Martin Wolf is chief economics commentator at The Financial Times. His latest book is The Crisis of Democratic Capitalism, now out in paperback with a new chapter on Donald Trump’s second term.
In this week’s conversation, Yascha Mounk and Martin Wolf discuss whether democratic capitalism can recover from the damage done by populists like Donald Trump, how artificial intelligence might upend both the economy and humanity’s control over its own fate, and whether a bursting AI bubble or France’s fiscal woes could trigger the next financial crisis.
Note: This episode was recorded on November 3, before the election results were available.
This transcript has been condensed and lightly edited for clarity.
Yascha Mounk: It’s always a pleasure to talk to you, Martin. I’m really excited to talk about a bunch of the contemporary economic issues facing the United States, Europe, and the world today. But I want to start by revisiting The Crisis of Democratic Capitalism, your really magisterial book about the relationship between capitalism and democracy. It is now a few years old and has just come out in a paperback edition with a new chapter.
Looking back from today’s vantage point, do you remain optimistic that, for all of the challenges we face, we are capable of continuing to reconcile our capitalist system with our democratic institutions, and that we will be able to avert the threats that risk undermining each of those elements?
Martin Wolf: Overall, I’m a little less optimistic. Not dramatically less optimistic, but less optimistic. Very briefly, I feel two things. The first relates to a basic theme of my book. Quite apart from the benefits of moral development in some sense, education and so forth, the capitalist economy that we created in the 19th century and roughly the first three quarters of the 20th century quite naturally generated a redistribution of power in society.
We moved, as it were, from feudalism and pre-modern structures based around land ownership to the modern capitalist economy, with an increasingly organized and powerful working class. Workers were genuinely powerful because they had education and they could easily coalesce and coordinate. That was an environment in which there was a very real risk of major war, and we saw that linked with the formation of the nation state. But there was also tremendous democratic pressure.
In the end, we seem to have reconciled all of that in the post-Second World War period, which was really very important. That underpinning has very obviously been cracking, and that was a theme of my book. But when I wrote it, I still thought that if we understood how damaging and how dangerous this was for living tolerably, we would prevent the worst.
That leads to the second theme, which is very much the theme of this revision. It contains an extra chapter on the Trump experience, Trump’s second term and so forth. What has become obvious is that we do now elect really quite radically reactionary governments. Maybe that’s not the right word, because they’re more complex than that: radically demagogic, populist governments. And we seem likely to do so now in other countries in Europe, which also seemed to be stable. The more I think about it, the more difficult it seems to undo that, because of the damage that is done.
You start not with something that is mildly damaged but with something that is, in some respects, wrecked. Finally, there are changes in technology and society, of which AI is obviously the most important and the most recent. But we’ve had others before that, social media and all the rest of it, which make it so difficult for people to organize politically in the way they used to. The parties of the left and right are just different institutions from what they were when I was young.
So I now begin to think that reversing all this and going back to the things we were familiar with is just a pipe dream. What that means for the future is really open, but in this respect I’ve become more concerned. We’ll see about the Trump experience in the next few years, perhaps. But undoing the damage he has done domestically and internationally—and the U.S. is so pivotal—is going to be an immense task.
One very last comment on this. The Polish experience, which I discuss, is actually live right now as we talk because of the argument with the president. It shows just how difficult it is to reverse.
Mounk: I’m struck by how much more negative and pessimistic the tenor of this added chapter is compared to the overall tenor of your book. The book was obviously written in alarm and with concern, but it still seemed to have a certain baseline confidence that, in the end, the enormous advantages and achievements of democratic capitalism would allow it to persist.
Perhaps we’ve gone on slightly different trajectories over the course of the last 10 years. Perhaps I was much more alarmed in 2017 than you were, and perhaps I’m a little less alarmed than you are today. Your book came out, if I remember correctly, before Donald Trump was reelected.
Wolf: Yeah, it came out in early ‘23.
Mounk: So obviously the fact that Trump was able to get reelected was very concerning and upsetting. But look back at January 2025, when Trump was taking office. He seemed to have a much more efficient staff than in 2016. He had control of a much more unified Republican Party. He seemed to be implementing Project 2025 and other plans for remaking the state very quickly. The Department of Government Efficiency was trying to dismantle large parts of the federal bureaucracy.
He was talking about taking possession of the Panama Canal and of Greenland. He wanted to introduce sky-high tariffs across the board. He had total control of the government. That felt like a very perilous moment. Internationally, it was a moment in which Narendra Modi was riding high in India, and in which there was good reason to believe that democracy had basically been abolished in Hungary by Viktor Orbán, in Serbia by Vučić, and in other places.
A couple of years on, it feels to me that Trump is now finally fading from the American political system. Many of my colleagues have written this once a week for the last 10 years, but I only wrote it for the first time a few months ago. He is at record-low approval ratings. We happen to be recording this on the day of the midterms, and it is very likely that Democrats will win the House of Representatives, and probably the Senate.
Most of the governorships are going to go to Democrats, so it’s going to be very hard for Republicans to skew the vote in 2028, because the vote will be administered by and large by Democratic governors and secretaries of state. More broadly, Trump has done a lot of damage. He has been enormously corrupt and has shot his own country in the foot. But the Supreme Court has struck down a lot of the tariffs. The U.S. economy continues to be surprisingly robust. The attempt to dismantle the federal bureaucracy quickly stalled. They have effectively managed to abolish USAID, but beyond that there haven’t been changes as big as were expected.
Internationally, Viktor Orbán has gone as a result of free and fair democratic elections. Even Vučić appears to be stepping aside. Modi looks more vulnerable than he did recently because of the cockroach movement.
So perhaps we need to move away from the idea that there’s a functioning set of democracies facing threats from populism outside, and either we win or we lose. Instead, we’re going to be in a period of much lower-quality governance, with many more irrational policies and bad governments that do really damaging things. But perhaps the threat to democracy and to the capitalist system is a little less existential than I might have thought five or 10 years ago. Am I really naive to try and see this optimistically, in this vein?
Wolf: I don’t think you are, because the truth is we don’t know. I’ve tried to think hard about historical analogies, and there’s never one that’s perfect.
So I’m beginning to think that maybe the best model for how politics might work in many countries is South America. They have dictatorial regimes, but those regimes don’t last very long. At some point they’re cajoled into elections, partly because of outside pressure, though a lot of the pressure has come from within. Then there is a democratic government, and there’s some respect for civil rights. But the fundamental problems aren’t resolved.
At one stage I called this a populist cycle, but perhaps you could call it a populist-democratic cycle. It goes on. Government is never really efficient, because institutions don’t last long enough. They’re not stable enough. The people in them aren’t secure enough or valued enough to be really very good. But on the other hand, they’re not catastrophic either.
You go on like this. When I look at the government of my own country since the financial crisis, it increasingly feels like that. We’ve had a succession of governments, none of which is really in control. They’re not really effective. All over Europe, I think it’s similar. They will quite possibly be replaced at some point by some sort of populist government of the right or left, but that won’t end democracy. It’s just dysfunctional. That’s much worse than what we had before, but it’s not a properly functioning autocracy.
In many of these countries—India is another example—you won’t end up with a Putin-type regime, because that’s very specific. But you will end up with regimes that oscillate between attempts to create a properly functioning, old-fashioned democracy and, on the other hand, decidedly corrupt, ineffective, but still malevolent populist dictatorships, which could be of the right or the left.
We hope you’re enjoying the podcast! If you’re a paying subscriber, you can set up the premium feed on your favorite podcast app at writing.yaschamounk.com/listen. This will give you ad-free access to the full conversation, plus all full episodes and bonus episodes we have in the works! If you aren’t, you can set up the free, limited version of the feed—or, better still, support the podcast by becoming a subscriber today!
If you have any questions or issues setting up the full podcast feed on a third-party app, please email [email protected]
That now seems to me quite a plausible future. In that respect, the parallels I once saw with the 1930s don’t seem right anymore. But at the same time, I don’t feel that the stable, self-confident democratic systems of my youth will survive, except perhaps in countries like Switzerland. Maybe the Netherlands, though that’s not so clear. Scandinavia, I think—small, very solidly bourgeois countries. But in the rest of the world, it doesn’t look like that. You could think of this as optimistic or pessimistic, but it’s not what I would prefer.
Mounk: I think we’re converging on a shared interpretation of a glass of water that is a quarter full or three quarters empty. One way of thinking about this from the perspective of political science is through regime types. I’ve been saying for a very long time that it’s not helpful to think about democracy and dictatorship as ones or zeros.
Wolf: Yes, of course. That was the Cold War scenario.
Mounk: Yes, but no system is perfectly autocratic and no system is perfectly democratic. There’s a huge range of gradations between them. One helpful political science concept is that of competitive authoritarianism, or at least of an uneven playing field, where somebody like Orbán or Vučić is able to capture power. They make the playing field uneven, and they favor themselves in all kinds of ways. That makes it hard for the opposition to displace them. But there is a genuine competition between the government and the opposition, and when the government becomes sufficiently unpopular, the opposition might be able to come in. And perhaps the opposition is great and democratic for a few years. But perhaps they themselves start to emulate the same norms and govern in abusive ways. And at some point a party that really doesn’t care about democratic norms is going to be back in power.
Another way of thinking about it is perhaps as what my colleague Roberto Foa and I once called a “dirty democracy.” Instead of the rules being set for everybody and everybody competing to get into government by the existing rules, a lot of the game consists in influencing, manipulating, and exploiting the rules of the game in such a way that you’re more likely to stay in.
That’s still a game, right? You’re not able to capture the rules to such an extent that the opposition can never win. But a lot of the effort in this polity is then taken up not with putting forward the better ideas for governance, not even with a battle of values, but with trying to maneuver in complicated ways to shape the rules of the game in your favor, to make it more likely that you then win the next election.
Wolf: But there is an important point, at least I as an economist would insist on. What you’re describing is a massively costly regime type, particularly for any functioning economy. So what makes a functioning economy work? People have to make really long-term, risky investments. And by long-term, I mean decades.
If they are to do that, they need to feel that the rules of the economic game, by which I mean the rule of law, their ownership of property, and the ability to trust contracts, are secure. They need to be able to assume that the government services they will need will be provided reliably, consistently, and without corruption, and that the trade regime and the monetary regime they’re operating under are predictable. And if those are changed, they’re changed as a result of a legislative process that involves serious debates and discussions.
It’s a slightly naive way of defining it, but this is pretty close to what was created in the West after the Second World War. And it launched a period of just staggering economic advance. I was lecturing about this yesterday to a group of Chinese business people, and I said that the world order created after the Second World War is incomparably the most successful in world history. And it was because of this, and because of the opportunities that it also extended to people in developing countries, including China.
Now, the world you describe—the world of South America, which, by the way, has done particularly badly in this period relative to what it should have done—is very, very far from that. It’s a world in which you constantly have autocrats and their coteries coming along and acting in whimsical, willful, and corrupt ways. Whether you can go ahead with your project will depend to some extent on whether you can seduce or corrupt some bunch of bureaucrats to help you. Of course, that means you will use them to help you against your rivals, whom they will have to act against.
So everything is politicized, as opposed to everything happening in accordance with well-established principles. The support for science, technology, free inquiry, and genuine free debate will again be politicized. And societies like this—indeed, this is mostly what we have in the world now—in my view don’t perform very well.
I agree with you that the U.S. is doing reasonably well economically, but I would say that’s because it is pursuing a radically unsustainable fiscal policy in the long run, which is, of course, generating demand in the short run. And there’s the AI boom, but the AI boom is the fruit of decades of technological development that happened to come to this crescendo right now. It’s got nothing to do with what this administration is doing.
So I would say you’re absolutely right: there are far worse things than this. It’s far better to have this than Stalinism or Hitler, obviously. On the other hand, it’s going to be pretty unpleasant, and everything will be, as it were, for sale. Everything will be unpredictable in every possible dimension.
And this comes back to the point we were discussing. I think the damage from the Trump experience has been bigger than you suggest. Telling people “we’re going to go back to the way it was before,” when they’ve lived in these circumstances I’ve just described—well, they’re not going to believe it. They’re not going to behave as they used to, because they’ll ask: “Can you promise us that there won’t be someone like Trump, or worse, soon? Can you promise us that the bureaucrats I’m going to be dealing with will be honest, honorable, non-corrupt people, which they used to be?” And I think the answer to that is no, you can’t.
So I think you’re absolutely right that there are far worse things. And it’s very encouraging that these would-be autocrats have had their wings clipped, and I assume that’s going to happen now in the midterms. But it’s very far from what we had before, when we could be confident that the system would work.
Mounk: Look, I think that’s right. Living in Argentina is a hell of a lot better than living in North Korea, or for that matter in Russia. But at the same time, while a hundred years ago it famously might not have been clear whether you’d rather emigrate to Argentina or to the United States, today, if you have a choice between living in the United States, with its economic opportunities, or in Argentina, it’s relatively obvious that you would choose the United States, because the country has done a lot better. And as you’re pointing out, that’s not because of the governance we’ve had for the last 18 months—or the last 24 months—it’s because of the governance we’ve had for the last 60 or 200 years. So I think we’re on the same page about that.
One of the ironies I find about this moment is that the world is being radically and fundamentally transformed by artificial intelligence, and we’re going to have to think very hard about the kind of government response we need. That’s partially about regulation, to make sure that people can’t steal financial information and bankrupt you through cyberattacks. It’s partially about making sure that people can’t manufacture viruses that are going to kill millions, or tens or hundreds of millions, of people. But it’s also about thinking about how to preserve and reinstitute the basic structural preconditions for a middle-class economy and for inclusive political institutions.
Now, 20 or 25 years ago, it feels to me like a lot of the major democracies in the world might have had the self-confidence, and perhaps the lack of polarization, to go about that in a rational way. Perhaps that was never really the case, but at least there would have been grand commissions by think tanks involving a bipartisan set of people that would have done briefings for top officials. There would at least have been the expectation that there would be some rational analysis of the problem, a set of proposals being debated and discussed, and policy made on the basis of that.
It’s very hard to imagine that world today, right? And so we’re confronting this huge challenge of artificial intelligence at a time when our collective self-confidence in being able to govern it is much lower than it would have been 25 or 50 years ago.
Wolf: Yes, I think that’s clearly right. I’ve suggested this in a few of my writings on AI. The first column I wrote on this, and I still think the best, started off with the history of the horse, which is quite well documented. I pointed out that up until the end of the 19th century, the horse was really—we could argue a bit about the sailing ship—the most important transport technology in the world. And now horses are just for entertainment.
So it seemed to me that there’s a lesson. Of course, my economist friends say, “Well, of course, we’re much more complicated than the horse.” Yes. But if it turns out that the most important thing human beings can do is to think, which I think is true—and that doesn’t just mean thinking intellectually, it means thinking in everything we do—and we can create AI and robots (I was just talking to somebody who does robots yesterday), then it doesn’t seem impossible to imagine a world in which everybody who has any money would rather have, apart from their few loved ones, machines around them. It’s an interesting question what having money will even mean in the future, but let’s leave that aside.
So what is the rest of humanity for? I sometimes begin to wonder whether the really deep reason why nobody’s having any children anymore is that, at some deep psychological level, they’ve realized that there is no future for their children, so they are acting preemptively. I’m not suggesting that’s true, but it does seem fascinating that there is this coincidence in time.
But the fundamental point you make is obviously right. This is a revolution. Now, here it interacts with politics in the following way. The most striking feature of the politics of the last 30 years—and it was essential in my book, though lots of others have written about it—is the rejection by the body politic of “experts,” of intellectuals, who have come to be seen and portrayed as the enemy.
I understand why, because the sort of policies the intellectuals and experts recommended, very much including economists, don’t seem to have worked from the point of view of most people. And so that opened room for people to get into power essentially by running against them. The most striking feature, in some sense, of the Trump revolution, or attempted overthrow of the old order, or whatever it is, is the war against universities. It’s a war that was possible because the intelligentsia here discredited itself. I think it did so.
Now, 40 years ago, the idea that any government in America would plausibly run against Harvard, or here against Oxford and Cambridge, would have been unthinkable. So that comes back to our thinking about AI. The process you are discussing should have involved ordinary people. I very much believe in citizens’ assemblies; it’s one of my democratic romanticisms. But of course, in discussing AI we’d need very high-level expertise, and indeed a royal commission would be the obvious thing to do if you were Britain or the equivalent. In the same way—though I know all the differences—when in the ‘50s and ‘60s they were thinking about how to control nuclear weapons, they talked to people like Thomas Schelling, very famously. And I think if we’d really been asked, we could have thought at least moderately intelligently about the real risks.
The idea—I’ve got this in my column for tomorrow—that Anthropic would put out in its IPO, for a business that seems likely to be valued in the trillion-dollar range, that, well, one of the risks is that we might end humanity, as it were, en passant Anthropic IPO risk disclosure. It’s a reductio ad absurdum of learned helplessness. I do think the reason we can’t do it is that, somehow or other, the intelligentsia has discredited itself, and it has been successfully discredited by its enemies. Running a really sophisticated, complex modern polity embedded in this complicated world, while being against the intelligentsia—any form of it—just seems crazy. It’s a theme of one of the things in my book.
Compare it with what the New Deal was like. I wrote a bit about Dean Acheson recently. And then you look at what the Trump administration has done to the State Department or the intelligence services. And then you look at what they’ve been saying about the Iran war. It’s all crazy stuff. Admittedly, it was pretty crazy in Vietnam too.
But the point is, I don’t think we can recover that confidence in our ability to control these things. So they are uncontrolled and uncontrollable, and the Anthropic IPO shows that.
Mounk: If they’re likely to be uncontrolled, what do you think the economic impacts of this AI revolution are going to be? On the one hand, if intelligence was the ingredient of the economy that was scarce, then suddenly having a country of geniuses in your laptop, or whatever the phrase is, should unleash huge amounts of productivity and that might produce more goods and services in ways that make us all wealthier.
At the same time, there’s deep concerns about the distribution of those goods and services, and there’s deep concerns about whether or not human workers will still be needed in the ways they are today, and what that might do to the size of the middle class, which is something that I’ve been thinking and worrying about. Stepping back, to what extent can we predict what’s going to happen? How should we go about thinking about what this totally uncharted territory might look like?
Wolf: I think all those questions are the right questions. But I actually begin to think they’re not the most important questions. I may be completely wrong, but perhaps because they’re so predictably economic sort of questions, I’ve been wondering whether we’re not being radical enough in our thinking about what this means.
First of all, we live in a world in which literally everything is run through the digital economy. I mean, just everything. It seems to be absolutely inevitable that AI-enabled computing, which will be everything, will take this over. They’re gonna run it. We’ve already seen pretty clear signs—to which, of course, Dario Amodei has been referring—that they’re not necessarily gonna run it in ways that we would think are sensible or right or moral or anything. They seem to be extraordinarily willful and erratic and out of control. So in some sense, they can do anything, whether on their own or when guided by hackers.
In Europe just today, in Sweden, I think it is, we’ve got a story about some hackers—might be an AI bot, might be some hackers—that have got control of the freezers in supermarkets and have decided to raise the temperature to levels that rot all the food. These all look like terrorism.
What happens if that’s the normal state? My nightmare—and it just seems impossible—is that we wake up one morning and discover that every single bank account in the world has been nullified. Everyone’s. In which case, the world economy stops. Period. It’s done. Can we guarantee that won’t happen in this world? I don’t see how.
Mounk: I’ve been thinking about whether I should take printouts of my bank accounts regularly, so that in such a world you at least have some kind of proof.
Wolf: Yeah, I think you probably should. Every day, the businesses of the world—every single business—are doing trillions of dollars, collectively, in financing on a daily basis. They can’t do that without the system. The electric power systems of our world, everything else, can’t do that. So we are going to lose control of the core institutions, the core operating system of our world. I don’t think it is possible to guess what that will mean.
The second thing I was thinking about—we have this idea that at the end of this, we go back, as an extreme, and it’s suggested by your model, to a sort of feudal economy, but accept that the vast majority of human beings actually have no labor value, because machines will do all that stuff. We won’t even need serfs, as it were. Most human beings are completely redundant. The machines that run the economy will be owned by a tiny elite of super wealthy people—the likes of Elon Musk and so forth—and all the goodies will go to them.
I don’t see why that should happen, because why should these machines pay the slightest attention to these notional ownership claims of these fragile little human things? They have their own purposes and aims. Why should we assume—I’m being very, very science fiction—that… So I think everything you say is right, these are all good questions, but I think AI is more radical, potentially more destabilizing, less controllable than that.
But, of course, just one last thing: it’s perfectly possible to imagine an upheaval which will make the deindustrialization process, itself radically destabilizing, look like a garden party, because people won’t be doing what you or I do. Most physicists and others will be replaced by brilliant machines. How quickly that will happen, I have no idea. Maybe human originality will survive. But the fundamental question, of course, is: what value do humans have? An even more fundamental question is: do human beings continue to control things?
Mounk: No, and I think those questions about existential risk—which come in different forms—are absolutely fundamental. One way it could go is that a machine decides to eradicate humanity, either because it wants to take over or because it’s misaligned in some kind of way. Another is just that machines might have some kind of use for us, or perhaps they want to keep us around as a museum of their own prehistory, of their creation, or something like that. But they’re going to be the ones in control.
It’s very interesting to think about bullfighting in this respect. I went to a bullfight many years ago; I was kind of dragged there by somebody, I wasn’t particularly keen to go. But I did actually enjoy it greatly, because I started to realize at a certain point during the event that, when it’s done really, really well, it’s not about celebrating the person with the biggest muscles, or the person with the biggest cojones. It’s not just about muscles or bravery, though those things are involved.
It is principally about the triumph of the human spirit and of intelligence. It is a human standing in front of a beast that is greatly stronger and more powerful than any human being can be—greatly more powerful and stronger than Arnold Schwarzenegger or anybody else. But it is through our intelligence that we are able to master the bull. So that raises the question: why should we expect that, in the long run humans, even if we created these AI machines, which might give us some kind of advantage for a while, should continue to be able to govern these weird creatures that are already about as intelligent as us, and in many ways more intelligent, and probably one, three, five years from now will be vastly more intelligent than we are? Why should we have all of that confidence? I think that we should both be thinking about those questions and thinking about the question of how, if we do manage to control AI, it’s still going to transform our social and economic world in deep, deep problematic ways—because both of those are deep threats to our world. But I certainly take the existential risk points very seriously.
Wolf: Well, one thing that is clear is in the world that you described—assume away the existential stuff, just look at the economics—I think in the world we can imagine, in which AI is the dominant instrument of organizing production and so forth, if humanity is to continue in some sense collectively to be the master of its fate, collectively and individually, then there’s going to have to be a far greater degree of socialization of the means of production and distribution and exchange than any of us have thought about. Because it’s very easy to imagine that the marginal product of a great many human beings, compared to the most efficient alternatives, will be negative, or at least not really at what we would consider subsistence level. What are people for in this situation?
Now, there are clearly some things we would still want other people to do—I’m assuming here that there are still people in charge. But I think we’d have to start thinking that the assumptions we have about how the economy will work, how the output we are potentially able to produce will be distributed, and what you need to do to get the things you need to survive and lead a good life—whatever that means in this context—how that will be done, would be radically different from the ways we have organized them.
One way I thought about this is that essentially this could be a world, at the most extreme, in which everybody lives like the aristocrats of old. The aristocrats of old didn’t do anything, basically. They owned land, and they owned the fruit of the labor of the people on the land, and they had soldiers who were theirs who controlled the latter. Well, they could live without actually producing anything. They were rentiers. In the world of AI, if all the production is essentially done by AI, we all become rentiers, and a decent society would share the rents. So we would live not on our ability to produce, but on our right to consume. I’ve never actually written about this. I suggested this in a seminar about 15 years ago in Washington. You could imagine a world in which everybody became an aristocrat. If we’re rich enough, we could organize that world. Everybody could enjoy lots of stuff. They couldn’t enjoy everything—you can’t all have the most beautiful view in the world—but you could imagine that sort of world. Politically, though, it involves a complete transformation of the fundamental assumptions of our economic, social, and political order.
Mounk: Let me ask you about two things here. The first is that I find that analogy imprecise in two ways. The first is that many aristocrats had a lot of leisure, but they were actually responsible for something. In some ways, they were running, in our language today, small- or medium-sized enterprises. A profligate duke could spend money and not care about his fields falling into disarray, and then he would leave a relatively destitute heir. But their social role was to make sure that the heir would get about as much money and fields and so on as they did. So they were responsible for managing, essentially, hundreds of thousands or tens of thousands of employees. Now, they did that in a relatively arm’s-length way, and the people they employed to go around and look after that did most of the dirty work. But they were actually, in that sense, running little enterprises.
They had a sense of responsibility, and were judged on whether or not they were able to pass on to their heirs what they had received from their parents. The other point you hinted at is that, of course, the lives of aristocrats—insofar as they were happy, and if you read Brideshead Revisited and other kinds of works, there are good questions about the extent to which they were happy; many of them probably had quite empty lives and felt that emptiness—but insofar as they had meaning, they had meaning in part because they had high social status. They had a high positional good. They could go around and say, “I’m a duke.”
I have a whole social world of which I’m a part, and in which I play an essential role in the legitimization of this strange political order, because of my social position. I’m somebody. But not everybody can be somebody by virtue of who they are by birth.
Wolf: That is right. That is something we can’t arrange. We can only arrange that they consume. We can’t arrange to give them this exalted status. I agree with you absolutely, of course. A book that influenced me a lot in thinking about how the morality—the sense of purpose in life—of people like that might collapse, because it’s a brilliant book, is Les Liaisons Dangereuses, whose idea was, in a sense—here are these signs of the French aristocracy; of course, when it was written is important—who basically thought, particularly the male ones, that all they can do in life is destroy the lives of young women by seducing them and abandoning them. That’s clearly part of how they lived. So it can be very, very destructive.
I think I’m obviously one of those people—a classic modern person, perhaps pre-modern given where we are now—who thinks that the value in life consists of fruitful work and contributing to society and being part of it. I justify what I’ve done in my life as contributing in some way to political and economic life. That’s value, and I’m paid for it, and I feel very good about it.
In a world in which those functions wouldn’t exist, because they will be done better by an AI program, or because there’s no political system left to advise, then I don’t know what I would do. I do think that without some sense of contributing through one’s own efforts to society… You’re right, of course, dukes felt that too, in the way they did things. Certainly people who held serious positions in the state as a result of their rank. If you take that away from human beings, they basically never grow up. They’re never adults. They are perpetually looked after. That’s not a life. This is, of course, what Brave New World was about.
There are lots of major contributions to literature which play around with exactly this theme: what do you do if there’s no struggle, no effort, and no reward for effort, when you’re basically playing, and you know you’re playing? Well, I think it’s poison. I’m in complete—poison, which is why, as I said, if I were dictator of the world and invited to this commission, I would say this whole thing should be banned, closed, finished. I’ve said this many times. It’s what I believe. But of course it’s not gonna happen.
Mounk: I’m not sure I agree with you on that, for I am very conscious of the dangers of AI, but I actually want to ask a question that brings the conversation a little bit full circle, back to the concerns of your book. You rightly defend the triumph of democratic capitalism in your book, and explain the dangers that await us if that system breaks down. But let’s try and imagine the world fifty or a hundred years from now, if the AI revolution totally takes place. If it does turn out that a lot of the ingredients to productivity that used to be scarce, and that therefore needed to be distributed by markets, no longer are scarce in that way—we have much greater material plenty, nearly unlimited compute, and we have intelligence at our beck and call for negligible or nonexistent prices—most people can be idle, perhaps happily, perhaps unhappily, and live in material splendor where we find it hard to imagine, and yet we somehow keep control of the AI.
I know these are all very, very large “if” statements, and I think there’s good reason to think a bunch of them don’t turn out to be true. But let’s take that as a premise, as a question, for a moment. Do we still need capitalism in that world? Is capitalism still the right economic system, or will we have to envisage a totally different economic system to govern that kind of a world?
Wolf: The answer, in a way, would relate to where I would challenge the assumptions—and, implicitly, the assumption you’re making, and this, of course, goes back through a whole history, including the prophecies of Karl Marx about the future. It’s been picked up strangely in a very popular recent book, which I’m sure you’ve read, on abundance.
The idea of limitless abundance seems to me to be very difficult to imagine. The reason is there are things that are intrinsically scarce. As long as the things that are intrinsically scarce exist, they have to be allocated in some way. The market is only one way, and some of them probably don’t want to be allocated in that way.
Let’s think about what might not be abundant. First, in your world of limitless abundance, you’re assuming away resource constraints—that means we have limitless energy, we can recycle things freely, as it were, at zero cost, and the basic raw materials for production, metals and all the rest of it, will be freely available. But the most important of these things, unquestionably, is energy. So the world you imagine must, I think, have a limitless energy source, and it probably would have to be fusion. So there are quite big technological problems to solve, which may or may not be soluble.
Then the second set of things—if we don’t solve them, things will continue to be scarce, and we have to have a mechanism for allocating them. It’s either through markets, in some way, which is capitalism, or it’s through hierarchy of some kind. The hierarchies are political, and we have to discuss how those will be organized.
Then the other set of things, which was introduced into my thinking by a magnificent book published in the ’70s by a man called Fred Hirsch, which you might know, called The Social Limits to Growth—which is basically about what are called positional goods. There are a vast number of things that are valuable because they’re intrinsically scarce.
Mounk: Well, and that goes back to the idea that not everybody can be an aristocrat, because the core of being an aristocrat is to be superior to most people in the relevant social hierarchy.
Wolf: Not everybody—for example, think about houses. Land is not freely available. Land is intrinsically constrained. We need houses, and I live in a particularly nice suburb of London, and everybody can’t live here. The result of that—the fact that everybody can’t live here—is that houses cost millions. They didn’t when I bought mine, but they do now.
The point is, everybody can’t live here. Everybody can’t have the perfect view. Everybody can’t own a Swiss lake or a beautiful island in the Caribbean. In other words, there are positional goods which are intrinsically limited by virtue of the demand from all the other humans. Somebody has to allocate this somehow, through social status or market competition. We do it through a mixture of the two. Above all, market competition generates the wealth. If you win in market competition, then you can buy the island in the Caribbean. But these are not things that can become abundant, and positions in society can’t be, by definition. Only one person can be the president or the prime minister at any time. We could imagine a competition which isn’t through voting, as we do it now, but one can manage.
Essentially, a world in which goods are freely available—that is imaginable for many goods anyway—but there are lots of things we really care about which will remain scarce, and they’re really important because they’re about social status. If they’re not done through the market, which is perfectly possible, then you have to manage them through some other allocation, and, by and large, historically, it’s been done through force.
That was, of course, Marx’s great mistake. He thought that if we got rid of private property, all these problems would be solved. I’ll leave aside his problem about abundance, but the problem he found was that it wasn’t solved, because somebody with the biggest gun—and the largest number of guns—simply took whatever he wanted and allocated the rest to the people who supported him. So I think the problem is much more profound than you suggest, even in this world.
In the rest of this conversation, Yascha and Martin discuss whether we are in an AI bubble, what would happen if it burst, and the state of the European economy. This part of the conversation is reserved for paying subscribers…