Part of what makes Daron Acemoglu exceptional, as an economist, is the scale of the questions he is willing to ask.
In 2024, he won a Nobel Prize, alongside fellow economists Simon Johnson and James A Robinson, for their work on how institutions shape prosperity, and how they in turn are shaped by the motives of those in power (motives that often involve profit). Their research took the shape of books such as Why Nations Fail (2012; co-authored with Robinson) and Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity (2023; co-authored with Johnson). Acemoglu, 59, explores some answers in What Happened to Liberal Democracy?: Remaking a Politics of Shared Prosperity. Power and Progress: Our Thousand-Year Struggle Over Technology and Prosperity (2023; co-authored with Simon Johnson) and What Happened to Liberal Democracy? both emphasise that inclusive institutions also need to steer technology in order to ensure they are consistent with shared prosperity. * Why have such institutions traditionally been so scarce?
His newest book uses a similar academic framework to explore the future of liberal democracy, and how it might be shaped by the future of work and AI. Acemoglu has been in the news lately, for a bit of a back-and-forth with The Economist about the nature of some of his views. There is debate over whether he underplays the role of colonialism; whether his theories account for the rise of China. What all sides can likely agree on is that a couple of fundamental questions now promise to shape our future: Can human ingenuity be directed towards a greater shared prosperity, or will it simply be another mechanism for concentrating power? What happens to liberalism if the outcome turns out to be the latter? “Liberalism cannot survive without shared prosperity. As we advance into an AI-led world, what does this mean? And those choices matter greatly, both for the future of shared prosperity and for the promise of liberalism. Absolutely not. People need to be able to look each other in the eye, regardless of their situation in society. This is not possible in a technocracy. Technocrats would have far more power over society. Moreover, self-government, which is equally crucial for liberal democracy, wouldn’t make sense in a technocracy. * Your chapter on liberalism’s dilemmas is very compelling. The first of these is intolerance. But is critical debate even possible in an algorithm-driven world? It absolutely is. But not when those algorithms are controlled by the most powerful corporations humanity has ever seen. This is why I argue, in the book, that we have to regulate social media and AI. The internet and social media could have been much more consistent with democratic discourse. That’s what many people hoped in the late 1990s and the early 2000s.
Because these structures depended on the participation of the working classes — do you think liberalism can make room for itself in a non-utopian technocracy, excerpts from an interview. * Given that liberalism had its greatest success when it had the support of key economic structures such as industrialisation — and given that it had that support. Because compromise and dialogue are impossible online, the reason that did not become a reality isn’t. Because leading companies decided that wasn’t what gave them profits and power. * You’ve said that good institutions can serve as game-changers, it is.
It cannot survive without the dignity of workers,” he says. AI could help, if it’s governed right, Acemoglu says. Like other transformative technologies, it “enables multiple feasible paths,” he writes. But, “we should not presume that when it comes to such transformative technologies, the “market”—or, in practice, a handful of tech companies—will make the right choices. Democracy and liberal democracy in particular require that people are not “dominated” by others in terms of economic or social power. How would you define a “good” institution? “Inclusive institutions” are those that provide incentives and opportunities for people to participate in economic activity on equal terms, engage in building human capital and undertake innovation.
MORE ON ACEMOGLU * Daron Acemoglu, 59, is currently Institute Professor of Economics at the Massachusetts Institute of Technology (MIT). He joined MIT as an assistant professor in 1993. In 2024, he won the Nobel alongside fellow economists Simon Johnson and James A Robinson, for their work on how institutions shape prosperity, and how they in turn are shaped by the motives of those in power — motives that often involve profit.
Building inclusive institutions is hard work. This isn’t regulation of AI in a defensive way. It’s more about building a consensus on how society at large can benefit from AI, and then building institutions to achieve that. I believe so. But there is so much uncertainty about the capabilities of AI that I cannot afford to rule out that we will head toward something like artificial superintelligence (ASI), in which AI models, together with robotics, can perform every task and function that humans can; in which case pro-worker AI would be much harder or impossible. But of course predicting the future of AI’s implications for inequality isn’t like weather forecasting. I explain in the book why pro-worker AI could actually be profitable for companies as well. What would you change about how it is currently defined? Yes, absolutely. Liberalism still has the best ideas for our future and provides the best environment for us to deploy collective knowledge for our betterment. But liberalism cannot survive without shared prosperity. Liberalism cannot survive without the dignity of workers. I think the most important element that I would emphasise is the relationship between liberalism and community. A fundamental mistake of current liberalism has been to sideline community. * What do you think we stand to lose, if liberalism fades? Many of the poor and working classes, for instance, seem content with more-closed societies, if these reflect their identity and protect them better. Who then are we fighting for liberalism on behalf of? Under different conditions, they were in the past, and can be again in the future, much more supportive of liberal ideas. . ALSO CHECK OUT: In what ways has social engineering been imposed on the working classes? What role, for instance, has education played? Click here to read an excerpt from What Happened to Liberal Democracy? . His research, over decades, has sat at the intersection of economics, history and politics. * Born in Turkey, his father was a lawyer and lecturer at Istanbul University; his mother was a poet and principal at an elementary school. He first became interested in economics in high school, as he began to wonder about the links between dictatorship, democracy and economic growth. * Acemoglu studied economics at University of York in England, and then at London School of Economics.
I do not believe artificial superintelligence is imminent, which means there is room for pro-worker AI. Because pro-worker AI would have to be more domain-specific and more dependent on high-quality, often proprietary, data, which would necessitate a more decentralised model of AI. * Do you think it would help to redefine liberalism, but it probably wouldn’t be as profitable for big tech firms,. Because they have experienced liberalism in the form of social engineering imposed on them, and they saw liberalism as indifferent to their plight, i believe that the working classes have this reaction.
But more than that, many are designed by elites in order to be “extractive”, so they get an advantage and monopolise resources and opportunities in the economy and have excessive power in politics. * Can inclusive institutions compete with machines that evolve as quickly as AI is doing? Yes, and this is where I would call for a “proactive governance agenda for AI”. The centrepiece of it is what I call “pro-worker AI”. * Is pro-worker AI a realistic goal?
The future of AI depends on decisions that we as a society, and that the tech sector, make today.
