want someone to bring you back to your default world for those hours.

And I have by a tourist been asked that question, they're like, oh, and you can immediately notice it. Like, there was there was just something off about this moment. And I'm not the only one who would say this, but clubbing in Berlin and experiencing community and self expression does ruin partying anywhere else in the world for a while.

**Danny**: How does the fact that you grew up in Mumbai shape your work today?

**Vishal**: I'm so grateful to have grown up in Mumbai. For me, Mumbai was a few things. One, it was about survival of the fittest, and we always had to work harder. And if I didn't get good grades at school and if you were my classmate, Danny, my parents would say, Hey, you and Danny play all the time. Why is he the topper in the class and it's not you?

But there was always that push to work harder because there were many people and fewer opportunities. So this idea of hustling and showing up for yourselves and going that extra mile was we were well conditioned for that, even if I think of just dodging Mumbai traffic. If I can drive in Mumbai, I can drive anywhere else in the world because it was about, hey, we're going to put you through chaos. And Mumbai traffic, by the way, is a great example of chaotic world that we spoke about earlier. It's minimal structure, maximal chaos, and yet we all find our way even if we take extra time compared to Switzerland to get somewhere.

And it's about finding your way. But more than anything, if I think of these things, it is the art of possibility. When I live in Germany, something that I struggle is when people say, nine. Can I just have these three changes made to my dinner order? I'd be like, No.

Or there are lots of things with bureaucracy that you'd get to know. Whereas in India, it was always possible. There was always a way no matter how crooked the way might be sometimes. There was a way of, yeah, absolutely, we'll find a way to work it out. So for me, was not about this can't be done, it's how might we get it done.

Let's find a way. Let's explore possibility. And that has helped me if I try to figure out my own path career wise, if I bike my way through the streets of Berlin, there is that grit and resilience that is that I'm grateful for.

**Danny**: What did you learn from playing cricket that you couldn't have learned anywhere else?

**Vishal**: I was really bad at cricket because I was one of the younger kids amongst the neighbor community that we lived in, and I was often the last one picked on the team, which wasn't great for my self confidence growing up. So while a lot of people learned amazing things about cricket and from cricket, my big lesson was to not to really choose your sport because it wasn't wasn't mine. And I could have been the last big player on a team in a sport for a long time. That have been my worldview. Or I could have chosen to go a different direction and pick a sport that works for me.

And for me, that was a big thing of India was about grades, was about being good at cricket, it was about having certain career aspirations. And it was very tough to go off roading from what was considered good. So for me, not to play cricket as much, for me choosing to drop out of a career in medicine, to choose to go take a gap year, were things that weren't done or celebrated, but it was me trying to figure out how do I do what's right for me, not what society expects of me. And cricket is a good example there.

**Danny**: You're charismatic. What's your theory or explanation of how charisma works?

**Vishal**: I appreciate your generous compliment, Danny. I think you're very charismatic yourself. So I should probably pass the question right back to you. For me, people that I find charismatic, what I appreciate the most is authenticity, is their ability to truly be and own who they are. I think of the metaphor of those Japanese broken balls with a golden thread around them and this kintsugi as an art form.

I mean, that's what I think of charisma is not perfection, but people who can be themselves, shine their light, own their shadow, embrace their flaws, even highlight them maybe, but, like, this is me. For me, so much of my life journey has been trying to be someone or be someone else or aspire, do more, more, more. And when I see someone at ease with who they are and show up with that effortlessness, It's that less is more that feels charismatic. It's that ability to pull not push, trust and flow with ease. And maybe those are the people who I feel are not working hard to be liked by everyone and diluting themselves along the way.

But hey, this is me. And there's a lot of work to be done to get there. We all seek validation in some way and to stop yearning for that and giving it to myself is the world that I would love to work on. If,

**Danny**: Let's say, your great grandchildren study the Jodhani method of facilitation, what would they find?

**Vishal**: My goodness. They would find that it is truly a family recipe that is coming from that has its own flavor and comes from opening your cabinet and just finding what is there. They would find it truly the signature feel is a sense of good connection and community and about creating belonging for more people, sparking joy, so joy would be one of the ingredients, and seeking for that overall well-being of the community.

**Danny**: Turning to a very different community, you've worked also in the social impact space in the past. And without being too critical of that scene, it's hard and I've also been part of this. It's hard not to feel that the scene hasn't quite fulfilled its promise. Why?

**Vishal**: For me, it goes back to this either or versus and. I noticed there was always this tension or the focus on doing good and the focus on impact and focusing on the work that was happening. In some way, there was a lot of possibility left on the table in terms of making it financially sustainable, which is sustainable in some way. And I'd always reframe it as another question, which is how might we do good and do well at the same time? And finding that balance between those, it's yes, how do we measure our impact?

It is okay for us to do good along the way and to do well for our teams. Our teams don't have to be underpaid. We can attract top talent and create good conditions for them to thrive and to recognize their effort. But it doesn't have to be stepped down for anyone. So for me, what's that is there was an abundance mindset and a sense of possibility when it came to impact, but not when it came to operations.

Yeah, so that's my personal take on it is how might we do both? How might we bring that sense of possibility and abundance to creating conditions for everyone to thrive along the way? There were so many amazing people who I met on the journey, people like yourself. And it was people would say, yep, I can contribute on an advisory capacity. I could contribute as a mentor, is also what I do right now.

And how might we create possibilities for people to invest more in it, not not out of charity or generosity, but because it makes sense for them. I feel I feel this is why we lose out on people somewhere. The people who do the work, they burn out for different reasons. So so when I say sustainable, how might it truly be sustainable for people? Mhmm.

Or is it just a stop in their career somewhere?

**Danny**: You do a lot of innovation work. What's your theory of innovation? How does that come about apart from well, I guess it's new, probably.

**Vishal**: You know, in my undergrad, I studied communication, and I majored in advertising. And for me, working with ideas and people was exciting then. Innovation was about how do we sell this in an interesting way. And I did that for a while, and then I realized we were so good at selling and being innovative in how we sold things. That was my limited view of innovation, that we might sell things to people who don't need them as long as things sell.

And when I went on to do my masters in innovation because I knew I like working with ideas and possibilities, this idea of human centered design or designing just with the human and the core of it was the period of innovation that had a lot of impact on me because it felt like innovation was going upstream. We can work directly with users and their unmet needs, understand their stories, their context. We'd sit with them in their living rooms, have coffee for an hour, get a sense of their lives. And then when we could come back with all those pain points and needs and then design from that place, that felt truly innovative. It wasn't innovation as in a bunch of creatives sitting in a room with bean bags and Post its and smoking cigarettes and coming up with ideas.

There was people out there in the field observing people in their day to day, having those conversations, getting inspired by possibilities in other fields, and then working as a multidisciplinary team to then create solutions with well crafted design questions, rather typical, how might we design a service that enables this? So for me, that has been the biggest shift and that has been my period of innovation is, one, how do we define the right design challenge and frame that as a good question, anchor in context and needs? Two, how do we then observe and empathize with the users? Three, how do we look at inspiration and diverge into a sense of possibilities? How do we ideate both from a divergence and a convergence point of view?

How might we then prototype something? So this idea of getting it right versus getting it good enough and then having as many iterations along the way until we find that fit and then scale it. But that journey often considered design thinking. But for me, it has been not just having a good process but facilitating that process well with the right people in the room and the right tools and giving ourselves permission to fail and iterate multiple times until it eventually makes sense for the end users we're designing for. And from a facilitation perspective, it's taking people along the journey, designing with users, with our teams, testing it together with people.

So I managed with my innovation work to bring in different parts of my life together, The advertising and the storytelling, the interest in humans and deep connection, the fascination for design thinking and all the work that has been done by companies like IDEO, and then just hosting good spaces to bring all of that creativity to the table. It's been really fulfilling as work.

**Danny**: Michael Pollan, the journalist, his advice on he memorably said, you know, what we should eat, and the answer was eat food, not too much, mostly plants.

**Vishal**: I love that. Yeah. You could read the book just you could get the essence of the book with the with the title.

**Danny**: Right. Eat food, not too much, mostly plants. What would be your equivalent for what you do?

**Vishal**: Show yourself, see the other, explore together.

**Danny**: What do you wish I asked for having so far?

**Vishal**: What keeps me energized about the work that I do?

**Danny**: Tell me.

**Vishal**: I've had a lot of iterations in my own work, including right now. And energy is a topic that I pay a lot of attention to right now. And I do a piece of work, then I pay attention, do an energy audit during the work itself and after the work, and I ask myself, did it give me energy, or did it take away energy? So it's still work in progress. And the learning that I have is finding my own inner compass in the work that I do.

And that compass for me right now has been just an articulation of my purpose, which is to bring joy, well-being, and love to the world. And it's not something I would say out loud in a business setting necessarily, but just intuitively knowing that for myself and naming it, When an invitation comes my way, even if it sounds big and scary, I tune into it and I ask myself, would it allow me to serve that purpose in some way? Or how aligned is it to my set of values? Like this is an important piece of work. Connection is a core value of mine.

So if I think of work that I do with facilitation or storytelling or coaching, like, oh, I really care about human connection, and this is why this word excites me. Or I say no to interviews or opportunities just because I feel it's not aligned. They sound great on paper, but they don't feel energy giving or as aligned. But it's like purpose and values. And it's also an invitation to others as we tune in because sometimes it happens implicitly.

Just explicitly naming it like, wow, what excites me about this work? I end up doing pro bono work. I end up taking on opportunities that are way out of my comfort zone sometimes, but it's more I just feel because we spoke about burnout earlier briefly. It is how do we find for ourselves, how might we find for ourselves that the deep inner source of energy to allow us to fully step into the work that we want to do.

**Danny**: With that, Vish, thank you very much.

---

## State Capacity and Government Reform – with Don Kettl

2025-10-08 | [Apple Podcasts](https://podcasts.apple.com/us/podcast/state-capacity-and-government-reform-with-don-kettl/id1839231559?i=1000730771456) | [Spotify](https://open.spotify.com/episode/7lqZTlPj1BhL2DeWMi8Q2o?si=44287339ff984566)

Don Kettl — prolific scholar of public administration — joins Danny Buerkli to talk about state capacity and government reform. They discuss what DOGE got right (and what it didn't), whether gradual government is possible at all, why Operation Warp Speed was so unreasonably effective, and what lessons we should learn from Paul Volcker.


**Danny Buerkli**: My guest today is Don Kettl. Don's the former dean in the school of public policy at the University of Maryland and was most recently a professor at the Lyndon B. Johnson School of Public Affairs at UT Austin. 

He has had and, in fact, still has a prolific output. He has written well over 30 books on government reform and state capacity, plus countless articles and columns, and has been an adviser to many government executives.

Don, welcome.

**Don Kettl**: Danny, it's so good to be with you today.

**Danny**: Thank you. Now to start us off, I'm wondering that, really, the last time government reform writ large, was sort of showed up in mainstream debate, I believe, was in the nineties under Clinton and Gore, of course, with the National Partnership for Reinventing Government. Why has this idea become so salient again now?

**Don**: There there are a couple of reasons. The first is that every president in The US since has felt the need to launch his own government reform agenda. And so there's been a president's management agenda for every administration since. But I also think that what had happened was that a lot of the reform ideas that people had on the table had in a sense, if not run out of gas, at least run out of enthusiastic backing. It was just not clear, for example, if Kamala Harris had won, what they would have done, what would have been the big new thing that they had rolled out.

And in fact, I was in conversations with with some of the people who were working on putting it together, and it was they they would have had an agenda, but it wasn't terribly exciting. Wasn't the kind of thing that would either have have grabbed the media or especially the people inside the government. So one of the reasons why it's back, I think, is that there's a sense that that the old ideas that have been working had essentially run out their course and there's time for something completely new. So that was one thing. The second thing is that there has been a a growing sense of of concern from both the left and the right about what government is and how it works.

From the right, there was, of course, the the sense that government bureaucrats had run amok, that they were uncontrollable, that had devolved into this deep state that Trump talks about all the time. So there's there's that piece of the problem that that the right had thought that that we needed something much more fundamental to try to shake things up and to try to especially shake up the power of the administrators who had gotten far too much power inside government. And then from the left, there's also been this problem that is rooted in the abundance movement, where there's an argument that the bureaucracy, in fact, if not out of control, at least had tied all of us up in too many rules and regulations. And so we needed a much more fundamental approach to changing what government does. And so the left is arguing that a lot of the problems that have happened are of its own making, and the right is arguing that, yes, that's right, the problems are of the left's own making.

And the left says what we need to do is to deregulate and get out of the way and to try to fuel abundance. And the right is arguing that, no, no, we don't need that. We just need less government. We need fewer government administrators. We need fewer government programs.

We need less government spending, and come hell or high water, that's exactly what we're gonna do. And so what had happened, at least in The US, was the sense that the reform movement had essentially just just run out of gas as it was time for something completely new. And what we got because Trump got elected was the perspective from the right that amounted simply to trying to to take government, not just down to the the basic studs and the framing, but down to the foundation or worse.

**Danny**: Without, of course, necessarily endorsing it, what did DOGE get right?

**Don**: DOGE got right the idea that there really needed to be fundamental change and that it was possible to instead of thinking about small incremental steps, that that big, much more dramatic changes were needed. And so they they came through barreling through government. And the other thing that they had was an instinct about where to go to try to to get leverage. So that they focused first on personnel and particularly about how to try to fire government employees. And so that was the the first thing that they focused on.

Second thing is they focused on the structures of government. And so they went after a handful of agencies, including, for example, the Agency for International Development, and they decided just to just to wipe them out. And so that they at least had an instinct about the fact that structures mattered. And then the third thing, which I think has gotten less attention, but which also was an important instinct, is their eagerness to try to go after information systems and understanding how information provides a key for doing what they wanna do and to get get leverage and control over government. And more even more fundamentally, they were trying to use information to weave different information systems together and then use AI to try to probe for points of weakness.

For example, who is committing fraud? Illegal immigrants might be collecting money? And so their effort was to try to use information and AI and integrated information systems to try to get at some of their important goals. And so they had the right instincts about where to go to solve problems in ways that I think the left probably had not fully recognized. And DOGE did this in an integrated fashion.

The problem, of course, is that they went way beyond anything that the law allowed. And then in addition to that, dove into a lot of these agencies without having really any understanding about what it is that they did or whether what they found mattered or not. The complaint about AI is that it you've you ask it for information and it'll give it to you. But what you can't be sure is whether or not what it's giving you is, makes any sense or what kind of sense it makes. And so for a while that they were talking about the number of people receiving social security benefits and the number of people who were over 150 years old who were receiving benefits.

Well, turns out that the problem is that the systems that they were using were largely built on COBOL and other kinds of antiquated data systems. When they didn't have complete information on people, they ended up just having to put a code in, which then made it look like they were very, very, very old. And they produced then these, these figures that did not reflect the number of people who were over 150 years old receiving social security, it just people who were receiving social security for whom they didn't have sufficient information and which the system had no way to be able to deal with. And so that they in case after case after case, they dove into into issues thinking that they had found problems when in fact they didn't know what they were really looking at.

**Danny**: You're in a sense saying it's a fundamentally, the approach may may not have been completely wrong. It's a skill issue almost in in its application. I think the one question that I've been interested in is if it's not DOGE, and if it's not what we've tried before, which also clearly does not seem to have worked, what is it?

**Don**: That is a huge problem in American government right there. And the the core of the issue is this, and I I end up making all my friends on the left and the right terribly unhappy when I say this. But I think that that Trump in general and DOGE in particular had the right instincts about the nature of the problem and where to go to try to solve it. Where they made an enormous mistake was in the kinds of strategies and tactics that they used to try to advance, that it was kind of the wrong answers to the right questions. And the left, I think, has not yet caught up to that that sense of things.

So you raised the question about, well, if you're not gonna do that, then what should you do? And I think that there are lots of answers, but there are, but they aren't nearly as exciting and don't grab as much immediate political support. And that's the core problem. But, we do have structures that get in the way. There's no doubt that we have problems not only of hiring people, but also dealing with poor performers, and also even more important of matching the people that we hire with the skills that are needed to do the job that has to be done.

And then last, there is this enormous opportunity out there through information systems and through AI to try to modernize government. And so those are the, those are the things that I think that we need to do that the DOGE got exactly right and did it exactly wrong. And in my mind, I think the the fundamental problem is this, that there was a a a big instinct and the part of of DOGE and the Trump administration more generally to think very vertically, to think in terms of control from the top and reaching down to the bottom, and to silo off individual efforts and initiatives and agencies in particular for particular targets, with the exception of what they did with with AI. But in general, what they did is that they approached what's increasingly a horizontal world with pieces interconnected among agencies and sectors, and approached it in a primarily vertical way. So when you approach horizontal problems in vertical fashion, then what you're going to get is a, is an enormous mess, which is in fact what they've succeeded in creating.

And the other thing that's happened is that they've, the initial phase of DOGE itself has run out of gas. It's run out of political support, and now that having broken all the pieces, they're not quite sure what to do about it. And they're finding that they need to rehire some of the people that they fired because they didn't really have a strategic plan about what they wanted to do and how to, how to get rid of what it is that they didn't need. And in addition to that, the problems of information have ended up just tripping them up because they've, they've violated all kinds of norms and laws about securing the privacy of information. And so they've, they've got themselves in trouble.

What I think we need is a system that focuses much more on flexible hiring based on matching people that we need and hire to the jobs that have to be done, which the system's very bad at. I think we need to go back and rethink about not so much abolishing structures per se, but about how to try to interconnect them better, understanding that the federal government, of all the work that it does, only about 5% of the federal budget is work that it does directly itself. And the rest of it has to do with, with leveraging lots of partnerships on the outside, and that's contracts and entitlements and loan programs. And then that's even before you get to regulations, which aren't captured by the, by the budget. And so we have to recognize that reality and find ways of making the government more capable.

So just abolishing structures doesn't really get to that problem at all. And the, the way to get at that, I think increasingly is through information systems that allow us to understand what we're doing and get leverage over the system and over the partnerships, and to understand how to apply AI to scale in a way that's smart and intelligent, and that focuses on it in a way that that's really focused on on producing results that matter to people, which is what what taxpayers care about. And it's that that connection about weaving all those pieces together in terms of results that matter to people that really need to be what it is that drives it. And so I think they've, they've missed that. I think that the, the left is yet to try to grapple with that, but that I think clearly is the way in which we need to go.

And it's not that's really not quite what abundance is arguing, and it's certainly not what it is that Trumpism is arguing. So I think we need a a kind of a third way of going at this in a way that is, think, much more focused strategically on the way in which government not only was ten years ago, but the government that we need ten years from now. Because, 2035 will get here pretty fast and get here even faster if we don't spend the time in the meantime figuring out the government that we need because it's gonna take a while to build it.

**Danny**: Right. On this exact point, it seems one key question is, these reforms that you're talking about, can they be done gradually, or is the political economy such that you have to go in with the disruptive shock? Because the gradual reform leads you inevitably to some kind of tar pit out of which you will never find your way out.

**Don**: And that's the fundamental dilemma. If and I think what we had discovered was that the, as I suggested earlier, the incremental pace of reforms that essentially just got yourself stuck in exactly that tar pit, There were, there just were not good ideas anymore. There wasn't, all those instincts were kind of there. We need something that's much more disruptive and we need something that, that really shakes the system to the core, but which does it in a way that focuses on, on results. And that's, that's a hard thing to do because it's a lot easier to focus on symbols than it is to focus on outcomes and to focus on results.

And so I think that the days of just gradually bit by bit incrementalizing our way into the future, I think are gone. Doesn't mean that we can't do it with small steps, but we need to be very agile about how we do it and making a strong disruptive run at things, but to build in enough feedback so we can, we can self correct constantly. And, it's so maybe it's incrementalism, but with instead of small steps at a time, lots of steps as if you're running a 100 yard dash, but we need to run fast because the the world and its problems are not gonna stop and wait for us to catch up.

**Danny**: And it seems like in your own words, the the problem you're describing is not necessarily new. I think you wrote in '98, so not quite, but almost thirty years ago. And I quote, the field, which here is public administration, has shown a constant tendency to reinvent itself only to rediscover that new approaches frequently rehash old ideas. And you've also said that the field has struggled with the basic problem, the accumulation of knowledge. So not only are we not learning anything, also all the new ideas that we're coming up with or that the field has come up with are not actually new.

Why?

**Don**: Some interesting reasons for that in terms of just the intellectual history and development of the field of public administration and public management. Part of it is that it is is captured by a paradigm that's existed for more than a hundred years that looks at the core problem of figuring out how you create vertical hierarchical structures and focus them on producing the goods and services. It's a kind of a, this is too cruel and too crude an approach, but it's kind of a Henry Ford Model T approach to the field. And the field has changed dramatically since then. There's been a lot of focus on networks.

There's been a lot of focus on information systems, lots of other kinds of innovations, but it's still, I think at its core, captured by this kind of vertical piece because it's the way in which everybody was trained and it's the way in which a lot of the work gets done. But then on top of that, I think there's, there's an emerging problem of younger scholars coming up trying to feeling the need to demonstrate their technical proficiency and lots of statistical tools to try to make this feel more like economics in a way in terms of the the methodology. And what that really means is two things. One is that you need big data sets to be able to do that. And big data sets, not always, but typically exist only for problems in the past, because that's where we collected all the numbers, as opposed to leaning forward into newer data sets that are just emerging, but are kind of clunky and messy.

It's, it's the AI AI problem there again, about what it is that you, that you're focused on. So there's that. And then there's also a tendency to, to use the definition of problems in the past with old data sets to be able to find ways of of advancing issues. And that means that we too often don't really focus on all the problems we're talking about now. And, instead talking about what we can learn about the problem that somebody wrote in a journal five years ago that we can now advance incrementally through, through new data sets and new techniques.

And so the field is, is captured by by methods, by method of deciding which problems matter, and by this basic hierarchical approach. And you put all that together, and it's you think about it kind of precisely what it is that Trumpism complains about. Not about the field per se, because it isn't paying any attention at all to the field, but about government in general. And so it's part of a broader problem that I think we need to try to work on, and again, on now, because I think the thrust of our conversation so far is that we are facing whole new problems where we don't have ourselves equipped with the right kind of approaches, where the approaches from the left and the right are both wanting, and where problems in the outside world are moving so fast that we better move in a hurry, but at the same time, the field is unfortunately not providing the intellectual capital that's needed to try to drive reform. There's really no kind of reinventing government by Osborne and Gabler to try to drive a decade of change.

And the whether you liked the book or not, whether you think it was right or not, it was effective at at motivating first the Clinton administration and then Al Gore, and then a whole decade or more of of reinvention of the field and reinvention of practice. And we just don't have that now because we have more of a sense of of just just wallowing out there in all the problems. And if you talk to lots of academics now, there's a sense of, if not despair, at least profound worries that, oh, this is just awful about what's happening. And people I think are drifting back into a sense of despair because what's happening doesn't match their view of the world, but there isn't an alternative view of the world that has emerged yet anyway. So that I think is really our big challenge.

Not to write a bestseller in the field, which I'm not sure is possible because people don't read books anymore and it's hard to be able to get stuff out, but we need in a hurry, I think, to be able to refocus the field, to focus on problems because there is an urgency to try to reframe what government does. And it's, this is of course not just an American problem. It's a problem that, that every government in the world faces and where other governments that are not as large have greater urgency in some ways to to adapt because they can find themselves buffeted back and forth by by waves that over which they have even less control.

**Danny**: If we look at a possibly slightly happier angle to this, if you look at Operation Warp Speed, which was the procurement of COVID vaccines, by any metric, it was an absolutely resounding success, just an utterly outstanding example of government getting something incredibly difficult and incredibly important done. And yet, it seems oddly underappreciated. Why was it so effective, and and why is no one, talking about it?

**Don**: That's such a great question because it it truly ranks as one of the greatest successes that government has had in a very, very long time. And as we talk now, it's, it's easy for people to just drift back and forget about how big of a problem we were facing. But at, at the beginning of COVID, there were public health experts who said, you know, we might have a death toll of maybe fifty to a hundred thousand. And people were saying, oh, would, that's horrific. We can't even imagine that.

And it turns out in The US, the death toll was 10 times that. It was over a million people. The the economy was shut down. Kids were were not in school. Problem after problem after problem was was piling up, and it was clear that we had an enormous crisis in our hands.

And the, the previous all time record for development and deployment of a new vaccine was four years. And so the, there was a sense that yes, we could develop a vaccine to try to deal with COVID, but the work in the past suggested that it was gonna take a long time. And what were we going to do with the, with public health and with the economy in the meantime, it took four years to be able to develop? Well, what happened was the Trump administration at the time, and it's important to remember, was the Trump administration funded a group of, of companies to be able to investigate a variety of different strategies. They seeded a whole bunch of different kind of approaches.

And there was a competition among the companies to figure out the best solution, because each of them knew that whoever it is who got it was gonna make a lot of money because this is an international crisis of enormous potential. And so the incentives were for private companies to try to find solutions and find them fast, that government wanted to try to do that, that the research and development was going to be expensive. And so what they did is they launched it at warp speed. And it's a term taken from those people who are Star Trek fans, who knew that the warp engine was designed to travel at faster than the speed of light. And so they, that was the strategy to try to do that, find incentives for companies that wanted to do what it is the government needed to get done.

The government realized that it couldn't do it itself, but it could play a role in catalyzing the development of innovation through competition with private sector partners, and they in fact produced the vaccine and had it deployed in nine months. That's just a stunning, stunning, stunning success. And it had to do with this, not as we were talking before, it wasn't vertical, it was horizontal. It had to do with the creation of innovation through in information, and it had to do with structures that were primarily horizontal instead of vertical, driving it through these incentives. And it's something that the the first Trump administration deserves enormous, enormous credit for, but over which it is, it's it's run away from since as fast as it possibly could because people didn't like the idea of maybe being encouraged or forced to take vaccines, vaccine mandates that it ran into the populist counterattacks.

And as a result, the administration now finds itself criticizing itself the first time around for what was an enormous success over which, for which it deserves just enormous credit, but which it hasn't gotten and doesn't want for itself.

**Danny**: If we trace that story to the end, it has a slightly less happy ending as it were because one thing is producing the vaccines. That was clearly, critically important. And then the second step was distributing them. And during the vaccine rollout, California and also other states had the doses, but no one knew where they were. It was incredibly difficult to find out who had them, which pharmacy was actually able to distribute them on any given day.

And there was a privately led effort by Patrick McKenzie and a handful of volunteers that built an organization called vaccinate CA, vaccinate California, to fill the gap. And, essentially, what they did was they built a website that would tell you where you could get vaccinated on that day. And because no one had that data, the way they procured the data was initially they called pharmacies. So they had a bunch of volunteers who would ring work the phones, call the pharmacy, and ask, do have you any doses today? And they would say, yep.

And then someone would enter that in the website. How come no one thought that this was part of their job? No one in the entire public health establishment either thought it was their job or was able to pull it off, given that the entire success of warp speed is utterly meaningless. It's an o ring problem. Unless you complete every single step in the chain, nothing is achieved.

Unless you actually manage to distribute the vaccine, everything upstream is meaningless.

**Don**: Yeah. That's a great point, and it makes a couple points. One is that you can have experts who are terrific at producing great ideas, but not necessarily in figuring out how to deliver them. And so there really was this problem at the beginning. There's still a sense of panic over COVID itself, a sense that the answer was at hand somewhere, but we just can't find it.

And so there was, especially among the the people who who were most eager to get vaccinated, there was this horrible problem of trying to figure out where to find it. And the next stage was the, was the realization, you know, that the way that we could get it out there was through private pharmacies, that there, everybody, everybody's got a pharmacy. Everybody's got in The US a CVS or a Walgreens or something else somewhere. And sometime, some places have a CVS on one corner and a Walgreens opposite it. And so that there were, there were pharmacies, you could train pharmacists to administer the vaccines, which is, was typically not part of the job description, but which wasn't all that difficult to be able to do.

And then you could, because people knew about pharmacies, went there all the time, knew and had confidence in them too, on top of that, they often knew their pharmacist, That that was a, an important critical link in making the bridge between the government funding of the development of the vaccine and private development of the vaccine itself to enormous production, and then actually getting into the arms of individuals. And it was that link of using private sector partners to be able to deliver what essentially amounted to a public good through an innovation and delivery that was built around systems of trust, he says. So talking like a wonky political scientist. But that's an important issue. I mean, how no matter how fast you move, can you connect with people?

And the answer was, yes, you can. And then the next stage was, how do you do it in a way that people's gonna, people are gonna trust it? And the answer is you use systems that they already trust in pharmacies. And then at the early stages, well, that's good, but I go to my pharmacist and they don't have it. And so then you develop an information system that allows people to be able to find it.

And you not only do it, but you notice that it was, the way that system developed and the way it rolled out in other places in the country was that it was based on connecting with private pharmacists in a way that deployed the information that was place based on a picture, that is a graph, a map. And so you could, if you wanted to find, I was sitting here in Austin and you wanted to know where could I get a vaccine? And turns out, well, my local pharmacist didn't have it, but, oh, wait a minute, I can, 10 miles away, if I wanted to drive there, they have not only vaccines, but I could pick which manufacturer vaccine I wanted. It was all on a map. So the idea of thinking about the chain of connecting lots of stuff with the people involved creating a delivery system based on trust that was nimble and flexible, based on information with the information, displayed in a way that was place based on a map.

And, you know, golly gee, that sounds like maybe what AI could do in the future, that you could create cross sector information systems that are meaningful to people by using easy to interpret pictures that display very, very complex bits of information. That's kind of cool. And that is, I think, an important lesson from all of this. So you're trying to figure out where should we go? Where can we go?

The answer is, I think, following that chain, which provides an enormous amount of clues about what worked. And the important thing to recognize is that it wasn't just a, well, let's see if this works. It they did check to see if it worked, and in the end they ended up with vaccination rates of, if I remember correctly, something like seventy percent or thereabouts of people who were able to be able to get it through the system that we described.

**Danny**: If we stay on the the question of technology, technological change has always brought about change in how government operates. It's an old story. Railroads made modern bureaucratic states possible in the first place by reducing monitoring costs. Harold Innis, the Canadian scholar, had this fun theory that the move from stone to papyrus made the ancient Egyptian administration much more effective and expanded its power because you could sort of take the papyrus and then run with it as it were, which was much harder, with a stone tablet. The printing press obviously enabled us to produce standardized forms.

Hard to imagine Without that, you know, we could go on. What would you expect to see with AI systems in the future?

**Don**: Before we get to AI is just the the comfort level of individuals dealing with technology to be able to deal with the systems that they want, and in some ways that are self-service. We have here in Texas a requirement that you get your car inspected for emissions, at least in the urbanized areas every year, and that you need to make sure that you have insurance, and that in addition to that, you need to display a little registration sticker on your car. And once upon a time, that was a really complicated process because each of those things had to be done individually and had to then require maybe a trip to a large bureaucracy that where you had to wait in line for a long time. I've yesterday just got my sticker for next year, and I, after I got my car checked for emissions, which it passed fortunately, I would and I had my insurance, which the people who checked for the emissions checked for me, they registered that into a comp a data system. I went home, got out my credit card, because they charge you for this of course, to register your car because it's a way of making money.

But I was able to tick, tick, tick, and enter all the information in. And then week and a half later, I got the sticker in the mail. And so we've short circuited a lot of what had been the, the bane of the existence of most Americans of having to go to the department of motor vehicles and wait endlessly in line. And all of a sudden the job is done electronically. I changed my I was able to update my driver's license all electronically, and I got a handy new shiny driver's license in the mail as well.

And so if you think about that, that provides a way of thinking about how individuals can interact with government in a way that's far less painful, and that provides better, faster service in ways that people are likely to trust. And you rely on the postal system, you rely on the internet, and you rely on information systems that are set up by government, as well as private partners who do the the inspections. Because I went to a private garage that that inspected for me. So you think of all the different players there, and that's the and just mapped it out. We have the Postal Service, a kind of quasi governmental organization.

We have the inspection, which was a private company that then checked with another private company about whether or not I had insurance that entered it all in a data system, undoubtedly provided by a private contractor, managed by the Department of Motor Vehicles, and then integrated all this stuff and all of a sudden it pops up in the mail. So you think about the way in which those systems work is a way to try to figure out, a way to, to solve these problems. Now, what, where, how can I figure out where to go to get my car inspected and checked for emissions? And so I went to this website for the, for the department of motor vehicles, and they have a way to be able to check addresses and up pops a map where I could be able to see who does it within a short drive of where it is that I lived. And there was one place that I checked and it was really close, but I knew the lines were likely to be long.

And there's another place that was a little bit further away. And I said, ah, you know, I could go there. Turned out to be, I was out of there in fifteen minutes and I was able to find that because it was a place based information system that drove the the inspections that were being put together. More of that, I think, with AI weaving the pieces together, being able to allow individual citizens to engage in self-service at a time of their of their convenience, in a way that works either on their phones or their tablets or their computers, and to be able to work through an integrated system that for them is seamless. And that I think is the way in which we're likely to go.

Place based systems created and constructed and integrated through information systems involving public and private service providers, where the idea of, of figuring out how individuals can connect with government as easily and painlessly as possible, regardless of sector, with information as the driver is I think an important way that we can go. It of course doesn't fit everything, but for a lot of things that government does, it it works, I think, pretty well. The problem, of course, is that you have to be have access to and be familiar with the internet. And there was a time not too long ago where there were you had a hard time trying to reach older people who didn't have a clue about how to try to do this. But now we're we're getting to the point where more and more people who were in their sixties and seventies and eighties, have had smartphones for a long time.

So we've crossed the threshold for being able to make this happen. So that's, that's an important thing. It also creates problems potentially for those who are, who don't have access to smartphones or to computers or the internet. We to be careful about, about inequalities and the delivery of government services that pop up because of income. And there are ways of being able to provide that in libraries, for example, where there's access to the internet for everybody.

But that's, that's something we need to need to think about. But it does, in terms of thinking about the delivery of government services for the future, have enormous potential. You can also leapfrog up to the highest levels of government. There really are problems of waste, fraud, and abuse in social security, for example, and in Medicare and in Medicaid, our big healthcare systems funded by the government. The reason is that most of those systems are operated through private contractors.

One of my favorite statistics about the federal government is that Medicare and Medicaid and related, child's health program constitute about 25% of the entire federal budget. The number of federal employees in charge of managing that are 6,000. So we've got, I'm sitting here just a few miles away from the University of Texas at Austin, and they've got 25,000 people in charge of managing 50,000 students, faculty, staff, and all the rest. So we've got six times as many, I'm sorry, almost nine times as many employees at the University of Texas for 50,000 compared to 140,000,000. And so what we have is a system that operates through private contractors.

So the way to be able to get to waste, fraud and abuse problems is to, what, to use data systems that allow you to track what's happening and who's doing what, and using AI to track down the likely sources of and so that's the that's the the the new frontier on issues that people really care about.

**Danny**: When it comes to outsourcing, which is something you've thought a lot about, is it ultimately as simple as it works really well when outcomes are measurable and cannot be gained, and it works significantly less well when the outcomes are not measurable and or can be gained? Or to make a specific example, the reason why NASA was able to outsource the development of some rockets to the private sector was because you can unambiguously know whether a rocket has achieved its objective or not. It's very simple to know. It's very simple to specify what you want from that rocket. The same thing, of course, is not true for many other things.

So is it really this simple in the end? Is it actually much more complicated?

**Don**: At least much more interesting. Let me put it that way. And the, the case of rockets on the one hand is, is fascinating because NASA decided that it was going to get out of the business of, of specifying particular outcomes, except for we want to be able to launch people into space and bring them back safely, but we're going to allow private companies to be able to compete for the business given that broad piece, instead of specifying the individual components as NASA used to do, it, was buying outcomes. And the problem is that it and what it is, is it encouraged multiple organizations to compete for NASA's business because it did not wanna be in the business of being captured by one supplier. Well, it turns out that that one of the companies, Boeing, has had a terrible problem trying to get anything to work well.

Elon Musk's company, SpaceX, on the other hand, has tended to to be far more effective and it has captured an enormous chunk of the business. And so there's not as much competition as the model suggested at the beginning. And Musk lately has had a problem of of of launching things that blow up. And so, well, we know that didn't work. But on the other hand, they've got, they say, a strategy of launching things and launching lots of things often, allowing them to blow up and learning quickly.

And so that's a new kind of model where NASA's sitting there keeping its fingers crossed that having found that it's got not all, but most of its eggs in one basket where the basket blows up, that the strategy of of of rapid learning will actually work for it. So it's it's a different kind of strategy with that. On the other hand, you go down to the grassroots and a large part of the social service system in this country is managed by nonprofit organizations that operate through federal money and state money and local money. And so you've got often very fuzzy outcomes that are difficult to try to nail down. And we have a problem of trying to figure out whether or not these programs work.

We have an administration that's come in and said that this is just a bunch of liberal left leaning progressives who are trying to fund their favorite kinds of programs and strategies, and it's not really based on anything that really works because we can't really measure the results. And so the fact that on the one hand, if things blow up, that's, that's not good, but at least you know that it's blown up and you know whether or not it works. On the other hand, for social services, it's very hard to be able to tell what the results and the outcomes are, and therefore it makes them more vulnerable to political attack. So you have your choice on the one hand of of highly technical, easy relatively easy to specify outcomes that turn out to be pretty hard to pull off. And on the other hand, you have social service programs that are pretty hard to pull off, but where the outcomes are are less fuzzy and where, therefore, the the problem of building political support is much tougher.

**Danny**: Personally, what did you learn from Paul Volcker?

**Don**: That's a great question. And I had a chance to be able to to get to know him, but also that built on a on work I did with for a book, The Leadership at the Fed, where I had a chance to be able to look at the the history of the Federal Reserve leading up through Volker and his efforts to to try to stabilize the economy. And the first thing I learned was you look at his work to stabilize the economy by putting the economy through the ringer with super high interest rates, which was needed to try to to beat down the the inflation that had grown up during the 1970s. And so it's not only as we look at the, the economy and some of the battles over the federal reserve now, you better be pretty careful about which you let loose, because if you screw up, the consequences down the road can be monumental. And so sometimes it's more important to stick to the pain in exchange for the long term gain.

And there's it helps to be able to look over the hill and through the problems to be able to do that. Second thing is that, he was he was the master of communication when he was at the Federal Reserve. And there he would sit there and he'd be chomping on a cigar and he'd be testifying and there'd be smoke coming up and people were trying to figure out after he got done talking, what did he just say? And he had this way of being able to talk to to be able to allow people to somehow divine his his words without him being specific about it. And sometimes it was like the ancient Romans cutting open a chicken to check the entrails to see what it was that the fed was about to do.

Because he knew that that people were not only just betting on the fed's actions, but betting on the expectations about the fed's actions. So it's important about not only managing what you're doing, but managing expectations about what you're doing. Another really important thing. It's useful to be, I think it was six six or six seven, and I went to lunch with him one time and we were sort of walking down the sidewalk and having a conversation, and I got a cricket in my neck just having to look up at him. And I was five'eleven, but I was just nothing by comparison to his sort of towering presence.

There was something to that as well. But most importantly, here's a guy who spent his entire career in central banking. He was president of the Federal Reserve Bank of New York, which was the single most powerful Federal Reserve Bank in the entire system. He was chairman of the Federal Reserve and was responsible for probably the the single biggest economic accomplishment of the last century in in stabilizing the economy in the after what had happened in the seventies. And after all this, instead of turning to consulting, which would have made him a fortune, he focused on the public service because the lesson he had learned through all of this was the importance of, not only the policies, but having smart people with a public service mindset in place to be able to administer what had to be done.

And it was a powerful, powerful driver for him. And the reason I was able to finally figure out, because I'm trying to figure why Mr. B, did you do this given what you could have done? And one thing he told me was a story. When he was little, his dad was heavily involved with, with public service kinds of things.

He's a city manager of a small town in New Jersey, and they were interested in trying to get some insights from the gurus who were at the Institute of Public Administration based in New York. And at that point, the the head of the institute was Luther Gulick, who was one of the real giants of public administration in the middle part of the twentieth century. And so he remembers having gone with his dad to visit Luther Gulick at the Institute for Public Administration to talk about city management. And the lesson that he drew from that was the importance of a kind of intellectual leverage over complex systems, coupled with the importance of the public service that drove his career for the rest of his life. And that story has really stuck with me because from the time he was small, he had in his mind the importance of of serving the public, of doing it through public service, making sure that there were others who could come along with him on the journey, and to be prepared for the the really tough battles that lay ahead because he knew that there was a core of values that the field could teach and that he could take with him in the way in which he did his job.

And he could as as he told that story, which I don't think he told very often, but he was talking to me about that and the as one of these the scales fell from my eyes as I was listening to him because it was clear where he had gotten it, where it had come from, and what drove him and the work that he did.

**Danny**: That's a great story. We've talked a lot about the US government for good reason. The US is also in a long term competition arguably with China, and China seems to have quite a lot of what you could call state capacity. What should we learn from the Chinese way of doing

**Don**: The first thing is that they have invested a lot in the state capacity, and they've they understood and still understand the importance of creating a system that allows them to, to move fast about what it is that they want to do. That's important. One, one small thing, two small points of history. One is that the Chinese in fact invented bureaucracy, invented the public service thousands of years ago. And they had, in fact, a testing system for civil servants that was incredibly complex.

And so there's a temptation to cheat. And if you've cheated on your civil service exam, however, the the penalty was death. And so only one to two percent of people passed, and you didn't wanna try to cheat to make sure that you were among that. There's always high prestige that came along with the civil service in China for thousands of years. The second thing is a kind of reverence for knowledge and for expertise itself, a real level of respect.

The something that I wish that they could bring to The US there in about the year January or so, professors were held in such high esteem that they had sedan chairs with people carrying them around from meeting to meeting. And so I've, I've often thought that would be, that'd be nice to be able to have, but I don't think that's gonna But again, it's the sort of the reverence for knowledge. I was struck at during when I visited China about the the kind of level of respect that I was shown, unlike the situation in The US, and not because of anything having to do with me, but because of the position that I occupied. So there is at the core, this appreciation for for knowledge and for capacity that lies at the core with an understanding that that's that's how governments, effective governments work. Related to that is something that we don't often pay a lot of attention to, but there was China's a big country, And unlike The US where there is a system of federalism where we have a kind of understanding about self government, one of China's big problems has always been trying to figure out how decisions in Beijing get transmitted out to the to the provinces.

And long, long, long, long ago, they've they've figured out the importance of of devolution and how to try to to manage that from the center. And of course they they now have a a communist apparatus that allows them to to do that. But it's based on much more than we have decided this is what you must do kind of approach. There's a real, a very sophisticated system of devolution goes along with that. And you've got a country as large as China, where the basics of service are important, you've got some importance there as well.

There's, I think in addition to that, the importance of investment in government and government services and technology, which you can see in both high speed trains and airports. I was struck the last time I was there getting on a plane in Beijing and then landing in LA and trying to figure out, well, which one is the third world country here? The airport in Beijing was just miraculous in comparison to, you know, LA works and the airport's huge and, and people don't seem to care a whole lot that it's not super fancy, but there's a big difference there in the kind of investment in technology. And so we we have that. But on the other hand, the risks of course that come along with a a country that has done all this, where we have both economic competition that's that's real and strategic problems that are, that are enormous, and now tensions coming through, through tariffs and other kinds of, of issues that really put on the table the, the risks of picking a fight.

So that there's that, that tension underlying it. But the, the importance of the respect for information or technical expertise for using that to try to, to manage relationships between the center and, and communities throughout the country, the investment in technology. Those are pretty interesting lessons that I think that the Chinese have, and there's a and it's more and more sophisticated by far than the impression that I think most Americans have, just saying, well, you've got an all powerful center, which, which they have, and they make decisions and everybody has to fall in line. It's a, I mean, it's a big country and it's, it's pretty easy to just sort of slide underneath the radar if you're, if you're doing stuff. And not to say that they're not in problems of of depriving citizens of access to some source of information.

But on the other hand, there's no, there's no secret that that Hollywood movie makers spend a lot of time figuring out how to appeal to to chinese audiences of the movies that they make because there's a as a large audience and there's a a a strong set of of both pride and and culture in China that they have to be able to tap into that's that's also part of the of the culture and part of the the way in which society works that that help to drive the nature of government's power and capacity.

**Danny**: What's something you used to believe about government that you've changed your mind on?

**Don**: Something that I used to believe that I've changed my mind on. In some ways, I've had one thing I that I haven't, deduct your question for a second, is, the importance of, of creating a government that really connects with people. That I started this business really focused on why it is we do what we do and how do we know that it works and understanding that that government isn't government unless it really connects with the people. On the other hand, something that I've changed my mind on, think, is the mindset that I had that I think a lot of the rest the field shares about role of of traditional structures and authority and hierarchy in figuring out how to deliver those services to citizens. Because I've seen through warp speed, I've seen through research that I've done.

I've, I've seen the way in which people connect with government, the importance of these, of these connections that the government increasingly has through the, the nonprofit world and through the for profit world. One of the things that I did was to, I served on a, on a task force once to advise the secretary of energy about nuclear waste storage, not what to do or where to put it, but how to develop trust and the decision that was being made. Because the storage of long term waste is important. You don't wanna mess with something that's gonna be lethal for ten thousand years or so. But on the other hand but on the other hand, technically it's a it's a solvable problem, but the problem is the political one of trying to figure out how you convince people that that having this 20 miles away isn't such a bad idea.

And so I, we toured the country, talked to lots of people, talked to contractors, talked to others about the problem of figuring out how to make that work. And so two stories still stick with me about that. One was that I visited the place where they used to make the the triggers, the plutonium triggers for nuclear weapons. So it was the the in some ways, one one of the most dangerous things on earth and got to the to the front gate. I I wasn't allowed to go anywhere without an escort, including even to the men's room.

They wanted to make sure that somebody was there all the time because didn't want to have any spies doing anything that they should do. So I had to wait at the gate for an escort to go to my next meeting. And the guy came out and, he, we were just talking and I said, oh, it must be a tough, tough job. He, this was a guy who looked at me and said, I don't wanna mess with him. Because he was, he was big.

He was hulking. You see the, the stories of, of North Korean guards at the border. They have nothing on this guy. And he said that he was trained in, in 20 different weapons. So I started my mind trying to begin, counting up how many weapons I knew.

And I got to about eight or nine, and I said, I don't, I don't wanna know about the rest of them. And on the then I looked at his uniform and there was a patch on the side that said, Wacken Hut, which is a very large private sector company that that had security tighter than at the White House, provided with over some of the most dangerous stuff on earth, provided by a private contractor. And I'm saying, this is, you know, talking about figuring out how to, how to deal with with issues, that gave me a flavor about the way in which things work. Then on a different trip, we went to Las Vegas and we had a public hearing. And one of the things that, Las Vegas was about 20 miles away from where they wanted to store a lot of the stuff deep inside a mountain.

And so there are a lot of people, as you can imagine, who are pretty worked up about all this stuff. And we had TV cameras there. There were people who were rabble rousing, we were being picketed on the outside, all having to do with problems of creating trust. But then there was a there's a woman who came up to the microphone to testify, and she said, my name is Cynthia of the desert. And given the issues that we were facing and someone who was obviously a naturalist, I was thinking, I better grab my seat belt because wow, I know what's coming.

I was dead wrong. She said, I live with my husband and my child in the desert, not far away from Yucca Mountain where they wanted to put the nuclear waste. And I was terrified about it and I was, I just really was strongly opposed. But then there were meetings out there with the department of energy staff who talked with me about what it was they were going to do, about how it was going to work. And I came away from those conversations really trusting what it was that they were saying.

And so I think that this is a good idea, and I think it's something that the country needs that I'm, I want to support it. And I was thinking, wow, here's somebody who you'd be sure in advance would be an enormous opponent of the, of the idea of putting nuclear waste anywhere, who started out in that position, who had gained trust in the process through conversations and interactions with government officials. And so things that I, where I've really changed my mind, or at least I've gotten more properly, probably a more sophisticated understanding about what happens is understanding the important role of the horizontal connections that make government work and the importance of, through those connections, building trust in citizens based on the nature of those interactions. And that's something that is really, that's sort of, that's sort of my Paul Volcker story about what matters, how to connect, the importance of trust, and the ways in which these networks operate to deliver government services. And it's, so it had to do with, again, not so much changing my mind, but rubbing shoulders with the the way in which government actually works that I think has has created a different sense, in my mind at least, about the way in which a public administration ought to ought to behave.

**Danny**: Final question. What is something I should have asked but didn't?

**Don**: I think we've explored a huge part of the world out there, but I think that not so much should have asked, but but the question we need maybe to focus on, because it's I still have kind of fuzzy in my own mind, but we really need, I think, now, to think about what government in 2035 ought to look like, how it's gonna operate. And I mentioned that earlier, but it really is not very far away. If you think about the number of presidencies that we have in The United States in the meantime, not many. If you think about how long it takes to develop, even if you're, if you're warp speeding everything, that took nine months and then another year or so to get things going. So you take two years and you, so you take 2025.

And so at that point you're at 2027 and maybe a new presidency is coming in, has to get his feet wet in 2028, 2029. At that point, 2035 is looking pretty close. And and the world in 2035 is certain to be different than what we have now. We can be sure that the the pressures of government are going to increase, that the, in some ways, the tensions between nations, but the importance of interconnections are going to grow. You can think about the interpenetration of government and the private sector, where I think we need to think about government's use of its private partners and private partners that increasingly have to think about, on their own part, a sense of citizenship, since they're gonna be heavily involved in delivering services.

Simply just delivering things on contract, that's not gonna be enough. And so we need to think about that and how that's gonna operate. I think it's clear, you look at the, the enormous pace of AI, it's clear that's gonna be a huge thing. The, I've, we, we can't today sit in the, toward the 2025 and even guess what AI is gonna look like in, in the middle of 2026, at the way things are going. I'm constantly stunned at what it is that we could do, but we, but we know it's, it's, it's stampeding ahead.

And so the question is, how we use information to make government work better and connect better with citizens? That's a big thing now that we can't solve, but which we need to put front of mind, I think. And then ultimately, not only how can we government work better for citizens, but ultimately, and this gets back to the Paul Volcker point, how can we hire the public servants that we need with a culture and the mindset that's gonna be required to be able to help the government navigate its way through. And for for a big democracy like The US, that I think is one of the one of the biggest challenges that we face. And it's pretty clear that that most of what we have now is not a very good answer to the question of what's gonna happen then.

So we need lots of smart people like you and and others who are sort of exploring the world, figuring out a road between where we sit now and where we know we're gonna need to be with intellectual capital supporting it all. The Chinese had that and have had that for thousands of years, and they're they've got that one approach that is developed here. What what's what's ours gonna be? And we have no choice but to figure out how to adapt, and we just have to make sure that we're not clumsy and stupid about the way we do it.

**Danny**: Don, thank you so much.

**Don**: Danny, it's been such a pleasure talking with you.

---