Judith Dada is Co-CEO of Langdock, a leading European AI company, Senior Partner at Visionaries Club, an early-stage European venture capital fund, and is co-author of Europe 2031.
In this week’s conversation, Yascha Mounk and Judith Dada discuss what the decline of German industry and infrastructure reveals about the broader crisis of the German model, why the European tech sector is so far behind China and the United States, and what Europe needs to change to become competitive.
This transcript has been condensed and lightly edited for quality.
Yascha Mounk: You live in a state, Mecklenburg-Vorpommern, where the AfD is polling particularly strongly. What do you think explains the rise and rise of this party and this movement more broadly in Germany?
Judith Dada: I think it’s the question that all of us are grappling with at the moment. I think the most interesting piece is that this is very much not a German problem. This is a problem, let’s call it a Western problem. We see the same tendencies in the United States, we see the same tendencies in France, we see them in the Netherlands, we see them in the UK. People feel left behind. I think they feel like tomorrow is not going to be better than their life today or their life yesterday. There’s a kind of melancholy, I think, that people feel for the good old days, for a world that was maybe slower, where the worth that they had relative to their environment was a little bit clearer than in the world today.
So I do think it’s a multifaceted problem, but it’s by no means a German problem. I do think that the AfD is maybe the ugliest expression of this problem, because looking at all the far-right parties across Europe, it is one of the most extreme parties. Some of the politicians who are still in the ranks of this party—it just baffles me that this party has not found ways to exclude these people or to more clearly draw a line in the sand. This is not just about conservative or right-wing politics. This very much, unfortunately, in certain aspects, has unfortunate parallels to Nazism and to Germany’s dark past. I think that is just a tragedy.
But at the same time, there’s a much deeper problem here, and one that is very much understandable, because otherwise more than forty percent of the German population in Saxony-Anhalt wouldn’t be voting for this party. There is a problem that is societal, and that is not just a fringe right-wing problem. I’m thinking about this every day, and I do think at the end of the day, what it requires is both a change—an actual drastic change in politics—but then also a message and a vision for the country and for the West, that is able to look into the future in a way that carries people with them and doesn’t give people the feeling that, yes, by and large the world is moving forward, but “I’m stuck in place.” Thereby the only way that I can progress is by finding scapegoats and giving an outlet to all this negative feeling, because it feels like now it’s just a matter of taking something away from someone else so that I can have more. We need to give people the feeling again that the pie is growing for everyone.
Mounk: One of the striking things about the AfD is that I think it’s more ideologically radical than other far-right populist parties. It’s also just less professionalized. This is obviously a problem and a challenge that newcomer parties always face—that’s true in certain ways of the Green Party in the 1980s when it rose to positions of political power. But in part because of a firewall, and the way in which the AfD has continued to operate as this kind of outlaw within the German party system, and in part because they’ve taken it as a matter of pride, the level of unprofessionalism is quite extreme. Among the people who were elected in Saxony-Anhalt, or who are currently candidates in some other states for election, there are actual former neo-Nazis. I think there are two of the roughly thirty-odd state legislators we now have who have been somehow involved in the porn industry earlier. There’s somebody—I forget whether it’s in Mecklenburg-Vorpommern or somewhere else—who’s actually been sentenced by a criminal court for being a bank robber. That sort of goes beyond ideology and makes it quite extreme.
In terms of your experience of living in Germany, which I no longer do, and living in a part of the country where the AfD does have particular support, you must speak to people who are voting for the AfD all the time. You visibly have—this is my least favorite German phrase—a “background of migration,” Migrationshintergrund. It’s a very weird German bureaucratic term. Do you find that a lot of the people who vote for the AfD are, in person, perfectly pleasant and charming neighbors, and it’s hard to square that with how they’re voting? Do you find in your daily life there to be a lot of disappointment and anger? One of the strange things about East Germany is that, for all of the economic challenges, especially over more rural parts of East Germany, it’s also a part of the world that is vastly more affluent today than it was thirty or thirty-five years ago at the end of communism.
Dada: This is a really good question. I would definitely say I live on the island of Rügen in Mecklenburg-Vorpommern, and I spend a lot of time here in Stralsund and the surrounding cities. My day-to-day experiences are not ones of xenophobia or of people showing any signs of mistrust or anything like that. I have lovely neighbors. I live in a lovely village with lots of people who’ve lived here for generations, and they have been nothing but welcoming and really kind to my husband, myself, my children, my father, who’s a Nigerian immigrant, who has much darker skin than I do. By and large, I think the experience has been very positive.
I think, however, what I’ve definitely noticed—my father was visiting last weekend and we went to Stralsund and went for a walk in the city, and obviously you see all of the voting posters, all the AfD posters everywhere, and you see the odd German flag hanging out of an apartment building or in front of a parking lot. There is a feeling of nausea that I feel inside of me, much more prominently now than I did years ago, because I walk around and I have this lingering fear: is my father, am I, being perceived in a way where we’re not belonging, or we’re not welcome here? Again, it’s not actually what I’m perceiving in the people that we then meet, who are perfectly lovely and welcoming, in the way that I’ve always experienced Germany. But I do notice myself having this fear more and more.
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 leonora.barclay@persuasion.community
I think it’s so unfortunate. If I was in France and I saw a French flag, I don’t think I would have the same trepidation as when I’m walking around in my own country, where I was born, and I see a German flag. I always ask myself, is this because they just love Germany, or is this because they’re very right-wing? I think it’s just a tragedy that I now get to have these questions, and I wish I didn’t, but it’s just my honest feelings. I’m sure if I’m feeling that, then there are many, many more people who look like me, or who don’t look like bio-deutsch, or whatever word you’d choose. I think that is something we must grapple with societally, because that’s just not how we live together in the best and most trusting way.
Mounk: There’s a debate about what it is that caused this vote. One progressive economist shared a poll in which people emphasize their concerns about a standard of living. He highlighted one answer to the poll and neglected to point out that the one just above it, with more people, was that people were concerned about immigration.
I think one of the things that the details of this debate sometimes get wrong is just a much broader sense of malaise, that things are going in the wrong direction. That might have to do with a feeling that the state can’t generally control its borders, and therefore we’re not really controlling who’s coming in, and that’s a challenge to public safety, etc. It can have to do with the fact that inflation is high and we don’t get the sense that things are looking up economically. But I think it is a more deeply rooted sense that our model is in crisis and the people who are in charge don’t really know what to do about that.
If you’re being sympathetic to that general feeling, there is something to it. I feel it very strongly when I’m in Britain, where I was last week for an event with Francis Fukuyama to launch his memoir. So if you haven’t yet listened to the series of conversations I’ve had with Frank, dear listeners, you should go and do that. But it’s a really stark contrast between when I came to Britain to study a couple of decades ago and today. I do also feel it coming to Germany. Famously the German trains no longer work—they’re now less punctual than the Italian ones—and people are always a little bit skeptical when I say it to them, but it really is very striking to what extent you encounter problems virtually every time you try to take the train, and people have just organized their lives around the idea that German trains are just never reliable.
It’s true of the German car manufacturing sector, which has really fallen behind the times. It’s also true in the sector in which you are active and make a living, which is European tech, and German tech in particular. What is your impression? Is that an overstatement, or is the German model really falling apart in ways that, however rational the response of voting for a party of bank robbers and porn producers may be, makes it rational to be very concerned about our ability to confront these challenges?
Dada: I think very much. Let me take a step back by saying that I do think the root of the problem is both a moral and an economic problem. I do think there’s a moral dimension here where people feel like there was a lot of progressive change that we’ve made over a relatively short period of time. Much of this change people like me benefited from, and much of this change I welcomed. But it was change that left people with a feeling of I’m no longer able to say this.
In Germany there was a huge discussion around gender—whether you say the female version of the word or the male version of the word—and a lot of this discussion, I think, in many ways didn’t really matter. This whole gender discussion, I always thought, got so much airtime relative to the actual relevance that it had for our day-to-day lives. But people felt very, very strongly about it.
Mounk: Just to illustrate this for the audience, because it’s in some ways a much more extreme version than anything we’ve discussed in the English language, it’s helpful to point out some of the changes that were made in the German language. In English, insofar as there’s a verbal or linguistic debate, it’s about the legitimacy of the third-person plural pronoun in the singular—is it fine to say “they” for an individual? That’s a bit of a change in the English language, but actually with antecedents before the question of gender really came up. It doesn’t particularly bother me very much.
In German there are two changes which are quite a bit more striking. One is that, as in many languages, the plural of nouns has often historically been gendered. So if you say Studenten, that includes both male and female students, but grammatically it’s the male form. There’s, I think, some quite bad linguistics behind the idea that if you say ein Aula voller Studenten—a room full of students—you picture a room full of just men. I’ve just never bought that that’s actually true in 2026. If I say I stood in front of an Aula full of Studenten, I don’t think anybody pictures a room that’s exclusively male. I don’t quite buy that. But be that as it may, the solution is to change to Studierende, which is a weird gerundive—that is to say, people who are currently studying. I have to say I have no particular problem with that. As somebody who left Germany before that became current, I’m just not at all in the habit of using it.
I did have one experience where I was on a panel and spoke the way that I grew up speaking, without any particular political implications, and got very hostile questions about why I wasn’t using this new linguistic form. I said, look, I’m a smart person who is pretty versed in language and who makes language my living, but I don’t live in Germany, so it didn’t come naturally to me. Think of all the people who are not up with all of the latest trends and who perhaps don’t have the same verbal facility because that’s not their job. Who then get that kind of hostility just because they spoke in a way that the German language has been spoken for hundreds of years.
The other change that I think is even more striking is that, to make a gender-neutral form in writing, you now put a star behind the ending—Student*in—to show that it could be a student who is male, or Studentin, who is female, or any other gender, indicated by the star. I personally don’t like that as a written form. But the thing that’s really quite odd is the verbal form of that, where you’re supposed to indicate—you can’t say “Student-star-in”—so you’re supposed to make a little pause between “Student” and the ending, and say “Student [pause] in.” Irrespective of how you feel about acceptance of trans people, etc., it’s very unnatural to introduce an artificial pause into a language. I don’t think it will ever take off in a broad way, and so it really just becomes a marker of educational attainment and belonging in a particular milieu. For understandable reasons, people then find it quite off-putting, and I think it just raises the price of acceptance of things like gender diversity in a way that I personally find unnecessary. Sorry, I didn’t mean to go on this little lecture, but for an English-speaking audience, I hope it is helpful.
Dada: No, I love this. A hundred percent—thank you for these very illustrative examples, because it’s exactly that. People felt like they were being outlawed in the country and in the language that they very much considered their home. Think of that what you want—change has always been around, and language has always been changing. So I think it’s not so much about whether we can change language to be more inclusive. I think it was more that there was a moral weight that came with this discussion, where the people who were not adapting to this change in a really rapid way were somehow seen as less, or different, or racist, or pick your favorite insult.
Speaking more societally, that was one example, and there are many, many other examples. You mentioned migration—I have a migration background myself, but I do feel that oftentimes in these migration debates, I was able to say something that my German friends just weren’t able to say, because I had a migration background. That’s just a bit weird. That’s just not a society I want to live in. I want everyone to be able to speak their mind and voice what they are actually feeling and thinking without immediately being seen as racist or as less than.
All of this was coming together in this moral change, and then next to it there was the economic change. I think those two dimensions are tied, because I think the people who most felt the moral backlash of not wanting to change so quickly were also maybe the people who were feeling like they were being left behind economically relative to some other people who were benefiting more from globalization, from technology accelerating, and so on. Both of those things are intertwined and reinforcing one another, bringing us to this current moment and to your question about whether the German economic model, which was very much the powerhouse of Europe, is still working, and what we do to fix it.
If you look at all the broad statistics, Germany was in a recession. Now we’re no longer in a recession, but the growth outlooks are not looking amazing, and I always find it funny how we get bullish about the economy because we’re correcting growth upwards by 0.1 percent. But really, by and large, we’re just not growing as an economy. We’ve been in this kind of stagflation period for too long. A lot of the industries that Germany has always taken pride in—you mentioned automotive, the chemicals industry is another great example—many of them, also tied to really high energy costs, are no longer competitive, and we’re seeing tons of jobs that are either being cut or moved elsewhere.
The German car is such a great example. We always took pride in our German automotive brands, BMW, Mercedes, and so on. If you now go abroad, because of having completely misjudged and underestimated the relevance of electrification and the ability of Chinese car makers to catch up in what was very much going to be a fundamental transition in mobility—in the way that we think about the value of a car, much more electric value creation, much more software-defined, and no longer the combustion engine that made us so strong for so many decades—it was unthinkable that Chinese brands could ever have the same prestige and brand recognition as the German car brands. But if you go to countries like South Africa, or anywhere else where German cars were always sold at a premium, you now see BYDs and all of the Chinese car brands flooding the market, and people no longer really care about the fact that this is a BMW or a Mercedes. So the last vestige of German pride—that no one will ever be able to have a premium car brand in China—is now being taken from us as well.
I think it’s another example of how people are feeling that their identity and their value are being eroded over time. At the end of the day, this was true a thousand years ago and it’s going to be true a thousand years from now: we want to matter as humans, we want to be seen, we want to know that what we do is good, that someone deems it as valuable. It’s a fundamental part of making sense of our identity as a social being within a community, and I do think people are feeling that that’s being taken away.
So, back to the question: is the German economic model still working? No, it’s very clearly no longer working. If you look at the tech sector overall, in the United States there are 150 companies that make over a billion dollars in software revenues. In Europe, there are about 30. There’s a very stark difference. We now have the Mag 7—some of the biggest tech companies in the world—which, if you take them all together, are almost half the total market cap of the world. So we see this extreme concentration of economic wealth and growth in one sector, which is tech.
Interestingly, this is not so much a European or German problem as it is a technology problem, because if you look at the established non-tech industry in the United States, it’s suffering from exactly the same malaise as German or European industry. U.S. economic growth is by and large boosted by capex investment into data centers—it’s driven by the tech and AI boom. The non-tech industry in the United States is growing at the same rate as German or European established industry. But that still doesn’t do away with the fact that we just don’t have as strong an industry in the technology or R&D-innovation-heavy sectors as the United States. That’s why I feel so passionately about Europe and Germany having to do significantly more to invest in innovation and in AI, and we can get into the details of all the ways we haven’t done that enough over the last couple of decades.
When we look at the value stack of AI, you look at the chip and infrastructure layer—the chips, and then the data centers that host those chips to turn them into compute for training and inference. Then you look at the model layer, the models that are actually producing the intelligence that all of us now need for AI productivity. Then you look at the applications that use this intelligence to build meaningful, tangible applications that bring value to people’s lives. We play too little a role in every single part of the stack, and the further down the stack you go, the starker this gap becomes. I’ll go into the details of Europe 2031 and why we think we need the most drastic political agenda in Europe that we have engaged in during peacetime, because without fixing this gap—without having a structural answer to what the tech sector is in the United States and what it is in China—I don’t think, no matter how much transformation effort we engage in, that we’re going to be able to solve for this root cause that is at the root of all the other problems compounding and growing from there.
Mounk: One thing that strikes me—to stay for a second with the car industry before we fully get into tech—is that, in general, the sense that everything is going to be worse ten or twenty years from now, that we’re on the decline, that things might still be okay today but are going to fall apart tomorrow, tends to be alarmist, and I tend to be rather skeptical of it, and it’s easily politically exploited.
I think when it comes to Germany’s economy, I worry that it really may prove to be true. It doesn’t take a crazy imagination to think that Volkswagen might go bankrupt in the next few years, that Mercedes or BMW might be bought by Apple or some other U.S. tech company that wants to push into autonomous vehicles.
While you’re absolutely right that the rise of the AfD is a whole German and a whole Western problem, I would even go beyond that, because countries like India, etc., also see the rise of various forms of populist movements. I do think one of the reasons the AfD has been stronger in East Germany than West Germany is that West Germany has a huge economic strength, which is quite striking to me. I was traveling a couple of years ago in Vietnam and I spoke to a woman who said she’d been to Germany once because she worked for, I believe, Baumarkt, and the headquarters was in some small town I’d never heard of. I thought, politely, you must be mispronouncing this—I’ve never heard of this place, I don’t believe a headquarters can possibly be in this town. I looked it up, and she had in fact pronounced it quite well. I’d just never heard of the town, because when you’re in West Germany, you’re never more than an hour’s drive—often half an hour—from a genuine center of world commerce. That’s because German companies are very spread out, and even when the very large companies aren’t spread out, the suppliers really are. That, I think, is why West Germany has held together not just economically but politically.
So what happens if Volkswagen goes bankrupt, and BMW moves a lot of its production abroad, and all of the suppliers are suddenly no longer needed, is going to fundamentally transform the German political landscape in a way that I’m very worried about, and I think that might very well happen over the next ten or so years.
But to get to the question of tech, which, as you’re rightly identifying, is the key reason Germany is so far behind—tell us a little bit about this project, Europe 2031, that you co-wrote with a few other people. It really imagines—and I recommend everybody interested in the state and future of Europe go have a look at it—how Europe becomes totally irrelevant, not just economically but geopolitically, because it’s so far behind on tech. It’s a story, in some ways, of two friends: one who goes off to San Francisco and makes his way in tech, and the other who stays in Europe and slowly climbs the greasy and not-so-appealing pole of European bureaucratic politics. It’s really about them both despairing about the fate of Europe and presenting to us the scenario of what’s about to happen. What do you think is about to happen?
Dada: We wrote Europe 2031 out of a deep love for this continent, and because we want to do everything in our power to stop this scenario from actually becoming reality. It’s a narrative scenario—we’re not super-forecasters, it’s not meant to be an actual authoritative, timeline-based scenario. But we still tried to engage in disciplined imagination by writing jointly, in a very long Google Doc, all of our different perspectives, criticisms, and nuances that in the end led us to the story.
As you rightly summarized, it’s about a young policymaker in Brussels and an AI founder from Germany who’s in Silicon Valley scaling his company. They’re chatting through text messages, exchanging and observing what is happening in Europe—he’s pushing AI from Silicon Valley, and she’s pushing AI from the policymaking perspective, trying to do the right things to diffuse the value of AI in Europe and to build AI champions of our own.
The scenario goes into three major points. First, that we massively underestimated the magnitude of AI. European policymakers particularly, but also to some extent the broader public—call them elected leaders of any sort, whether in the economy or the political realm—thought that this was going to be a little bit more like a tool, maybe like the desktop computer, or maybe like the mobile phone, when actually it was much more similar to the Industrial Revolution in the explosive force it would bring to the economy—or not bring, to the economies that are not adapting in the right way.
The second piece is that we underestimated the speed at which this technology was going to progress. I have to say, in 2026, it’s now September, and I’m feeling this part quite strongly looking at the most recent models coming out of OpenAI and other labs over just the last couple of months, which have been absolutely crazy in terms of progress. This was something Europe thought was going to happen maybe over decades, certainly over years, when really it was just a couple of years, and then suddenly a lot of progress happened in a very short period of time—massively underestimating what that would mean, both from a societal-preparedness perspective and from an investment perspective, and the urgency and type of capital mobilization then required to still play any meaningful part in this race.
The third piece we really underestimated was how important certain messengers, or certain companies, were, in consistently speaking the truth about where this was all headed—because most of those voices were squarely located in Silicon Valley, in some way or form entangled with the current politics of the United States, and associated therefore with MAGA. In that whole—we call it the “tech-brosphere,” as European media often refers to them—there’s a deep skepticism of any voice coming out of that corner of the world.
Mounk: I have my serious reservations about Elon Musk and all of that—but the way in which, just in the last few months, I see it with people I speak to, with the media, my students: the phrase “tech bro” has just become this thing where Silicon Valley, AI, is just tech bros doing stuff. It’s become nearly a slur. I don’t mean it’s offensive, but it’s become this super-easy way to dismiss the importance or relevance of any of this—let’s not worry too much about those tech bros in Silicon Valley and their games, we can ignore it because it’s just for tech bros. It’s very strange.
Dada: I fully agree. I would have hoped that maybe our piece would have shifted the narrative a little bit, but I think it’s actually grown. The “tech bro” narrative has grown—it’s not decreasing in any way or form. It’s exactly that: an immediate dismissal of those voices, seeing them as non-serious, as power-hungry people who just want to feed their own bank accounts.
I do think, to some extent, at least, the power-hungry people do want to feed their bank accounts. That can be true, and they can still be telling the truth. That’s what we go into in the piece: you can be critical of the political views, or the worldviews, or the moral fiber, of these people, but that does not give you the right to ignore what they are saying, because they are, virtually, the ones who are very much building this exponential technological wave. This is not a wave that is being pushed by a hundred thousand companies—this is very much a concentration of power in the hands of maybe a dozen companies that play massive roles in this race. They’re very much the people who are at the center of power of everything that is happening here.
At the same time, if you look at Elon Musk, for example—I’ve been very critical of Elon Musk, and written publicly about his political views, and how he’s pushing the AfD in Germany, and all the ways in which I think that ought to be criticized—still, I think he has been consistently speaking the truth, and also being a brilliant entrepreneur, for what it’s worth, in capturing the value that is being generated in this AI boom. Europe can’t say the same thing about our continent.
The piece goes into those three factors, and it basically leads to an economy that is no longer able to be competitive, where people are either blocked from using AI, or, more towards 2028, 2029 in the scenario, there’s an actual compute crunch in the world, because Europe has not built sufficient data center capacity and is very reliant on U.S. data center capacity. At the point where everything’s exploding and AI is really starting to take off and work, there’s just not enough compute, so there’s basically compute rationing, and Europe is not able to get as much as we would like in order to uphold our economic weight at the table.
All of that leads, at the end of the day, to the European Union falling apart and the European economy being sold for scraps and pieces, with the United States and China coming to the negotiating table, both making offers for various parts, almost like a cow, picking the beef strips that you want, and starting to negotiate on that. Europe is just absolutely powerless, no longer able to coordinate, to have a united front—there’s infighting happening. At that point, the scenario ends, and then we have an epilogue in 2035, where one of the characters is looking back and trying to grapple with her anger at not having done enough, asking what would it have taken for the leading class in Europe to see what was happening and to drastically change, even if it cost them their careers, even if all of those things were hugely unpopular, because Europe is already in a pretty difficult position, and a lot of these things require you to do things that, in the beginning at least, may look very unpopular in society. But people doing them anyway—not because they’re easy, but because they’re hard, and because they’re the right thing to be doing for Europe at that moment in time.
That’s one of the last pieces that was really important to us: the importance of, if you want to ask people to engage in this kind of change and to make these kinds of contributions—for example, labor market reforms are one of the contributions we think are hugely important. I live in Germany, I also invest a lot in France. These are two of the countries where making any type of organizational change is extremely expensive, to the point where it’s almost impossible for certain companies, because it’s just so expensive to let people go, to do even performance management, performance tracking. These are really difficult things. We need labor market reforms to be able to have more diffusion of AI in society, to allow companies to drive change with much more confidence, and thereby remain competitive.
But these are obviously things that society doesn’t like. So what is the story? Who’s the politician—I always ask, who is the person who’s able to stand there and have a vision that people get excited about, and say, this is why I see the light at the end of the tunnel, and why I’m willing to contribute my fair share of change in this wider sum that will then get us to higher ground? Europe right now has none of that.
Mounk: I was on Italian TV this morning, and there was the usual general chaos—I woke up to seven missed calls about where I am, because they’d moved my appearance forward by an hour without telling me. So that was a fun way to start the day. But they had a brief discussion about the lack of opportunity for young Italians, and talented young Italians moving abroad, as one of the crises for the country, which I very much agree with. It’s one of the tragedies to see how many talented young people are forced to move abroad even though they’d love to make their careers in Italy, or who stay in Italy and just see their prospects diminished in a very serious way.
But the solution that seemingly everybody on the panel proposed was to say the problem is the precarity of young people, and we need to make sure they have more secure employment contracts and that it’s harder to fire them. I wanted to scream and say, no, the problem is that you don’t have any economic dynamism—nobody wants to hire a twenty-five-year-old if you have to employ them for the next forty years, no matter what the future of your economy and your company looks like. We’re in such a malaise that some of the solutions people try to embrace are actually reinforcing the roots of the problem rather than solving them.
But I want to get back to the solutions you propose, and to the fact that you also advise the German government on AI policy. First, I want to dig into why it’s such a problem if Europe falls behind on AI. One problem is perhaps economic—there are a lot of economic benefits to the tech sector in general, and perhaps AI specifically, though there are questions about that too; perhaps Anthropic and OpenAI will prove to be economically unsustainable, etc. More broadly, there are two lines of disagreement and skepticism I want to tease out.
One disagreement is: what kind of AI does Europe need? Does Europe need frontier models, where we’re particularly far behind and it’s going to be particularly difficult to catch up? Or do we simply need to be good at creating AI-powered companies and having a lot of AI adoption? I had an interesting discussion about this the other day, on a podcast episode that will soon be released, where Louis was arguing that actually it’s fine if Europe is behind on foundation models—we don’t need those, as long as we have compute in Europe, as long as we actually apply AI in our businesses, we can remain competitive in everything that matters, and just leave the foundation models, the frontier models, to the United States or China. Other people seem to argue that this still makes you a vassal—that you’re still militarily dependent on the United States, that they can switch off access to those frontier models, that no matter how hard it is to catch up on frontier models, unless you’re doing that, you’re just giving up any hope of being an equal player in the twenty-first century.
So, to start with, where would you sit on that debate? Do you think Europe needs to catch up on actually developing frontier labs here? Or can we have this kind of free-rider attitude that we had toward military security for much of the second half of the twentieth century, saying, well, the United States is going to provide our frontier models, and perhaps sometimes we can use open-source models from China, and we rely on the fact that we’ll somehow always have good-enough models to defend ourselves, for our companies to compete, because they have access to those models?
Dada: This is a great question, and I think it deserves a little bit of nuance. Let me try to untangle it. The way I would think about it: yes, there is a big economic argument to be made—and this is just a vibes-based number—that eighty percent of the economy, of the total value creation of AI, is not going to depend on the strongest Pareto-frontier model. There are many other models that are powerful but more efficient, cheaper, maybe open source, that are still covering whatever economically valuable task I’m trying to achieve, within their intelligence distribution, so I’m able to get that done. I don’t actually need the most recent version of whatever the closed-source models from the big labs are able to perform. I think eighty percent of the economy is going to be just fine and booming, just trying to get a unit of intelligence at the cheapest price. That’s going to be a big market. In fact, the company I’m leading is also massively focused on this—this is where a lot of the term “model routing” comes in, where you actually want to be independent of the underlying models. You shouldn’t care if it’s model A, B, or C—you should just care whether it’s the best intelligence at the cheapest price, lowest latency, at any given point in time.
Then there’s the other twenty percent. I struggle to pin down that twenty percent—it may turn out to be significantly more, or significantly less—because this is where frontier intelligence plays a role, and I still struggle with valuing frontier intelligence in the right way. How do you value intelligence that goes beyond human intelligence at scale? I find that very difficult. I can’t properly do a market sizing for that, because who knows? These are things like drug discovery. These are adversarial markets—think of cybersecurity, where there’s always a better hacker, so you always want access to the best frontier intelligence to hack the next system or defend against the intrusion. These are markets like hedge funds, generally structured in an adversarial nature, where your gain is very much someone else’s loss. There’s also the securing of critical infrastructure. There are a lot of markets where I’d argue having a lot of Einsteins in a room matters enormously.
I always use one example to make this tangible: if you had access to a thousand Einsteins right now and you were a hedge fund, I think you’d be able to boost your total success pretty drastically. Versus if you were Coca-Cola, I don’t think having a thousand Einsteins would make Coca-Cola that much more successful, because it’s a big brand, a recipe that’s been around for a hundred years, a lot of distribution, physical stuff. Einstein is not going to meaningfully change your warehouse. There’s an elasticity to intelligence, where a marginal additional unit of intelligence is valued differently by different industries.
Coming back to Europe, and how to think about this from a sovereign perspective: for that eighty percent, I think it’s fine if we just have access to the current model landscape. It’s wide and beautiful—unfortunately there are very few European names, but we can use the Chinese models, we can use the U.S. models. This is not likely the part of the market that’s going to close down if push comes to shove. But for that twenty percent, I think it’s crucial, absolutely crucial. I don’t think you can honestly tell a European person that you’re able to defend their lives ten years from now, or secure their access to the modern medicine that will be available at that point, if you don’t secure access to frontier intelligence.
How do you secure access to frontier intelligence? You’ve got two options. Either you produce it yourself—but if you produce it yourself, it’s not fine to be number three, it’s not fine to be number four, it’s not fine to just do what DeepSeek does, minus twelve months. If Europe does engage in a lab effort, it needs to be a Manhattan-style, ultra-ambitious, ultra-well-funded effort, because the piece where it actually matters is being able to produce outcomes that are frontier. Looking at the recursive self-improvement loop—AI starting to improve AI research itself, which is happening at OpenAI, at Anthropic, and other leading labs—we are somewhat late in the game to build this up ourselves. But I honestly think this is such an important race that I don’t care if it’s late in the game. It’s not over until it’s over. Even if it’s the last meter of the marathon, I will still try my best not to come in last place, and try to somehow make it over the finish line.
If Europe engages in an effort, we need to do away with all of the non-ambitious approach and just make this an ultra-ambitious effort. Now, even if we were to do that—if Europe was able to get its act together, mobilize the capital, mobilize the talent—I think success would still not be guaranteed. That’s why the second perspective comes in: frontier intelligence is going to matter for the areas I just outlined, so let’s engage in our own very ambitious effort, but let’s also plan for a scenario where it doesn’t succeed. That comes down to: what are the other areas where we have value, where we own certain valuable bottlenecks, where we can increase our weight at the table so we can negotiate access to the frontier, however it’s shaped, whoever owns it at that given moment?
This is where a lot of the let’s build more compute in Europe comes in, because compute is the actual infrastructure that powers all of this. Compute matters enormously, both for training and for inference, and for the economic weight you’re able to push, how you’re able to diffuse the value of AI. So let’s do a lot more here to become a much more relevant player. Let’s also look at the very meaningful choke points, the valuable assets we control in the overall AI supply chain, because we still have companies like ASML, like TSMC, Siemens Energy is a great example, powering the entire data-center boom we’re currently seeing. We have a lot of industrial AI companies. If you look at, for example, humanoid robots, a lot of the suppliers—the “hidden champions,” these companies no one’s ever heard of, in some random German village, who are still the world market leaders producing that one actuator, that one critical piece. We have a lot of these assets, but I don’t think we have a coherent enough strategy of putting them together and using them to create leverage and increase our weight at the table. So I think we need to be doing both—securing access to frontier, or building frontier, while also recognizing that the eighty percent of valuable tasks that don’t require that also exist.
Europe, I think, is well positioned if we actually get the diffusion part right, because the productivity piece is so crucial here. One given euro of inference spend that goes to OpenAI, Anthropic, or Mistral—pick a model of your choice—can unlock three, four, five, ten euros of productivity for companies that are able to use this intelligence to increase their competitiveness in the market. That’s the economic piece that matters enormously, and there’s also the security piece that matters enormously. That’s why it’s a little bit more nuanced, and we need all things at the same time, if that makes sense. I think that would be the sensible strategy—you need to plan for different scenarios and be ultra-ambitious about pushing each of them, because we don’t quite know where we’re going to land in the end.
Mounk: That’s a super helpful way of thinking through it. If you want to have that effort, before we get to solutions, we need an accurate diagnosis of the problem. I’m struck that in these conversations I keep hearing a set of overlapping explanations, but often people don’t get on the same page about which of them really matters. One explanation is that we just don’t have capital, because we’ve missed the tech boom, so there’s a lot less capital that can be allocated to startups in Europe. A second is that there’s a cultural problem—in a narrower sense, that there just isn’t a strong enough startup ecosystem in Europe, because it’s so much stronger in the United States, or for that matter in Shenzhen or Beijing; or in a broader sense, that students here—and I certainly feel this talking to American and European students—are more likely to aspire, as your scenario implies, to be a bureaucrat in Brussels than to be a startup founder.
There’s also a set of concerns about the general weight of regulation—how hard it is to do everything from incorporating, to, in Germany, having to listen to very expensive notaries read out a contract in front of you, which Peter Thiel recently commented on, but which my friend who works in startup and VC land in Berlin has complained about to me many times before—to taxation, where you might be taxed in some European countries, not all, on the theoretical valuation of your company, even if you don’t have any of that money in your pocket yet as a startup founder, to the lack of a single European market. I’m probably missing three or four other explanations.
You have the vantage point of being a prominent commentator on this in the media, but also of being in the tech industry and investing in tech companies in Europe. What are your top frustrations, and how does that perspective inform your assessment of this debate?
Dada: It is a multifaceted problem, and all of those dimensions matter, but they don’t matter in the same way. I do think there is a very significant capital problem—not so much that Europe doesn’t have capital, because that’s not true. We have our own pension funds, big insurance companies with big balance sheets. But what is true is that, structurally, Europe does not have an equivalent to the current biggest funders of AI, which in the United States are the likes of AWS, Google, Microsoft, xAI, OpenAI—who have either access to capital they’re raising from the VC market, or balance sheets that have no singular equivalent here.
The concentration of economic power in the United States doesn’t have a structural equivalent in Europe. Our economy is actually not that much smaller—I believe the European economy is about two-thirds of the US economy—but we structurally lack a small number of players who can coordinate at very low cost, because xAI can just call OpenAI, can just call Anthropic, can just call Microsoft. People say they’re doing circular deals with Nvidia, but at the end of the day, the beautiful advantage the United States has is vast resources concentrated in a small number of players who are able to coordinate. In Europe you’ve got country borders, different regulation, different companies that have capital, but the coordination costs are significantly higher. We don’t have a single market. We don’t really even have a homogenized financial market.
What I’ve been advocating is that Europe is what it is—we won’t be able to do away with these structures tomorrow. That just means we need to make an even bigger effort from the private and public side coming together: a small number of players, the biggest corporations in Europe and the biggest governments, who also have access to very meaningful funding and to regulation they could change—for example, allowing pension funds or insurers to allocate more money to venture capital, because in the United States there’s ten to twenty times more capital allocated to venture capital by these bodies than in many European countries. A small enough number of players need to come together and engage in an ultra-ambitious effort to make up for the fact that we don’t have the structural equivalent of the United States, where most of this can just be done by the private sector and a small number of companies. We won’t have that small number of companies, but I think it can be a small enough number of companies and governments, a coalition of the willing, so to speak, that come together and engage in very meaningful reform to mobilize the capital.
The second piece—I don’t so much think it’s a talent problem. I do buy the talent piece to some extent, but there are many Europeans in the United States who would really love to come back to Europe and give it their all, because they want Europe to succeed. It’s a domino effect, a chicken-and-egg problem: they don’t see the capital side coming together, so they think, I’m not going to come and run some startup that’s going to get 200 million in funding, because I know it’s just not going to go anywhere. At the end of the day, it needs—and this is the interesting piece—an Elon Musk-type figure, someone with as much influence, power, and orchestration capability, to bring all of these puzzle pieces together. We don’t have these ultra-connected, mega-connected individuals in Europe—it tends to be a lot more fragmented, and it’s difficult for us to put all the ingredients together. The Commission is trying to do some part, national governments are trying to do some part, VCs are trying to do some part, but no single party by itself has enough power to truly get the job done.
That’s currently my biggest concern. I do think many people are starting to understand the problem, or what a solution could look like, but we’re missing the end decision-making power and the weight of someone who’s able to lock people in a room for three days, Manhattan-style, meet in the desert, and just make the deal happen. I think that’s actually the biggest structural gap we have in Europe. I think we could overcome many of the other challenges—which are indeed real challenges—but I think they’re minuscule in comparison to this much more top-down structural and capital challenge.
Mounk: So how do we overcome this? As you’re saying, there’s a large economy, a lot of Europeans who’ve saved a lot of money—but most Germans, for example, until very recently, saved money through insurance companies, taking out insurance products, life insurance products that would give them guaranteed two-point-five, three percent returns a year. They’re not directly invested in the stock market, which obviously makes it much harder to allocate capital toward the most productive and risky things. The lack of comfort with the idea of investment is still deeply rooted in the European middle class, which tends to think of it as some kind of gamble.
There are all kinds of cultural obstacles here. Presumably the only coordinating instance that exists is the state—when you’re talking about a Manhattan Project-style undertaking, presumably this would be the European Union coordinating with the member states on some kind of project. I agree with you on the need for a Manhattan Project-style undertaking. I guess I’m just very skeptical that, if led by Brussels in coordination with twenty-seven member states and all of their political demands, in a culture of people who are very skeptical about the wisdom of this in the first place, this can work.
Dada: I agree. I’m under no illusion that this would be easy. The German government is, as we speak, negotiating reforms of our pension system and the Aktienrente—the pension that could be fueled by investment into capital markets—and it’s proving a very, very difficult undertaking. So I can only agree with you that structurally it is very difficult. But at the same time, it is the right thing to do, or almost the only thing to do. Just pointing to the private sector and saying VCs and companies, go solve that doesn’t account for the structural difference between the European market and the U.S. market. The only way to fix this gap is for the public side and the private side to come together and work together.
I don’t think it needs all twenty-seven member states. I think a coalition of the willing—if you look at the size of the economies in Europe, there are a couple of economies that matter quite a lot more than others. If you focused on the top three, four, five economies—I would also love to see the UK included, even though they’re no longer part of the European Union. I actually think that some of the ministers there, the people currently leading AI strategy, have a lot of really good ideas that we could replicate—for example, looking at the AI Security Institute in the UK, or a lot of the work they’ve done around the sovereign AI fund, trying to invest very seriously in the next breakout technology that could emerge five to ten years from now, with the capital and talent required to do that properly. There are many things the UK is doing that we could be modeling in Germany or other countries in Europe.
It’s difficult, but just because it’s difficult doesn’t mean we shouldn’t try to do it. I’d also say we need to look this fact in the eye: Europe and the EU are a construct that was built in a time when stability and peace, and never going to war with your neighbor again, was the most important focus. It’s amazing that we were able to build the European project in the first place—what an achievement. As someone who has grown up in Europe, I am nothing but thankful for a lot of the energy and effort that went into the European project. However, times change. The only thing that is for certain is that if times change but you refuse to change and remain the same, the result is eventual decline. We are but one big arc of evolution, and all of us are here today because we found ways to adapt. Had we not adapted, we would have long been eradicated.
The EU, or European national governments, or Europe, is now at a crucial crossroads. We need to look the fact in the eye that we have the right principles at the core of this continent—self-determination, human dignity, equality, valuing the life not of the richest person but of ordinary citizens, equal contribution in the democratic process, and so on. These are all things that ought to be protected. But to protect them, we need to look the fact in the eye that the processes that served us well in the past are no longer fit for the times we live in. I don’t care what type of cutting of red tape, what kind of extraordinary policymaking and rule-making we now need to engage in, to be able to make change and get to progress. We need to look the fact in the eye that the current processes meant to protect what is actually valuable no longer serve us, and will, in the end, lead to us upholding the processes but losing all the things they were actually meant to protect.
The only solution I have is that things need to drastically change, and we need to find new ways of regulating and of making change happen in disruptive ways—but ways that I believe will actually serve us well in this overall transformation we’re currently going through. I don’t have perfect solutions, but I can only say: it’s hard, but just because it’s hard doesn’t mean it shouldn’t be attempted, or shouldn’t be done.
Mounk: I feel like we’re ending on probably a conscious, perhaps unconscious, allusion to my favorite novel, and the line from The Leopard that for everything to stay the same, everything will have to change. I very much share the same sentiment about Europe today.
If you have not yet signed up for our podcast, please do so now by following this link on your phone.
Email: leonora.barclay@persuasion.community
Podcast production by Jack Shields and Leonora Barclay.
Connect with us!
X: @Yascha_Mounk & @JoinPersuasion
YouTube: Yascha Mounk, Persuasion
LinkedIn: Persuasion Community