Interview: Simon Johnson

Power and prosperity in the age of AI

© Nobel Prize Outreach. Photo: Clément Morin.

Simon Johnson is a British-American economist and the Ronald A. Kurtz Professor of Entrepreneurship at the MIT Sloan School of Management. In 2024 he shared the Nobel Prize in Economic Sciences with Daron Acemoglu and James A. Robinson. He previously served as chief economist of the International Monetary Fund. His books include Power and Progress, 13 Bankers, and White House Burning.

What follows is a transcript of our conversation, which took place on September 10, 2025. The transcript has been edited for length and clarity.

[Note: The opinions expressed by the interviewee are their own and do not necessarily represent those of the interviewer.]

Do College Degrees Still Pay?

CB: To touch on something you addressed in your book, Power and Progress: To what extent do college degrees still confer prosperity, and what is the relationship between college education and prosperity over the last few decades?

SJ: Well, there’s no guarantee that any individual course will make someone prosperous. It’s a very competitive world out there, and becoming harder for people at all education levels. But if you look at what’s happened to real wages since the 1960s, on average, people with a college education have done better than people without, and actually postgraduate degrees have done even better. So the jargon among economists is “skill-biased technological change,” meaning the way technology has changed has disproportionately favored the productivity and therefore, ultimately, the income of people with more education.

And what’s also striking is not just that people without education have done less well, that’s pretty obvious to people, but when you parse the data really carefully for men who didn’t finish high school, real wages have not moved much since the 1960s, and that’s pretty dramatic, and actually quite unusual for the industrial phase of economic development. When the tide rises, all the boats are supposed to rise. That’s the metaphor, right? Some boats have risen a lot. Other boats have stayed entirely flat in real terms.

So despite all the changes we’ve seen, despite all the technological changes and everything else in the world, a big chunk of the population hasn’t done well. And remember that about 60% of the current American workforce doesn’t have a college degree. So this is a lot of people.

Elite Overproduction

CB: Do you see there being a kind of reversal of the trends of the last few decades, where college degrees have become a lesser guarantor of prosperity? I think we’re seeing what Peter Turchin and others might describe as elite overproduction in the cities, and some people are attributing recent urban political trends to this phenomenon. Do you think there’s a problem with elite overproduction?

SJ: Overproduction is a possibility. And I think India, for example, is a country that has had exactly that. A lot of educated people can’t find productive jobs. I don’t think it’s one of our longer-term problems in the US. I do realize and understand the unemployment rate among people with college degrees has gone up, as it has for lots of other younger people. I think that’s temporary. I think better-educated people will find new opportunities and do better, including as artificial intelligence arrives.

I’ve never said that it’s going to be easy. I think the world is really quite competitive. And if you think about the apex of American society, what are the firms at the top? What are the jobs at the top? How many government officials do you have at the top?It’s a fairly fixed, limited number of people. And the population of the country, of course, continues to grow. When I came to the US 40 years ago, there were about 100 million fewer Americans than there are today. So that’s a big expansion in population.

The number of places at top Ivy League schools has barely changed over 40 years. So there is a problem of a lot of talented people trying to fit into a relatively few top places. But on the other hand, you can create new things. You can create new spaces. It’s a big country. There are a lot of places you can live, a lot of places you can be productive, a lot of places you can have a good life. And college-educated people are generally able to grasp new opportunities.

So economists talk about “human capital,” meaning, in other words, your education. But then within human capital, they talk about being very specialized, or having a general level of human capital. The people I work with most closely are MBAs and mid-career executives at MIT, and they often have very specialized skills. But also, and this is what the MBA cultivates, they have general human capital. So if they’re pursuing their passion project today, and all of a sudden they realize it won’t work, it’s been automated fully, or somebody else is going to dominate their space, they can generally switch to something else and do really well in that.

And that’s the kind of training and ability you need. College education is not necessary or sufficient for success, perhaps, but it is pretty helpful. And the kind of technology we’re seeing develop today in and around AI is almost surely going to reinforce that.

Can We Still Write?

CB: What do you make of the state of writing as a university professor and as a writer yourself? Do you think there’s any societal problems surrounding the ability to write?

SJ: Well, I think it’s certainly a problem if anybody relies overly on AI and assumes that AI can do the work. People talk about cognitive atrophy. That’s probably, almost certainly, a thing. You know, I think the question is related to: do you think that the advent of the pocket calculator undermined the mathematical abilities of Americans, or educated Americans, or anyone? And I think the answer is no. Pocket calculators, or being able to do calculations on your spreadsheets, is just fine. There are other problems with math education, other problems with math literacy, surely, but just having this technology available is not the problem. The problem is probably the incentives that we have to write well or to communicate effectively and so on.

But I’m sure, I’m confident—I mean, I talk to people who are employers of folks with all kinds of educational backgrounds—and the ability to communicate in written and oral form is very important for the job market. So the incentives are still there for people to apply these skills.

Who Really Holds Power?

CB: I was at a speech given by Steve Bannon in Washington, DC last week, and he made a very interesting claim. He said the most powerful institutions in the United States at the moment are the Federal Reserve, the CIA, and CENTCOM. Those were the three. I thought that was interesting. And I wanted to ask, in your view, what are the most powerful institutions in America, or internationally if you’d prefer?

SJ: Well, I certainly agree that the Federal Reserve is important, and other central banks around the world also are important, including the European Central Bank. And then they form a network of some cohesion among developed countries. And that’s important. I think ministries of finance are very important, and how they think about budgets and debt and so on. And I think how they interact with financial markets is incredibly, incredibly important.

I also think the universities matter a lot. I mean, the secret source of American technological progress after World War II wasn’t so secret, it’s quite loudly open. The federal government gave money to university scientists, and they did work as specified under contract, and then either brought that back to the government through the DOD, or they commercialized it, or otherwise made it available to the private sector. And that’s a phenomenal, was a phenomenal idea. Really brilliant breakthrough had not been done before. It really came out of World War II, lessons from that, from that effort and how to harness the scientists.

And I think the pressure that universities are under now on the funding side and on the governance side is potentially quite damaging. So, that’s an institution that is taking a bit of a beating, and we’ll see how the financial markets feel about what the ministries of finance are doing. Obviously, France is very interesting right now, because they have a serious problem, and, you know, they have some advantages. And I’m not saying there’s an imminent disaster or anything, but that’s the kind of pressure that other countries may expect to see.

From Robber Barons to Tech Barons

CB: I think power is like a magnetic pole in that it tends to shift location over time. I wonder if you could trace the location of power in the United States over the last century and a half, maybe from the beginning of the industrial era. Where would you say power has shifted during that time period?

SJ: I do think that one point that was not anticipated by the framers of the Constitution or by anybody else in the late 18th century was that invention would become so important, both for commercial activity and for, let’s call it, national security. So the world is, you know, humans have always been inventing. They’ve always been tinkering with stuff and making new things. But the pace of invention really picked up at the end of the 1700s into the mid-19th century, and the US went from being a pretty backward, mostly agrarian country in, let’s say, the 1840s to being the leading industrial power in the world by the 1890s. Other countries also obviously industrialized.

Now we talk about the robber baron period in the Gilded Age at the end of the 19th century. I think that’s correct. That was very important. The US Senate was called the Millionaires Club, because you had to be a millionaire, which was a lot of money then, to be a senator. And then we went through sort of a re-founding of the Republic, a bunch of constitutional amendments and changes in political norms, so that the power of that moneyed elite receded, and what rose instead were the professional inventors, the university professors, the industrial labs run by big companies, who were sort of corporate inventors. And they were very powerful, including in the military-industrial complex through the 20th century.

But I think the people who are in the driving seat now are the people again inventing technology, making large fortunes, setting the rules and also creating the expectation. So, you know, with regard to AI, which is surely the biggest invention of our era: Is this going to be pro-worker, enabling a lot of people to work better, and pro-democracy, enabling us to have a better, regulated or reasoned democracy? Or is it going to be something in the hands of a few people, making them really, really rich, and the rest of us get crappy jobs, and the democracy becomes excessively and extremely polarized in a way that’s not productive?

You know, these are really consequential choices which lots of people can weigh in on, but the direction of that technology is in the hands of relatively few people, and I think those are the people who are really driving the economy now. I could take you back to the 1840s. We could talk about Eli Whitney or Samuel Colt or other people who built new systems of manufacturing. And we could lay out the antecedents and draw a line. The inventors have mattered all along since the Industrial Revolution. But the power of invention is new with the Industrial Revolution.

And remember, this power is commercial. It’s about national advantage and national disadvantage. And within countries, there are consequences, but it’s of enormous consequence for international competition. Do you have better fighter aircraft? Do you have better ships, naval ships? Do you understand, right now, the really pressing question: who understands drones? Drone warfare, anti-drone warfare, are really, really important. And if you’ve got the inventive capacity and the people to comprehend that, to come up with new solutions, to manufacture them, to create material representation, obviously new ideas, and then to innovate further, you’re in a very different place in the world than if you are purely on the receiving end of drone attacks.

CB: One of the things that fascinates me about the winning coalition that converged around Donald Trump’s candidacy in the 2024 election was that you had the populist forces, which I would say were more closely aligned with a pessimistic view of technology. And then you had the great tech barons of our time, the Elon Musks and the Peter Thiels, who are very much techno-optimists. And there’s a fissure in this movement already, I would say, that has to do with this very issue. How did it come to be that these forces, diametrically opposed as they are, converged around the same political platform?

SJ: I certainly agree about the techno-optimism point. These are the people who famously had the seats close to the president at the inauguration, right? The tech barons. They want a future, or at least a near-term future, without regulation. Nobody stands in the way. The government facilitates their investments and helps them project their influence or power around the world. That’s what’s going on there. And for sure, they take note of it. They are the optimists; that’s what they tell us.

On the techno-pessimist side, I’m not sure the most pessimistic voices are in Mr. Trump’s coalition. What I hear from Steve Bannon and others is more about globalization. They’re pessimistic or negative about the global economy, supporting tariffs and other restrictions, and thinking the US will get a better deal that way. I don’t see the full span of technological views reflected in the Trump coalition.

But any successful, winning coalition in a country as diverse as the United States is going to have a range of views. The interesting question is, can anyone stand up to, or is anyone willing to stand up to the tech barons who are on the charge right now? And I think to your first approximation, speaking roughly, they are essentially unfettered in the political sphere, and substantially in the social sphere as well.

If you look at people writing books, including those on the left, even those developing alternative agendas, there’s very little about corporate power. People talk about all kinds of issues, of course, behind closed doors or in conferences. But in terms of leading, salient ideas of the moment, no. Nobody wants to stand up to the tech sector. They’re too powerful.

CB: They are also responsible for a very large portion of equity growth in the markets over the last 15 years. Disproportionately, right? The FAANG stocks plus Tesla and Nvidia have largely carried the S&P.

SJ: The power of persuasion is not just about writing things down. It’s also about helping people make money and feel richer. I’d emphasize that even if we had more competition and more diffusion of power, you could still do very well in the stock market and with investments in and around the tech sector.

The main effect of strong monopoly-type power is to concentrate profits in the hands of a few people. Elon Musk being offered a trillion dollars worth of compensation by Tesla, for example. Look, technology is important. It creates value. The stock market tries to look forward. Sometimes it prices things wrong. Before 2007, a lot of stock value was concentrated in the financial sector, which turned out to be somewhat illusory.

The tech sector, by contrast, has created remarkable products that we all benefit from. The question is, how widely will those benefits be shared? And what I would worry about, Carson, is what I think the Trump coalition should worry about, but doesn’t apparently, which is interesting, is that prosperity is not going to be shared by the highly concentrated tech sector.

So yes, you’ll get more progress, and yes, you’ll get more change. But you’re not going to get higher real wages. And I think that’s going to be a problem, for all of us, not just that coalition.

CB: There’s something else, which I guess you could tie to the rise of populism, but also to a historical mega cycle. I once read an author who wrote about the age of constitutionalism as a result of the development of the mercantile, fiscal-military state, right? So conscript armies and very expensive navies, which required more revenues drawn from the populace. That influenced a rise in constitutionalism, because you needed the people to cooperate in order to achieve your aims as a state. Today it seems the populace has much less leverage in dealing with the state, and when dealing with power in general. What do you make of this?

SJ: And of course, it also is again, a very broad question, but that’s not a problem. I think, just on the history, remember, there were some very powerful monarchies in Europe, in Western Europe, in the 1400s and 1500s and 1600s. And then there were some rising, kind of scrappy mercantile powers attempting to take them on and carve out some space themselves in Europe and around the world. And that would be the Netherlands in the first instance, and Britain as the primary driver, and some other places as well.

And in those scrappy upstarts, you’re quite right that they found, after a lot of struggle in the Civil War in Britain, they found that they could strike a balance where you kept a king, but the king had constrained authority, and you had a parliament, and you needed the assent of Parliament to raise taxes. And within Parliament, the merchants were very important, and this is before the Industrial Revolution.

And I think it’s important that none of this was built with anyone imagining what was coming next. But it turned out those conditions, that representation for merchants, and that principle of representative democracy, limited representation, turned out to be very useful, and actually what you needed for the Industrial Revolution. So the Industrial Revolution started in the north of England, where I’m from. I’m from Sheffield, which was a metal-making, metal-working city. Manchester got in first with textiles. Birmingham made a lot of machinery. Other cities like Leeds, Sheffield, Leicester, Liverpool, got in on the act.

And the industrial class, the inventors, were very small-scale. My family made screws in Sheffield for about 100 years.

CB: Like the Chamberlains.

SJ: The Chamberlains made screws in Birmingham. We were beaten out, I think, by the Chamberlains. The Chamberlains were quite big already by 1900. My family never made it big time, but you’re right, the Chamberlains came from screws and similar sorts of metal-working. My family never employed more than eight people, that was maximum employment. So very small workshops, very tough market conditions, but they made some decent money sometimes, and the business survived for a century. Pretty impressive.

But I think it’s that legal framework and representation. Of course, the British had to go through a number of parliamentary reforms in the 19th century to expand the franchise and to give more votes to the north of England. Both the middle class wanted that, and the workers’ labor movement wanted that. That turned out to be really helpful for prosperity, because that encouraged people to make investments, and that was an adjustment of the relationship between the state and the private sector that was possible within that constitutional arrangement.

Now, is it the case today that people have lost their power or voice? Well, to some extent, you’re right. I mean, certainly the labor movement is weaker than it was. That’s an issue. And there are people who want there to be new monarchs, and there are voices claiming that autocracy, or a CEO approach to the country, would be helpful. And I think that’s fundamentally misreading history, because, you know, there have in history been some autocrats who were quite good for economic growth, either headline growth—I mean, Stalin had good headline growth, although he murdered millions of people. But there are also autocrats who have good headline growth without killing lots of people. But they never last, right?

I mean, the problem with autocrats is the autocrat gets old, they get cranky. Who’s going to be the next autocrat? There’s a fight for power. Autocracy is a failed system. I mean, not just once. It’s a multiple, repeated, in-your-face failure. But if you’ve got some oil, or if you’re standing at a useful place in terms of geography, you can jolly people along for a while, and you can persuade them this is a low-tax place or whatever. So, you know, autocrats will always be with us, but they’ve got no staying power, and they’ve got no long-lasting innovative capacity.

Unless you believe that China has changed the world completely, and that the Chinese Communist Party dictatorship is going to become somehow incredibly inventive and innovative. Despite the fact that top-down systems just cannot manage to generate the pluralism of ideas and competition that generates prosperity. But sure, people believe in China right now. We’ll see how they go. Give them a couple of decades.

Again, it’s democracy, and its constitutional constraints and decentralized initiative that has brought us the modern world. That doesn’t mean decentralized initiative couldn’t run into some problems. We could certainly, with the wrong incentives, overheat the planet massively and deeply regret that. So it’s not that that’s the only thing you need. I think Churchill is widely quoted as saying democracy is the worst of all possible political systems apart from the alternatives.

CB: I would say one of your central theses in Power and Progress is that innovation does not inherently lead to broad-based prosperity, right? Would you agree with that?

SJ: Certainly, there have been episodes, prominent episodes in world history, when there’s been a ton of innovation, as in technological change. Congratulations. You now have windmills and water mills, new systems of irrigation and power in medieval England and France. And some people do really well. I mean, innovation only happens when somebody says, oh, this new machine is pretty good for me, let’s do it. But in those same episodes, including medieval Europe, we didn’t see increased real wages or living standards to the extent we could measure it.

And in the early Industrial Revolution, people made a ton of money. Entrepreneurs made a ton of money in Manchester, for example, from the mechanization of weaving. And the mechanization of weaving, per se, was not particularly good for weavers, either in the factories that adopted the machines or in the artisanal ecosystem that existed previously.

So yes, we have seen episodes of profound technological change and progress or improvement in productivity that have not been at all correlated with, or leading directly to, or within half a century, higher real wages or better living standards.

CB: In today’s era of artificial intelligence, we also live in an era of mass surveillance capitalism. I wonder, what leverage do populations have in 2025 to achieve something like the collective bargaining power of the labor movement? What sort of tools exist at the hands of the people, if you will?

SJ: I think that there’s a lot that can be done in terms of understanding the situation, explaining it. I think agency can be acquired through more decentralized innovations that actually figure out how to use AI in different, better ways that help lower less-educated people. And I think that the government could play a useful role in terms of limiting how much surveillance is on us, protecting privacy, and throwing out some grand challenges to encourage innovation to move in particular directions.

The current obsession with self-driving vehicles was really catalyzed by a DOD competition about 20 years ago with a total prize of $1 million to drive vehicles by themselves around a patch of desert. But that really got people thinking, and it pulled a lot of talent into that sector. And we’ll see where that goes in terms of jobs and prosperity.

But I think catalyzing the use of technology around better education, more education for more people, more useful education, better health, better understanding of health, more individual control over health in an informed way, and more good jobs. I mean, the government could work on all of this in a non-heavy-handed way if it wants to. It’s not doing it. I understand, you know, that’s not the flavor of the month. But we’ll see what the AI barons bring us, and we’ll see how happy people are.

Innovation, War, and Deterrence

CB: The epic defining periods of innovation tend to have something to do with warfare and conflict. I mean, look at the innovations of Germany and the United States during the Second World War. You look at the Cold War era, the creation of Silicon Valley as a technological force, largely thanks to Pentagon funding. On a smaller scale, you look at Israel, which has pioneered a number of innovations thanks to its robust military state. There seems to be a wave of innovation underway relating to defense, relating to automation, drone warfare, etc. What do you make of this chapter of our history? Do you think this could steer us toward conflict, or do you think this is ultimately good for deterrence?

SJ: Well, I completely agree that there’s a symbiotic relationship between commercial innovation and military applications. And you and I have talked about this in terms of air force development before World War II. I think that was a space in which that was happening, and there are other spaces, but I would emphasize more today. But I think the point is the same: does investing in defense make war more or less likely? That is a very old question.

Of course, if people think that you can defend yourself effectively, then attacking you becomes more costly, and they should take that into account. But a lot of times, these defense systems, particularly when there’s new technology, are not proven, not understood by anybody, including the people developing them. And, you know, I think May of 1940 is a leading example, where the Germans didn’t know what would work or not work in terms of air power, for example. And neither did the Allies.

The Germans thought they had some theories. They believed some things. They thought they’d had some success. They’d seen some things in the Spanish Civil War. They’d seen some things in Poland, very limited things. And they had a concept of what would work. And in May 1940 they really shocked the British and the French, and of course the Belgians and the Dutch, whose doctrine of air power was not effective in defending France.

Obviously, it was partly about tanks, but a lot of it was about how they used aircraft. Very new, very brilliant. But it turned out that the British had thought about air power, and they had thought about air defense, but in a slightly different context, through the lens of protecting the British Isles against foreign invaders and against German aircraft. And the British had a remarkably successful air defense system that, in July, August, September, completely turned back the Germans.

It turned out that the German air force, the Luftwaffe, didn’t have the right doctrine. They had not understood strategic bombing and what it meant. They had underinvested in four-engine bombers. They had not extended the range of their fighters. They had made a lot of mistakes that cost them dearly. But they didn’t know that, and the British didn’t know that.

And we go back and say, oh, it was a near-run thing, and so on. Actually, I don’t think it was. I think the Germans completely defeated the Allies decisively in May, and in July, August, September, the RAF completely defeated the Luftwaffe. It was completely unable to achieve its objectives. And without that, of course, you couldn’t invade Britain, and you couldn’t win the war in the West. And there were a bunch of other consequences.

Now, if Hitler had known all that, would he have still attacked? Maybe. I mean, he liked to roll the dice, right? It was all probabilistic. So I don’t think you know. I do think going around threatening countries and/or attacking them, and being aggressive, is often a mistake. And being provocative can backfire. But I do think defending yourself effectively is entirely reasonable. And then there’s a legitimate discussion about where you draw which lines, both in terms of technology and in terms of doctrine and actual action.

CB: It’ll be very interesting to see how some of the large-scale procurement programs that are now underway will play out in the coming decades.

SJ: Yeah, look, I think, Carson, that the big lesson on the military innovation front is it’s all about owning the future. And I was talking to some prominent European official who said, ah, yes, in the face of the Russian developments and what we’ve learned about the US commitment to NATO and so on, we’re going to build more jet aircraft. And I said, well, you might want to do that. But are you really sure that jet aircraft are the future of air power?

Because haven’t we learned from the Ukrainians and the Russians that drones are going to be a huge part of what you need? And exactly how drones interact with jets, and where the jets fit, is not clear to me, at least. And if you can confront and defeat or neutralize a $100 million jet with a bunch of drones that cost $1,000, you’ve changed the economics of warfare in ways that are hard to anticipate.

I write science fiction now, Carson, I think I may have told you this in Helsinki. In part because I want to imagine these futures. I find economics to be a useful, but fairly limited tool. I think you have to consider the holistic system. And if you write fiction, and if you force yourself to be consistent, and so do your readers, I think you start to see things a little bit differently.

So I don’t know what the future of warfare is. I speculate on it in my book. But I think pushing ourselves, not necessarily into big procurement programs, but into massive innovation efforts, that’s the key. And then the question is, who’s capable of innovating? How do you innovate? Can you use AI to leverage your human assets in innovation? Those are the key questions.

I may have mentioned this to you, but the engine in the Spitfire and the Hurricane, the Merlin V12 liquid-cooled engine in 1940, was something the British started to develop in 1929. So Winston Churchill spun this story about how the British were not ready for World War II. At a strategic level, sure, they weren’t prepared, and they certainly didn’t want to fight another war in northern France in the trenches, and so they didn’t prepare for that one. But they were building for war and pursuing technological paths that changed completely how wars were fought, 10 years before the war broke out.

If they’d started to try and develop that engine after the Munich Crisis in September 1938, they would have had nothing, right? So it’s not about going all-in on your military stuff at all. It’s not about building huge armies. The French had the biggest air force in the world in 1923. Didn’t do them any good in 1939, did it? Because technology had moved on.

So when technology is moving, you want to be at the front end. You want to be driving it. And what drives it is your creatives. And you have to ask the question: who are these creatives? Where do they work? How do you motivate them?

The Soviet Union had a bunch of them working in Siberia, and the Soviet Union doesn’t exist anymore. China has a bunch of them working under central control. We’ll see how that goes. I put my money on the American universities: very creative, very competitive places, where ideas rise and fall on the basis of their merits, very driven by results. You can give them different sized carrots. You can give them sticks if you want, but I’d recommend carrots. Then see what you get.

Because if you try to pursue the civilian productivity-enhancing innovations at the same time as the military ones, you don’t have to choose. A lot of people frame it as, do you want guns or butter? The classic discussion. Well, that’s a fair question if you’re doing procurement. But if I want to own the future of innovation, I want the future of butter and the future of guns. What’s beyond butter? What’s beyond guns? If I’m creative enough, and really push, and have the right infrastructure, I can own both.

And we have a lot of this in the US. It’s not perfect. It could be a lot better, and it could be spread around the country to include a lot more regions and people. That’s the magic that was understood first in the 1940s, that was implemented in the 50s and 60s. We kind of let it drift away from us and let other countries rise up. But we could get it back.