Daron Acemoglu (@DAcemogluMIT) header
Daron Acemoglu (@DAcemogluMIT) profile picture

Institute Professor @MIT, @MITEcon. Co-Director of @MITShapingWork. Author of Why Nations Fail, The Narrow Corridor, and Power & Progress.

Joined April 2023
Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)Photo by Daron Acemoglu (@DAcemogluMIT)
Daron Acemoglu retweeted
Ahead: @NinaTotenberg on what’s ahead for US courts after the mid-terms. Nina’s is the voice we all want to hear. And @DAcemogluMIT on what’s ahead a new congress might do that might lower prices.
3
13
11,342
Daron Acemoglu retweeted
Spotted today in Amsterdam @DAcemogluMIT!
1
3
28
6,898
Daron Acemoglu retweeted
Die Krise der liberalen Demokratie ist nicht vorbei, warnt der Ökonom Daron Acemoglu, auch wenn in den US-Wahlen Donald Trumps Partei verloren hat. Ein Gespräch über KI und linke Parteien, die die Arbeiter vergessen haben. faz.net/aktuell/wirtschaft/m…
3
12
1
36
14,460
Daron Acemoglu retweeted
And Brazil defeated extremism, guns, environmental destruction, militarization, democratic threat, and organized crime ruling politics. It’s not going to be easy for Lula, he will need to do a very different government, but what a beautiful mobilization of civil society
12
102
1
772
18,856
Daron Acemoglu retweeted
Hi @elonmusk , we’ve conducted the most comprehensive study to date on tax unfairness in the United States. We found that you, like your fellow billionaires, pay a lower effective tax rate than the average American. So if you really want to help the American people, I have a better suggestion than charity: pay more taxes. 🧵
Next year, I will give one of my SpaceX shares to every kid in America. This will be the first of many.
3,089
1,088
285
4,597
602,859
Daron Acemoglu retweeted
Team Erdoğan’s corruption has become unsustainable. In a normal country, the leader would hit reset. Erdoğan may no longer be able to. Too many of his cronies now have vested interests in the status quo. A real cleanup would jeopardize the arrangements that keep them loyal.
9
242
8
1,345
61,837
Daron Acemoglu retweeted
Over 400 researchers and I have signed this open letter calling for the AI and security communities to come together to improve alignment and containment approaches in order to protect our critical infrastructure and digital systems. ai-security-letter.org/
67
206
45
681
58,720
Are you also worried about the future of democracy? Come and join us for this conversation with @DAcemogluMIT on Tuesday October 13! Prof. Acemoglu is asking the right question: Can ‘working-class liberalism’ heal liberal democracy? brookings.edu/events/can-wor…
2
5
46
19,774
Daron Acemoglu retweeted
Dear Followers Over the next three days I am going to respond to @DAcemogluMIT 's questions on AI. There are lots of reasons to do this but I also wanted to send a big FU to that Economist piece that suggested we economists aren't willing to challenge Daron. That is not my experience and I know people who challenge him all of the time. None are shy about it. If anything, they do so in our normal circles more than, say, here. So let's do it more here too.
1
7
1
167
20,919
Daron Acemoglu retweeted
Economist Daron Acemoglu is not optimistic about the future of humanity if AI develops on its current trajectory. Full interview is live now. @DAcemogluMIT
1
4
1
19
12,362
Daron Acemoglu retweeted
Really great to see this: six Nobel Prize-winning economists (Duflo, Krugman, Diamond, Banerjee, Acemoglu and Stiglitz) just came out in favor of the California billionaire tax. You don't have to be a radical leftist to see why it's a good idea. First, look at the numbers: ten years ago, the richest 0.001% of Californians were worth $700 billion. Today, that same group (about 250 people) is worth $2.3 trillion. From 2019 to 2025, their wealth grew by $1.4 trillion. The California income tax they paid over that period was just 1.6% of that gain. A nurse or a plumber pays a higher state tax rate. Yes, many of these people built great companies, and they've made a lot of money in the process. Good for them! The problem is that wealth on this scale stops being just money. It becomes extraordinary power. In 2000, billionaires accounted for about 1% of all donations in federal elections. In 2024, it was 19%. The Washington Post, the Wall Street Journal, the LA Times, Facebook, Instagram, TikTok and the platform you're reading this on: all owned and controlled by billionaires. Now add AI. If you think power is concentrated now, just wait five years. It's going to get much worse. So, what does Prop 40 actually do? It's a one-time 5% tax on the wealth of California's billionaires. That would raise about $100 billion, enough to offset the federal Medicaid cuts. "But they'll all leave!" the critics say. Nope, they can't. The tax applies to any billionaire who lived in California on January 1, 2026, and the initiative only became public about five weeks before that. "But it'll kill Silicon Valley!" Uh no, that won't happen either. Since the tax was announced, California's share of all new US venture capital has gone up, from about 50% to 80%. Nvidia CEO Jensen Huang, who would be one of the biggest payers, said he would be "perfectly fine" with it and had "not even thought about it once". Sure, some other billionaires can't stop thinking about it. Sergey Brin has already spent more than $100 million to defeat the proposal. At a holiday party in December he pulled Gavin Newson aside and said 'he could not stand the state’s proposed billionaire tax', because he thought the 5% tax would turn California into some kind of Soviet Union. That is, of course, nonsense. This is not going to be some kind of socialist revolution. The proposal is about restoring balance to a mixed economy. You could even argue it's about saving capitalism itself from oligarchs like Sergey Brin. I think that's exactly why Nobel Prize-winning economists are coming out in favor of this tax. If Prop 40 passes in November, it will be the first tax aimed specifically at billionaire wealth anywhere in the world. I think it could be the start of a global movement. Anyway, read the whole letter here. It's really good 👇 gabriel-zucman.eu/files/prop…
164
1,967
110
4,975
272,113
Daron Acemoglu retweeted
AI firms are subject to already-existing laws, @linamkhan tells @cwarzel: Part of the conversation around regulating AI “seems to assume that these current developments are happening in some legal vacuum,” she adds, “and that is just false.” theatln.tc/65NUk0Q9 #TAF26
8
126
9
354
48,971
Daron Acemoglu retweeted
Honoured to have chaired the LSE event with @DAcemogluMIT on his new book 🎙️ Podcast here: lse.ac.uk/lse-player/what-ha… We covered a lot in an hour: Acemoglu's core arguments, plus a few questions from me.
1
6
100
11,724
We asked co-director Simon Johnson how things have changed since his book "Power and Progress" (co-authored with @DAcemogluMIT) came out in 2023. Is it too late to redirect the path of AI? Watch the full AMA video with @baselinescene: youtube.com/watch?v=OgMbxjEA…
3
6
6,543
Daron Acemoglu retweeted
Replying to @DAcemogluMIT
Having read the summary, i found it super-convincing, if this is indeed a correct summary. In particular, that the focus should be on application and diffusion, not pursuing AGI fairy-tales. I studied a bit China - AI diffusion is ahead of the West, despite somewhat weaker models - E2E ai orchestrated transactions in SuperApps (contrast with awkward attempts to slap ads on chatbots in the West), 10,000 +ai produced mini-dramas per month and other applications are much ahead. Incidentally, i read somewhere that in China people are optimistic about the AI contrary to the west. Also AGI “wet dream” implies replacement of humans while the approach that augments and enhances humans to make a pair human+ai a super combination worth 10x a human alone implies empowerment. The former is misanthropic, the latter inspiring.
3
11
6,463
Daron Acemoglu retweeted
There's an attention trap in AI. Yes there are legitimate questions regarding AI-related existential risk and the welfare of AI models. But these come *way down* the list in terms of urgency as compared with other AI-related problems. The disproportionate focus on the former distracts attention from the latter and even impairs our ability to engage with the more urgent problems.
4
5
39
8,916
Daron Acemoglu retweeted
A capable president would move swiftly to secure and reassure the American people, rush inspectors into frontier labs, fortify the nation against bioterrorism, demand legislation from Congress, and lead the world toward an AI treaty. But President Trump is compromised.
1,202
1,757
235
7,279
1,356,923
Daron Acemoglu retweeted
Terence Tao: The math behind today’s LLMs is actually simple. Training and running them mostly uses linear algebra, matrix multiplication, and a bit of calculus, material an undergraduate can handle. We understand how to build and operate these models. The real mystery is why they work so well on some tasks and fail on others, and why we cannot predict that in advance. We lack good rules for forecasting performance across tasks, so progress is largely empirical. A key reason is the nature of real-world data. Pure noise is well understood, perfectly structured data is well understood, but natural text sits in between, partly structured and partly random. Mathematics for that middle regime is thin, similar to how physics struggles at meso-scales between atoms and continua. Because of this gap, we can describe the mechanisms but cannot yet explain capability jumps or give reliable task-level predictions. That mismatch, simple machinery versus hard-to-predict behavior, is the core puzzle. ---- Video from Prof @Briankeating YT Channel (Link in comment)
125
834
76
4,981
495,908
Daron Acemoglu retweeted
Two mistakes in responding to hype about AI existential risk - 1) taking it seriously and ignoring more pressing issues around the dehumanisation of life, and 2) dismissing out of hand the very idea that AI is a transformative technology which can yield great benefits if properly directed.
4
8
40
9,779