CEOInterviews.AI
Start App
Get an email when Daron Acemoglu gives a new interview
One email per new interview. Unsubscribe in one click.
Daron Acemoglu
Economist and Nobel Laureate, MIT

AI, Democracy & the Future of Prosperity

📅 Nov 19, 2026 Institute for Information the Internet & Democracy 60 MIN 3 VIEWS 35 SEGMENTS · 2 SPEAKERS
AI, Democracy & the Future of Prosperity with Daron Acemoglu, hosted by the Institute of Information, the Internet & Democracy at Northeastern University November 18, 2026 Daron Acemoglu—Institute Professor at MIT and recipient of the 2024 Nobel Prize in Economic Sciences—joined us at Northeastern for a conversation about artificial intelligence, technological change, and the future of liberal democracy. In his new book, What Happened to Liberal Democracy? Remaking a Politics of Shared Prosperity, Acemoglu examines how the rapid development and deployment of AI is reshaping work, economic p...

Questions asked in this interview

12
  1. 28:30Joe, you gonna call on them or?
  2. 28:48What are the new concepts that we need as social scientists to start figuring out is AI doing the thing that we want it to do in terms of pro-worker AI?
  3. 32:05So where do you see us going in terms of building the trust and building communities, and also where do you see us going in terms of actually building AI models that can be more democratizing rather than centralizing?
  4. 37:54What are some key values that you suggest we should leverage our sort of future changes and legislations based off?
  5. 38:29Do you think that this is something that needs the nuclear bomb treatment, where we have a global coalition to try and restrict AGI usage, or is this something where just domestic policy is enough?
  6. 43:40And so how can that happen, and who participates in creating that solution?
  7. 44:06If not, how do you differentiate yourself?
  8. 48:13So how would an alternative work, and how can you convince the institutions to switch from a profit-driven motive?
  9. 48:56And if yes, is there any way that Western democracy can evolve to improve the institutions?
  10. 52:48And my question is, do you think liberal democracy can truly recover while economic inequality keeps growing, or do we need to solve the inequality first?
  11. 53:10And why won't we have that today?
  12. 56:06I just wanted to know that if AI supposedly became software and kind of, you know, took over the government by chance, what would be the implications or pros that AI would have?
John Wibby 0:09 ↗
Hi everyone, welcome. I know everyone is still getting seated and getting food, which is wonderful, and a full house today. So that's terrific. My name is John Wibby. I'm a professor of media and communications technology in the College of Arts, Media and Design. I also serve as the assistant vice chancellor for strategic AI initiatives. And I'm also a big fan of Professor Acemoglu, who's kindly come from across the river to give us some wisdom and some guidance. But before we formally introduce Daron, I want to just talk a little bit about the sponsorship of this event and all the people who have come here and put it together. So first of all, I'm also a co-founder of Northeastern's newest institute, which is the Institute for Information, the Internet and Democracy, and my co-founders are David Lazar, Professor David Lazar, professors Rebecca Tromble and Laura Eden. It's an interdisciplinary institute that's really focused on the digital and information environment. We hope to do more events like this. This is somewhat of our sort of hello world event. We're kind of doing a gradual ribbon cutting, but if you're interested, please come see me after the event and happy to talk about that. The president's office has also been instrumental in helping us pull this together. Shout out to Mike Armeni, the senior vice president. Also Leila and Maggie who are on the advancement team. We also want to thank the John S. and James L. Knight Foundation, which is our big supporter in the new institute effort, as well as the office of the provost. And I especially want to thank our what we call Triple ID, that's the name, the shorthand for the new institute, our team, Alisa Zmer and Libby Phillips over there, who have helped pull us all together. So today's show is going to run until 1 p.m. roughly. And Daron's going to speak for about 30 minutes, which I'm really excited about, because if you've been following sort of his intellectual journey over the past few months, you know that around the world, people have been debating his ideas. And so hopefully you not only have a great lunch but a great intellectual lunch. And so we're going to also offer the opportunity to speak with him, to ask questions, and we'll do that for about half an hour. Today's event is being recorded, photographs and whatnot. So just heads up on that, that you might appear in photos or video. Please silence your phones. We are raffling 15 signed copies of the book, of the new book. So you'll receive an email around one o'clock if you got a copy, and I think a lot of them are signed, if not all of them. So that's pretty cool. In any case, Daron Acemoglu is an Institute Professor at MIT. He's a faculty co-director of the James M. and Kathleen D. Stone Center on Inequality and Shaping the Future of Work. He's an elected fellow of the National Academy of Sciences and the American Academy of Arts and Sciences. You may know about his previous books, co-authored with James Robinson, Why Nations Fail, or Power and Progress with Simon Johnson. This is his newest and maybe his most important. We will see what history shows. His academic work covers a range of areas including political economy, economic development, economic growth, etc. Among his many distinctions, he won the John Bates Clark Medal in 2005 and in 2024 he won the Nobel Prize in Economic Sciences along with Simon Johnson and James A. Robinson. So we have a Nobel Prize laureate here, which is amazing. You know, here at Northeastern we are really trying to figure out what the future is with respect to AI technologies and the labor market, and we have a whole bunch of commitments both in terms of pedagogy but also entrepreneurship, and we're doing a whole range of things. Now we can do awesome stuff here within the university, but of course the world is moving very fast too, and so I thought it would be terrific to have Professor Acemoglu over, who's probably the world's leading expert trying to think through these issues of how AI is going to fit into the global economy and help us think through how exactly do our efforts here on campus, which include lots of experiential AI commitments, faculty capacity building, we are revising courses as we speak, we're doing all kinds of things, but that world is changing very fast and he's going to help us map that today. We need analytical rigor, we need wisdom, we need guidance, and I think we're going to have that today in spades. In particular, I just want to point out one thing, an idea that caught my eye and actually made me reach out, which he's been talking a lot about pro-worker AI. And I think pro-worker AI is a really interesting conceptual framework for us thinking about how this technology is going to reshape the economy. So with all of that, I'd like to welcome Professor Acemoglu and the floor is yours. Thank you.
Daron Acemoglu 5:15 ↗
Thank you very much, John. Thank you. Thanks everybody. It's my real pleasure to be here and fantastic to see so many of you interested either in the food or about talking about liberal democracy. As John very kindly said, this is my most recent book and in fact it's the book that I have written with the greatest urge and in the shortest amount of time among all the books I have worked on for one very simple reason. I have a deep belief that liberal democracy is one of humanity's greatest inventions. It is unthinkable for me that we could be living in such complex, huge societies made up of tens of, sometimes hundreds of millions of people, very, very powerful states and armies, but still we have laws and, by and large, state institutions and norms that support several aspects of individual freedoms, try to create opportunities for people, strive for things like limited inequality, helping the discriminated or the disadvantaged in society. None of these, in my opinion, would be feasible if liberal democracy and its ideals went away. And unfortunately, I believe that liberal democracy is in a deep crisis. Moreover, for those of you who may be tempted to think liberal democracy is in a bit of a difficult state right now because of attacks from various extremist, anti-establishment, anti-liberal ideologies and movements, yes, that's true. But I think the strength or apparent strength of those movements in Europe, in Latin America, in the United States, around the world is just a reflection of the problems that liberal democracy itself is having. So its internal problems, and therefore it's not enough to defeat these movements at the ballot box once. But we do need liberal democracy and liberal ideals to be revitalized. And that's my aim in this book. Not to propose solutions per se, but hoping to contribute to that debate. In a nutshell, the argument of the book is that liberal ideals, especially when they fused into other movements, including those inspired by socialists, led to the left corner of the liberal tent in the form of liberal democracy, Roosevelt's New Deal programs, Labor Party in the UK, and they spearheaded rapid, rapid changes in norms and institutions in line with broad liberal principles. And those principles are even more necessary for the 21st century, for the age of AI, and for the age of even more complex and more diverse societies. One proof of that comes from philosophical arguments, but I'm not going to go into those given that time is short, though I believe they are important and that's why I did spend, you know, the first 20, 25% of the book on those, but they also come from liberal democracies' successes in the past. Again, simplifying it a little bit, I argue in the book that liberal democracy promised three sorts of things to people in exchange for legitimacy and support: shared prosperity, meaning economic growth, economic improvements, which were spread out evenly across society, no group left behind, no single group or minority or majority enjoying all of the fruits of economic growth; public services, starting from a point in the early 20th century or even at the end of World War II where the state of public infrastructure in Europe was completely pitiful, especially for more disadvantaged people in minor cities, in working-class neighborhoods, in much of the country in the United States also, public services that provide security, health, education, communication, infrastructure, and key opportunities for people to participate in society. And of course, critically, as the democratic part of liberal democracy emphasizes, voice in public affairs for people of all sorts, culminating to what I call self-government, meaning that people themselves decide at the local level, in their communities, and then national level how they should be governed. These are very lofty ideals. But by and large, liberal democracy for the three and a half decades following World War II, and to some extent even before World War II in places where it was practiced, it did deliver on these. Not perfectly, but to a remarkable extent. Public services improved tremendously in the 20th century, especially for more disadvantaged members of society. And not just in the United States and in Europe, but even in other parts of the world. Shared prosperity especially is a remarkable achievement in view of the fact that we always talk about how inequality is an inevitable part of market economies and there is some truth to that. But if you look at, for example, the United States, but the same is true in the UK, in Canada, in much of Europe, especially in Nordic countries, in Germany and France, economic growth was rapid from around 1945-46 to the end of 1970s despite the oil crisis, and that was associated with actual declines in inequality. Real wages increased very steadily for all types of workers. They increased faster than the pay of managers or profits. The wages of workers without a college degree rose faster than those with college degrees. Very, very different situation. And people participated in politics in various different ways, especially starting in their local communities and in their associations. Trade unions were a very critical part of the democratic governance. So was local communities governing themselves and national politics. But today you see a very different picture. Inequality has skyrocketed in many countries, especially in the United States. Not just that the gains have been distributed unevenly in a much more jarring pattern that I think has no parallel during peace times anywhere else. Real incomes, real wages of about 50% of the population stagnated. Some segments actually declined as other groups, especially those in professional occupations, technical occupations, engineering occupations, those with college degrees, especially those with post-graduate degrees, rose rapidly. Public services became much harder to provide and to satisfy the needs, and people have become very much convinced that they have no voice. Those are all aspects of liberal democracy's crisis. Inequality or affordability are on the lips of everybody. People continuously complain about health care, rightly, about the state of schools or infrastructure, and respondents in survey after survey, they say that they have no voice, they have no representation, that politicians don't represent or think about people like them. And in fact, there is also statistical evidence to suggest that they're not entirely wrong. Representatives and senators in the United States vote much more in line with the preferences of their richest constituents rather than those who are lower income, and there are a number of other statistical irregularities. So what happened? What happened to liberal democracy? Is it that the liberal democratic project was bound to fail from the beginning, as some conservative thinkers have argued? I don't think so. I think we can trace the diagnosis, the roots of the failures, clearly to a single phenomenon which has then impacted politics and the mismatch between aspirations and reality in various different ways, and that's what I would call the rise of post-industrial society. The economy and politics surrounding it were very different when liberal democratic institutions started to take shape. You can see that in the origins of social democracy, for example, in Nordic countries, Denmark, Sweden, Norway, all around industrial production, trade unions. In the United States, with FDR's New Deal coalition, where labor was a central part of it, but it was also a broad coalition. Local communities and local governance was much more important, trade unions were much more important. But most critically, in the industrial economy, there was a clear pathway from economic growth and investment for firms to shared prosperity. Firms needed workers, and in the framework created by trade unions and democratic oversight, that meant that they had to pay workers more if they wanted to get them. They couldn't coerce them. They couldn't shortchange them easily, and as a result, economic flourishing for companies went hand in hand with greater labor demand, leading to faster employment, faster wage growth. The car sector, for example, in the United States in the first half of the 20th century is a brilliant example of that. They were so hungry for workers that Ford and GM sent representatives all throughout the country to every state trying to recruit workers with promises of stable employment, high wages, good careers, etc. Post-industrial society completely changed that in two ways. In post-industrial society, economy became much more globalized and it became much more digitized. Both of these severed the link between companies flourishing and the prospects of workers, especially blue-collar workers and white-collar clerical workers in mid-level, low-level occupations. Now companies could automate work with robots, with dedicated machinery, with software systems, and now with AI, cut employment, cut labor costs, and still increase profits. So there was no link. In fact, sometimes there was a negative link between their expansions and profits and the wages of the bottom half of the population. Same thing with offshoring, same thing with globalization. But that's not the end of the story. Post-industrial society also fundamentally changed the nature of politics. Two aspects of that are particularly worth emphasizing. One is that the communities that I emphasize, which were always a critical part of liberal democratic institutions and liberal ideals, although we have forgotten much about them, were mixing grounds. So were firms, companies like manufacturing firms, GE, General Motors, etc., employed workers of all different skill levels. Neighborhoods were mixed, and associations from churches to baseball clubs and to other local organizations were also mixing grounds. With the rise of post-industrial society, the number and aspirations of the college graduate, people like us, started changing, and their lives became much separated from the rest. Not everywhere, not for everybody, but as a general rule. So in the last 50 years, we've seen much greater neighborhood segregation between educated and less educated segments. The college-educated today befriend other college-educated, postgraduates befriend other postgraduates. They marry within their groups. Their values have diverged between what I would call working-class communities that don't have college degrees or that don't have a majority or a significant fraction with college degrees. So there is much less mixing both idea-wise and practice-wise between these key groups. But even more importantly, as the economy was changing, there were new strains on the livelihoods of blue-collar, non-college workers and the value divergence that I have highlighted. The face of modern politics in all industrialized nations changed. The working-class parties that I have highlighted, like the Social Democrats or the Labour Party or the Democrats in the United States, used to receive the majority of their votes from the working class. The college-educated, being much richer and a small select group, voted more for the center-right in all of these countries. In the last 50 years, starting, for example, in the 1970s in the United States, that picture started changing. Today, center-left parties receive an overwhelming majority of their votes from the college graduates, and working-class communities support the center-right or increasingly the hard right, a complete change. And the center-left parties, being the drivers of many of the aspirations and the institutions of liberal democracy and its ideals, now started changing their character. In that process, they became much less attuned to the plight of the working class and much more driven by their internal problems. The kinds of coalitions that emerge, for example, between more left-leaning versus more libertarian ends of the college spectrum, supporting cultural politics but not agreeing on any of the economic policies. There is much more to be said about how post-industrial society changed the economy, changed politics, changed the nature of cultural struggles. None of this, I think, was an entire surprise. Many of the issues that became central, such as questions around inclusion, intolerance, education, discrimination, were real problems in countries like the United States, the UK. But the nature of the way in which cultural politics emerged in a much more top-down manner was again a child of post-industrial society.
But I want to leave plenty of time for discussion and your questions. So let me conclude in the next three or four minutes by outlining what I see as the way out. The way out is not to go back to the industrial economy. That's not possible. The way out is obviously not to abandon liberal democracy or liberal ideals. I think that is a dark place. One thing we've learned over the last 15 years is that whatever the promises of the anti-establishment parties are, their practice is not going to satisfy the people because in fact they don't have any solutions. Worse, they don't have the internal balances, checks and institutionalization, and they don't have the ethical standards to even uphold minimal degrees of control of corruption or good governance. So if we are left with these alternatives, our situation is dire. What can we do? I think the best hope is to revitalize liberal ideals for the 21st century, for the age of AI. This is what I call in the book working-class liberalism. I use that title precisely because today working classes and liberalism have been alienated from each other. But I argue, I believe firmly, that liberal democracy cannot stand but with the working class at its center. So it has to gain back the working class. That has to have three legs. The first one is shared prosperity. Jobs, wage growth, careers for people of all skills. And indeed we actually need all skills. I'll argue that very briefly as well, that despite the AI tsunami, human skills are still essential, especially human skills that electricians, plumbers, blue-collar workers, office workers bring are still going to be needed in much of the economy even if some occupations do get eliminated. So the question is, can we find a way to support flourishing for these groups economically? Second, community. Community needs to become central to the liberal agenda again, starting with local governments. And then third, a reaffirmation of individual liberties and self-government both at the community and at the individual level, which must start by creating a bigger tent than cultural litmus tests, because I don't think it's possible for the center-left to gain back the working class with its cultural litmus tests. But that's also consistent with liberalism's foundational ideas. I argue why it's not just important to have individual liberties, but it's also important to have self-government at the community level as part of the experiments from which we learn to develop new technologies, but much more importantly, new norms and new institutions for governing our changing society. All of these may appear far-fetched given our current situation: social media having destroyed community even more, inequality having skyrocketed, and cultural politics ossified. But I don't think that's the case. There are many pathways for us to achieving these objectives. And let me just end by mentioning the most important one perhaps, which John already highlighted, which is shared prosperity in the age of AI. Indeed, if AI goes in the direction of automating many, many, many jobs and sidelining workers, there is no way you can form shared prosperity. You can perhaps create transfer programs such as universal basic income or universal dividend, but they're going to be just patchworks. There is no substitute for creating well-paying jobs. But AI, despite what we're always told, is not a natural automation technology. Human intelligence is very different than machine intelligence. And it's only one particular agenda to have machine intelligence mimic human intelligence under the guise of AGI, so that at the end we become much more dispensable. Instead, what I have argued over the last 10 years in my work, academic work, and in some of our other writings, and I do in the book, is that there is a very clear and very feasible pro-worker AI direction, which means that we use AI as an information technology to expand the expertise and capabilities of workers, enable them to perform new tasks so that technological change increases productivity indeed as promised. But it doesn't sideline workers. It creates new jobs, new more exciting possibilities for workers. The problem is we're not devoting any resources or enough resources to that. It's not part of the agenda. There is literally close to a trillion dollars going into AI investment, most of it to data centers supporting the AGI agenda. So it does require a redirection of AI. It does require a very different public conversation around AI. And I think it does require democratic oversight of AI. But the most important thing is that neither the separation between the college-educated and the less educated, nor the cultural divides that have made our politics much less pleasant, nor the decline of community supported by social media and other technologies, nor the relentless increase in inequality supported by automation and AI, none of these is inevitable, and we have a narrow window while democratic governance is still alive. But during that window, we do have the power to start the foundations of a revitalized liberal democracy. Thank you very much, and I look forward to your questions and comments.
John Wibby 28:30 ↗
Joe, you gonna call on them or?
Yeah, I think we're gonna have them come up. I'll, I'll, um, I'll thank you so much, by the way, for, for a brilliant talk, and I think there are many people who are quite eager to read What Happened to Liberal Democracy if you haven't already, and we're gonna have a signing over there. But first, we're gonna engage with a bunch of questions.
Daron Acemoglu 28:46 ↗
They're already signed. The books are signed.
John Wibby 28:48 ↗
Okay, great. So you are the most cited economist in the world. You have the ear of an incredible number of policymakers who are eager to figure out, like, what should we do. In one, I think in Power and Progress, you came up with this idea of machine usefulness as a metric, and we have a lot of social scientists and computer scientists in this room. Maybe I'll ask, I'll start with a wonky question, and I'll let the students ask about, like, personal finance questions since you're the most cited economist in the world. What should we be measuring that we're just currently not measuring? What are the new concepts that we need as social scientists to start figuring out is AI doing the thing that we want it to do in terms of pro-worker AI? And with that, I hope everyone will start lining up with questions, and please be bold and ask tough questions of our Nobel laureate here.
Daron Acemoglu 29:37 ↗
Well, that's a great question, John. You started with a tough question yourself because right now we do not have a clear, unambiguous metric that would say this is a pro-worker technology, this is an anti-worker technology. The broad idea is that technologies that automate could still be useful sometimes, but they need to be coupled with technologies that create new opportunities and new jobs. Measuring those, I think when you see them, you recognize them, but measuring them with metrics is not always easy. One way to do it is, which we have data on this but not in real time, we have several sources of data that starts with the O*NET, before then the Dictionary of Occupational Titles, and now various different data sets like LinkedIn and vacancy postings. They have a lot of information about what kind of skills, what kind of tasks different jobs involve, and we can measure when they are being expanded, when they are being contracted. But I think we do need even better measures. It's like, but it's the same thing, like for example the whole sort of revolution around supporting renewable technologies would not have been possible without having a measurement framework where we, for example, measure the carbon content of technology. So we need something like that. One other thing, but again, it's not a perfect measure, is that when you have, what this is what economic theory predicts and empirical evidence supports it, is that when you introduce automation technologies, the labor share of value added declines, and when you introduce new tasks and more pro-worker technologies or human-complementary
technologies, labor share increases. But that's not a perfect measure because labor share also increases due to bargaining or declines due to bargaining or other things related to trade. So that's not a perfect measure, but we might for a while need to look at a dashboard of measures and then try to develop better measures. But also some things are very clear. I think whatever confusing things we hear, the AGI agenda is clearly an automation agenda because the AGI aims at making machines better than humans in almost all tasks, better than the best humans. And once that happens, why are you going to use humans? Why are you going to pay them much? So that is a clear automation agenda. The AI industry sometimes says we're going to create new jobs, but where is the new expertise and new capabilities that workers are going to use in their occupations that are going to support that? Thank you.
John Wibby 32:05 ↗
Thank you, and thank you for your talk. I largely, broadly agree with a lot of what you said. And I do see kind of another world where we did do AI in a more pro-democracy way, and I'm very curious to hear a little bit more about how you think we can solve some of the present challenges before us for that. Some of the specific challenges that I'd love to hear you elaborate on are firstly trust and community. I agree that emphasizing bringing back more to our local communities and building communities, whether it's local or otherwise, is an important part of how we are going to protect liberal democracy. But AI is also largely anathema to community building because of how it erodes trust within the community, within institutions with authority, when we consider how it's so easy to create deep fakes and misinformation and these things. So that's one question about community and trust. And then the other part of this is that currently AI is a very centralizing technology. It is controlled by these large corporations. It requires massive resources to train these models, and many people kind of believe they are unbiased when in fact they are very biased, and it requires a certain level of understanding to interact with these models in a way that accounts for and acknowledges the inherent limitations of them. So where do you see us going in terms of building the trust and building communities, and also where do you see us going in terms of actually building AI models that can be more democratizing rather than centralizing?
Daron Acemoglu 33:54 ↗
Okay, fantastic. Great questions.
John Wibby 33:58 ↗
Oh, sorry. Hello, I'm Kylie Beimis. I'm a professor. I'm an assistant teaching professor in the Khoury College of Computer Sciences.
Daron Acemoglu 34:05 ↗
Great. Thank you so much for those excellent questions. I'll answer those questions. After this, perhaps I can take two or three at a time to economize on time, but this is an excellent question. So yeah, if I had more time, I would have put the amazing decline in trust in institutions, in each other, actually, as part of the crisis of liberal democracy. I think it isn't a necessity, however, it is a consequence of our institutions and our aspirations not being realized, our institutions not rising up to the challenge. I have academic work, and I would like to do more, showing that trust in institutions is very much related to whether institutions actually deliver what they promise. So when they do, they create demonstration effects, and those demonstration effects start building up trust. But of course, there's a catch-22 here. When you don't trust institutions, it becomes much harder for institutions to function. The same thing with community. Now, for community, however, I'm very pessimistic in some aspects, but I'm not as pessimistic as you are on AI. I think AI's challenge to communities can be overcome. I think the one that's even more challenging in my mind is social media, because I don't think the problem for communities and the problems for trust is so much deep fakes. It's disengagement. And what social media has done is that it has made us disengage from our real social networks. If you look at the data, Americans, for example, have many, many fewer close friends. They spend one-third of the time that they used to with their friends. And that's not AI, that's social media. So I think it's really a twin challenge. We have to rein in social media's effects and create a social media environment that aspires to do the things that people believed was almost definitional for social media in the early 2000s, like pro-democracy, broaden access, etc. So those, I think, are not going to be easy, but I'm not as worried about deep fakes as reining in social media. In terms of centralizing, again, that's another thing I would have discussed, and I do discuss in the book, that AI is centralizing in two ways. One is the way in which it's financed, and it's led by three companies to a broad extent. The other one is that the architecture of AI aims to centralize all knowledge, so you have a single oracle or two oracles that compete. And again, none of those are inevitable. This is redirection of AI. But there is one little hope there too. Despite the amazing capabilities of the foundation models that are true, and I never really signed up with the view that they are stochastic parrots. There are real capabilities, but you don't see them having much of an impact on the economy at the moment. Why is that? Because we lack applications. We don't have good, easily usable applications that embed the right kind of expertise, and so the application layer is going to be key, and that's where I think centralization could be broken. Ideally, we would like to break the centralization at the foundation level as well, and that's going to be much harder because of the resource cost that you've very rightly pointed out. But I think at the application layer, who owns the data, where the data is, where the expertise is, I think those are going to matter a lot. So even for developing economies, actually, which you might think they really have the raw deal here, they're going to be left behind. Well, many of these economies can start producing, working towards their own applications that are more appropriate for their own conditions. Again, it's not an easy thing, but it's at least a feasible path, and that's the way I would see the small silver lining or small hopeful path there. Thank you.
John Wibby 37:54 ↗
Hi, professor. My name is Nada Mogadam. I'm a senior studying economics and linguistics. I found this interesting, it makes us think about society in a very different way. My question is, one of your solutions is to gain back the working class and in that reaffirm self-governance within society. What are some key values that you suggest we should leverage our sort of future changes and legislations based off?
Daron Acemoglu 38:23 ↗
Thank you. Thank you. Let me take one more question and then I'll answer those two.
John Wibby 38:29 ↗
Hi, my name is D. Jane. I'm a political science and economics undergraduate student. My question is something that I've heard a lot in the media around the restriction of AGI. In part, I think some aspects of it feel like fear-mongering and some aspects feel very real, where the argument is that if we don't maximize our AGI abilities, like another country or somebody with bad intentions will maximize their ability to use it. Do you think that this is something that needs the nuclear bomb treatment, where we have a global coalition to try and restrict AGI usage, or is this something where just domestic policy is enough?
Daron Acemoglu 39:15 ↗
Thank you. Thank you. Let me start in reverse order since you just asked that question. You know, I think that's also the China question that you're referring to, and I sometimes joke that if China did not exist, Silicon Valley would have had to reinvent it. So it's been an extremely useful argument for acceleration and lack of regulation in the United States. And of course, it is true, loss of control of AI systems is a real threat, absolutely. And that loss of control includes malicious actors controlling it, using it for either national or private gains. So we have to be very cognizant of those. But when you look at AI, you need two things. Definitely you need this redirection, but you also need global governance. And it's impossible without some sort of collaboration or coordination between China and the US. And look, there is actually plenty of reason to do that. I've mentioned in response to the previous questions the importance of applications. And if you look at where applications are most fruitfully being developed right now, it's China. China is far ahead of the United States in trying to integrate AI and robotics into the production process. They have much more experience with that because of their large manufacturing bases, much more digitally aware consumer workforce and e-commerce structure. And they are very behind in certain aspects of chips and some aspects of foundation models. So there's a lot actually for both countries to learn from each other, and whether that can be done within the context of security rather than a race to the bottom, I think those have to be worked out. But again, we did it for nuclear weapons. Nuclear weapons are very different, but the problem of governing them internationally was very challenging. So I don't think that's an impossible task. In terms of the previous question of community, yeah, I think your question is very apt, and indeed I completely ignored all of the difficult problems of bringing community back and broadening the tent because the details are much harder. And one issue is that the broader tent requires us to respect individual liberties, but at the same time that is a two-edged sword. You have to respect community and communities' aspirations for self-government. But what if those communities themselves are violating other people's liberties? That creates a real paradox that we have to grapple with. There is no way of avoiding that paradox. I think what we have done over the last several decades is pretend that we can create liberty by not allowing people to govern themselves according to their views. That's completely false. That's a very top-down approach. And liberal democracy, I don't think, can survive that kind of top-down approach. So the way that I argue in the book is that we have to seek compromises, but we have to then set certain boundaries, and those boundaries have to stop at protecting individual freedom of speech, freedom of expression, freedom of religion, and freedom of association. So there has to be a minimal set of liberties that are protected. And then if communities then practice other types of things that we may or may not like, they have to be given the right to do that self-government within the constitutional framework that the rest of the nation chooses. And there are no black and white there. There's going to be a lot of gray areas. And how are we going to sort those out? Not with a sort of a constitutional rule that says this is right, this is wrong. That's another conceit that somehow there is a social contract that tells us everything what's right, or there is an Immanuel Kant-style categorical imperative that says this is the right ethical behavior, this is not. I think liberal democracy is a work in progress, or liberalism in general is a work in progress, and it gives us a general framework, and then we have to work out our own compromises as a community, as a nation, to navigate those gray areas. Thank you.

15 more exchanges in this transcript

Sign in free to read the rest of this interview. No card required.

Sign in to read the full transcript

Cite this transcript

APA, MLA, BibTeX
APA

Acemoglu, D. (2026, November 19). AI, Democracy & the Future of Prosperity [Interview transcript]. Institute for Information the Internet & Democracy. CEOInterviews.AI. https://ceointerviews.ai/interview/3351466/

MLA

Daron Acemoglu. "AI, Democracy & the Future of Prosperity." Institute for Information the Internet & Democracy, 19 Nov. 2026. Transcript, CEOInterviews.AI, https://ceointerviews.ai/interview/3351466/.

BibTeX
@misc{acemoglu2026_3351466,
  author       = {Daron Acemoglu},
  title        = {AI, Democracy \& the Future of Prosperity},
  howpublished = {Interview transcript, Institute for Information the Internet \& Democracy. CEOInterviews.AI},
  year         = {2026},
  month        = {nov},
  url          = {https://ceointerviews.ai/interview/3351466/},
  note         = {Speaker-attributed transcript with timestamps}
}