Daron Acemoglu

Such an incredible publication day for #PowerAndProgress, thank you all. This is a message from my co-author, @baselinescene, and me. You can now buy the book here: shapingwork.mit.edu/power-and-prog… Please share the message and join in the discussion.

Dear followers, I’m happy to share a link for 30 free copies of What Happened to Liberal Democracy #Giveaway for What Happened to Liberal Democracy?: Remaking a Politics of Shared Prosperity goodreads.com/giveaway/show/…

It's awesome to have you as a co-signer of the statement, Daron. I always enjoy discussing this topic with you. We've both long made the case that the greatest upside is in having AI complement people and help us do new things, rather than simply replacing workers. It's gratifying that so many fellow economists and AI leaders publicly agree.

Why did I sign this statement? First, I had a hand in revising it, after the organizers reached out to me. I did not feel like I could sign the initial version, but I felt that finding a statement that would reflect the overlapping views of a number of AI researchers, economists, and social scientists was important. Second, I agree with much of the revised text. Indeed, there is a possibility (or perhaps more than a possibility) that AI may become more powerful over the next 10 years. I’m still not convinced that we are going to see the very large productivity gains that industry insiders are predicting. But more powerful AI may (again no certainty, just may) lead to significant job displacement. This is a big economic and social risk. It could also have myriad consequences on human cognition, starting with K-12 and all the way to advanced science. Some of these consequences are good, some of them are dangerous. Third, while I do not like the comparison to the Industrial Revolution that much (feels like comparing apples to oranges to me), it is true that AI will have complex effects on the economy. Finally and most importantly, I wholeheartedly agree with the ending: “to build the incentives, guardrails, and institutions needed to steer AI in a direction that complements humans and benefits society.” This is what I have been arguing for over a decade now. Good AI needs to complement humans, and this requires a redirection, because the current focus on AGI is, in all but name, an agenda for displacing humans from meaningful work. That’s why steering AI must be a first priority. I’m happy that many thought leaders have agreed.

Dear followers, please see this new paper on the labor market and macroeconomic effects of low birth rates. Although almost all analysts and policymakers are very worried about the prospect of stagnant and aging populations, the evidence from cross-country and within the United States points to the opposite: lower birth rates predict faster growth of GDP per worker and wages. The reason: labor scarcity encourages more technology adoption.

Whose work most influenced you? Institute Professor of Economics at MIT and Nobel Prize winner, Daron Acemoglu, answers this question on a special episode. Hear from many of today’s leading social and behavioral scientists now: ow.ly/qPlY50ZlwXL

The World Cup was shaping up to be a spectacular success for the United States. The hospitality the country provided to the teams and fans was exemplary. The interest from American fans dwarfed the interest seen when the USA last hosted the World Cup in 1994. The matches throughout the US, Canada and Mexico were energizing, showing why this is the beautiful game. And the US team played exciting, dynamic soccer (football), demonstrating that they are up to the task of taking on even the best in the game. Then, it all got ruined. Not by the fact that the USA lost Belgium 4-1. That can happen to the best teams in the world. No, it was ruined, perhaps predictably, by President Trump, interfering and trying to reverse the one-match automatic suspension for the American player Folarin Balogun after he received a red card in the round of 32 game against Bosnia and Herzegovina. We will never know for sure whether it was pressure from Trump that reversed the decision, but this is what Trump has claimed. In any case, the decision to postpone Balogun’s one-match automatic ban by one year is completely unprecedented. Something looks fishy. The track record of FIFA and its president Gianni Infantino does not give anybody much confidence that this was a decision made on objective grounds, without interference. What is sad and dangerous is that President Trump’s repeated disregard for rules and institutions is having an effect in every domain. It has further polarized US politics, damaged US democratic institutions, significantly hampered global cooperation, and made security and peace more tenuous around the world. Now, it is poisoning the beautiful game. As in democracy, once you damage the institutions of world soccer (football), and the trust that people have in the effectiveness and impartiality of these institutions, the consequences will be long-lasting. Say what you will about Trump. He is consistent – consistently corrosive to the institutions our future depends on.

This is a longer version of an earlier thread. Thank you.

And here is the link to the podcast. How Will AI Affect Jobs? - Exchanges | Podcast on Spotify open.spotify.com/episode/77vj2r…

Dear followers, Please see this interview with Goldman Sachs on AI and jobs. How Will AI Impact the Labor Market? | Goldman Sachs goldmansachs.com/insights/goldm…

Speaking of AI, this story is v interesting. A benefit of the Microsoft-OpenAI breakup! If Microsoft were serious about deviating from the current path, that could have productivity benefits Microsoft’s Satya Nadella: We Can’t Let AI Giants Eat the Economy wsj.com/tech/ai/micros…

Thanks Economist for featuring my views, but they make me sound more negative than I am. I think AI is transformative, but I argue near-term productivity growth will be less than what industry insiders believe. Meet the world’s top AI-pilled economists economist.com/finance-and-ec…

Dear followers, Please find enclosed this conversation with Fortune on AI, productivity, inequality and the future of our society. Nobel Laureate Daron Acemoglu on 'brainless' AI discourse, myth of capitalism and Gen Z revolution | Fortune fortune.com/2026/06/21/nob…

This is a post about lessons from the World Cup for ethnic integration and patriotism. When the German footballer Deniz Undav, brought in as a sub in the second half, followed his first goal in the 68thminute, with a second in stoppage time to give his national side a 2-1 win over Ivory Coast, there weren’t many German fans who didn’t see him as the hero, epitomizing the spirit of their national team. Undav, the son of a Turkish-born and Syrian-born Kurdish Yazidi parents, didn’t have the classic German look. No matter. He was delivering an exciting victory to the German side and a glimpse of what makes the beautiful game so exhilarating. Undav was also offering a rebuke to the obsessions of both the hard right and the hard left. Not just him but the entire German squad was a living demonstration that people of very different origins and appearances can be woven into a shared enterprise that, at its best, becomes a source of collective pride rather than a threat to it; and loving your country and welcoming newcomers into it are not incompatible impulses. The right insists the first half of that sentence is dangerous; the left often flinches at the second. This isn’t the message from just the German team. The World Cup, played this summer across the United States, Canada, and Mexico, puts ethnically-mixed squads in front of billions of fans. Every serious contender in the tournament is fielding players whose families arrived a generation or two ago. And almost every one of them is generating a joyful patriotism shared by their team and supporters alike. The combination of diversity and pride, together, is precisely what the political extremes insist is impossible. Consider the right first. Across much of the industrialized world, mainstream politics has become consumed by anxiety over the immigration of culturally and ethnically different peoples – and the supposed discord it brings. Ideas once confined to the darker corners of the internet, from the “Great Replacement” to “remigration,” are now aired and debated in respectable forums. In a recent YouGov poll, 45% of Brits, 50% of Danes, 51% of the French, 53% of Germans, 51% of Italians, 52% of Poles, and 46% of Spaniards expressed support for a scenario in which immigration stopped and many recent migrants departed. To be clear, the harshest version of such an agenda, which involves forced expulsion (sometimes bordering a modern version of ethnic cleansing) commands little support, and survey wording may be inflating softer concerns. But even allowing for those caveats, this is a striking reversal from less than a decade ago, when many of the same publics welcomed those fleeing war in the Middle East. The left has traveled in a different but still troubling direction. Among some activists and philosophers, the world came to be read almost entirely through the lens of oppressor-oppressed dynamics, with Western nations placed firmly in the former camp. Patriotism – which at its best allows one to identify with the customs and history of one’s country while remaining free to criticize its failings – came to be viewed with suspicion, even disdain. The shift shows up in the data: the share of US Democrats reporting “extreme pride” in being American, according to Gallup, fell from the mid-60s in the early 2000s to just 22% in 2019 (even if it recovered somewhat since then). That last figure deserves a caveat. Much of the collapse coincided with a presidency many Democrats sharply disagreed with, so some of it may reflect disaffection with who governed rather than a principled rejection of country. But the deeper trend is real: on parts of the left, the very vocabulary of national pride has become an embarrassment. This is troubling because without a shared identity it becomes harder for national politics to coalesce around policies that will watch out for and support those who are losing out – as the working classes, especially those without four-year college degrees, have been over the last several decades throughout the industrialized world. Sport, of course, is not a perfect mirror of society. A national squad is a small, lavishly resourced, intensely managed group united by a single unambiguous goal-– hardly the same as integrating large populations into housing, schools, and labor markets. Nor is sporting integration as frictionless as the cheering suggests. England’s Black players were deluged with racist abuse after losing the Euro 2020 final; France’s “Bleus” are perennially dragged into arguments about who counts as truly French; and the US has an ugly history of racism not just in baseball but throughout many sports. Integration in sport is celebrated by the many and contested by a loud minority – exactly the pattern we see in the wider society. The World Cup does not prove that integration is easy. What it nonetheless shows is something the extremes deny categorically: ethnic integration (rejected as impossible or even undesirable by the hard right) and patriotic pride (often looked down upon by the progressive left) routinely coexist. Most fans, watching their multi-ethnic team carry the national flag, simply experience both at once and think nothing of it. If the hard right and the hard left could climb down from their high horses long enough to watch a few matches, they might rediscover what the rest of us already know: that integration and pride are not enemies. As on the pitch, they can be teammates.

A post about the ethical challenges of designer babies and the parallels this has two artificial intelligence. Designer babies, which are genetically engineered (“gene-edited”) for desirable health, physical and intellectual features, are now within reach of current scientific know-how. They raise a host of societal and ethical questions, not unlike some of those brought about by recent advances in artificial intelligence (AI). There are also ways in which AI and gene-editing can become symbiotic, feeding on each other, both technologically and socially. Yet in some ways, both an ethically-clear path – by conceiving and treating gene editing is a public health measure for all – microphone of command and the guardrail institutions to support it may be within easier reach with designer babies than with AI. The big breakthrough for gene editing came in the 2000s and especially the early 2010s, with the development of CRISPR-Cas9 technology, which enables high-precision genetic engineering. The technique uses a guide RNA to locate a specific sequence in the genome of a living organism, where the Cas9 enzyme makes a cut for disabling or editing a gene (for example, turning off a disease-inducing gene). The applications are vast. The technique can be used to modify seeds and agricultural products and to design custom-made medicines and vaccines. It offers the potential to tackle inherited genetic diseases (such as sickle cell) as well as later-in-life, disease-inducing somatic mutations (such as those involved in cancer). It also raises the possibility that genes can be removed and new ones added in embryos, in vivo or in vitro, thus producing what is commonly referred to as “designer babies.” Human genetic engineering could be purely somatic (affecting the individual during their lifetime) or take the form of germline editing, altering the genome of an individual in a way that would be passed on to their offspring. AI is technologically complementary to gene editing. AI can sift through massive amounts of data to find gene combinations that can change not just disease potential, but looks, physical strength and various dimensions of cognitive ability. The promise is real – and so is the ethical conundrum. Designer babies can take us one step closer to a truly two-tier society. A group of wealthy humans would use this technology to improve their cognitive and physical performance, not just their lifespan, and with germline editing, would pass this on to their offspring, making one class of society not just richer but also superior in terms of various important characteristics. To the extent that cognitive ability can be engineered this way, even if imperfectly, their offspring will now have an even more formidable advantage, going beyond the already (arguably) unfair privilege of inheriting much greater levels of wealth much higher quality education. With their physique also engineered this way, they will plausibly become more and more likely to match within their class, laying the foundations of a society divided between the privileged, endowed with an array of desirable and superior characteristics, ruling over the rest. This is not merely hypothetical. Embryo selection services based on polygenic screening are already sold commercially to IVF patients. AI makes this future both more likely and scarier, because it will likely create a cadre of billionaires and trillionaires intent on cementing their economic privileges with biological ones, while many people will struggle to find good jobs, or any jobs at all, especially if AI continues in an automation direction. A first reaction to this prospect would be to ban germline editing and anything looking like designer babies altogether. But this may not be the right reaction. We already “design” our bodies and our children by means of vaccines and operations. Parents in the industrialized world (and increasingly in the emerging world) test for a variety of diseases, such as Down syndrome, and can decide to undertake selective abortions. The line between these interventions and genome editing is not hard and fast, but amorphous. Why is it different to edit genes so that smallpox cannot occur than to inject children with a vaccine against smallpox? The difference isn’t about interfering with nature or impacting the genome. It is about two other aspects of central importance. First, the implications of germline editing are poorly understood. Living organisms are constantly attacked by pathogens, and some genetic changes, especially those passed from parents to offspring, can create unforeseen vulnerabilities. For example, the first gene-edited babies were born in China in 2018, as a result of the scientist He Jiankui editing the CCR5 gene to confer HIV resistance. But the same edit appears to increase susceptibility to other infections such as West Nile virus and influenza. We know what smallpox vaccines do, both because of their simplicity and from more than two centuries of experience. The far more complex and intrusive process of genome editing can do myriad things we cannot anticipate, including creating new population-level vulnerabilities to infrequent viral or bacterial infections. Much more experimentation and caution would therefore be necessary before moving forward even with the most basic interventions. Second and in my mind more fundamentally, vaccines are a public health intervention, whereas genome editing, at least in the foreseeable future, will be something enjoyed by the privileged few create and propagate privilege in ways that we have not even conceived. These two principles together create a clear set of guidelines for how we may want to approach germline editing. First, make sure, extra sure and then double extra sure that unforeseen consequences are limited both in number and scope. This requires not just research by individual scientists and labs, but a public infrastructure to do an even more rigorous testing and permission process than the Food and Drug Administration (FDA) is currently implementing for drugs and vaccines. Second, legislate and enforce the norms within the scientific and medical community that germline editing should be seen as only and only as a public health measure. No germline editing services should be allowed to be purchased in the private market, and any well-tested and understood beneficial genome-engineering intervention should be made available to all (and associated with an accurate and widely available information campaign, in the way that vaccines were initially supported). Implementing these guidelines isn’t pie-in-the-sky. We already have institutions that house relevant expertise and are well respected and norms within the scientific community are aligned with these bodies (which doesn’t of course mean that some irresponsible, risk-taking individual scientists wouldn’t try to break the laws and regulations). These bodies include the US National Academies of Sciences, Engineering, and Medicine and the FDA, both of which have published highly visible reports and guidelines for research and practice in this area. The International Summit on Human Gene Editing, held in Washington in December 2015 and convened by the US, UK, and Chinese academies, adopted a hard line against germline engineering, especially the alteration of gametes and embryos that would generate heritable advantages for the privileged few. Subsequent summits in Hong Kong (2018), where the He Jiankui case broke, and London (2023) have reaffirmed and refined that stance. The reason why I find the challenge in the area of designer babies a little more manageable than the ongoing changes in artificial intelligence is precisely that such bodies exist, have been proactive and are respected and listened to by the majority of scientists in the area. The tragedy in AI is that we are at the cusp of similarly transformative changes, with equally far-ranging implications, but there are no equivalent bodies with authority (AI safety institutes and EU-convened bodies lack the authority to influence the industry meaningfully) and individual scientists and labs appear to be oblivious to the societal effects that they may unleash. We need the same urgency when it comes to taking the systemic, ethically-clear stance against AI risks. Such a stance could also facilitate dealing with the various symbiotic challenges of genetic engineering.

And for those who think that pervasive automation coupled with repression is too far-fetched, please look at the book and manifesto by Palantir co-founder and CEO Alex Karp: techrepublicbook.com

Dear followers, I would like to share this new working paper. economics.mit.edu/sites/default/… It’s a theoretical exploration of the political consequences of pervasive automation. The worrying implication from the analysis is that the combination of rapid automation and redistribution used in order to quell discontent doesn’t seem stable – more automation necessitates more and more redistribution. This creates a natural complementarity between pervasive automation and repression as the method of choice for holding pitchforks at bay – a particularly concerning conclusion as AI is expected to lead to a lot more automation in the years to come.

Dear followers, Please see this new article from @MITSloan on pro-worker AI, which draws from my recent paper with David Autor and Simon Johnson for @BrookingsInst mitsloan.mit.edu/ideas-made-to-…

We're live from #ABCDE2026! Join Nobel laureate Daron Acemoglu and @WorldBankGroup Chief Economist @IndermitGill for a discussion on creating good jobs at scale for the next generation. Watch now: x.com/i/broadcasts/1…

I recommend Columbia mathematician Michael Harris’s wide-ranging, informative and thought-provoking essay in Boston Review on AI and mathematics: bostonreview.net/articles/knowl… Harris rightly worries about the possible negative effects of AI-generated proofs and mathematics on collective knowledge in the field, as I have also argued here: economics.mit.edu/sites/default/… Of course, used in the right way, AI is a tool that can be beneficial in many fields. The question is whether we can develop institutions, norms and practices to support its beneficial use and whether the current direction of the technology in Silicon Valley will enable us to do so.

Dear followers, 10 more copies of What Happened to Liberal Democracy?: Remaking a Politics of Shared Prosperity will be given away today. Unfortunately, only to US residents goodreads.com/giveaway/show/…