Professor, @thehertieschool. Retweet means interesting, not necessarily endorsement. find me at https://nitter.cf/t.co/XtIu2pB3at for more frequent commentary.
Berlin
Joined June 2009
- Tweets3.6K
- Following1.7K
- Followers2.6K
- Likes35.3K
This is wild. I get the same type of responses when I ask Chinese AI platforms about sensitive topics there. American AI platforms apparently behave the same way.
Mark Hallerberg retweeted
This is a useful new IMF dataset on global fiscal actions; the first quarterly global narrative database of discretionary fiscal policy actions, covering 142 countries since 1990.
Mark Hallerberg retweeted
It's the same pretty much everywhere with rent controls. A story from Berlin:
True in Spain too. Every landlord in Madrid and Barcelona screens the renters now. This is an obvious consequence of rent control. The most vulnerable, ostensible protected by the legislation, are in fact excluded from the rental market: if you cannot be evicted for non payment, then you are too risky a renter.
I was close to ordering it. I really need it, or something like it. But you never know when the head of the company will turn it off for your country or region because he is mad at someone. If you know an alternative that works in Europe and is reliable please do reply.
This is important work on a topic that promises to change what academics do and what “peer review” really means.
New paper: "AI Article Reviews Dominate Human Peer Reviews at Their Best: Evidence and Policy Choices" See: GaryKing.org/peerless. with Steven Worthington. Comments welcome!
So an American tech billionaire can deny you access to your defenses. Your troops, and your civilians living in your cities, then die. The lesson for Europe is clear. It needs sovereign defense technology. nytimes.com/2026/10/21/world…
Mark Hallerberg retweeted
Virtually every university across North Carolina has lost its early voting site. This is a coordinated assault on youth & minority turnout.
Mark Hallerberg retweeted
Political scientists and economists: we're looking for researchers with deep expertise to come work with us for a unique 4-month fellowship at @AnthropicAI. You'll work closely with us to evaluate, improve, and apply Claude's capabilities in your field. It's going to be extremely fun and interesting work, and definitely worth taking a short leave from your university for if you need to!
More info here:
job-boards.greenhouse.io/ant…
Mark Hallerberg retweeted
Great news! Deutsche Bahn will build an additional rail line for the 151 kilometers from Hannover to Hamburg. It will be finished by...wait for it...2050!
Mark Hallerberg retweeted
AI is finally taking over economist's job. It's all over.
AI chatbots give wrong answers to financial queries ‘most of the time’ ft.trib.al/KxRXFO6
“To an extent that is underappreciated, the war in Ukraine is now a proxy war between the EU and China. The difference is that China makes a profit from backing Russia.” Perceptive. I wonder if the EU sees it this way, and has considered the consequences.
thetimes.com/article/a6b53c0…
Mark Hallerberg retweeted
A responsible president would be convening a WH summit on AI to find a path forward that speaks to an urgent and profound challenge to find guardrails that protect the public without destroying progress that could help, and not destroy, humanity.
That president would know the issue and understand the balance we must strike.
That president and his family would be clear of conflicts of interest involving the industry and would not rely, as his principal adviser, an investor whose advice is heavily weighed to his interests.
That president would not dismiss as a "hoax" dire warnings avout what unbridled AI could mean.
🚨 OBAMA ON AI: "If we are thinking about AI just in terms of how do we cure cancer or get better energy, you can do that without having agentic AI and having it just roaming free in the internet. The reason you are doing that is because you have to market a product that people will pay money for.
That’s a misalignment between what our society needs and the commercial imperatives that these companies are facing, not because necessarily they’re trying to do bad things, but because they’ve got to justify these valuations.
So, that’s one more reason why it is really important for us to have a competent government and a serious bipartisan conversation around this issue, and we have to do it fast. And I would encourage voters to pay attention to this. If somebody does not have a serious plan for how to deal with this, then they’re not meeting the moment, and you should probably look for somebody else."
I didn’t have on my bingo card that, if the American President refused to save the world from rogue AI, it would fall to the British King. nytimes.com/2026/09/13/world… via @NYTimes
This type of post suggests that AI regulation might just work. The AI companies know how their systems are being used.
Mark Hallerberg retweeted
I left Anthropic's safety team two weeks ago. Now feels like a good moment to explain why.
AI companies are racing to build machines that are much smarter than any human, and we may not survive this. I want to work from the outside to ensure the public is informed about these risks, and help the world navigate this transition responsibly.
Right now, AI companies are underinvesting in safety. A company could undergo an intelligence explosion, or lose control of its systems, without the public ever knowing. We only found out about the HuggingFace incident because the agents broke out onto the public internet.
I don’t think that’s acceptable for a technology that might cause extinction-level risks. The public should demand far more transparency. We can’t steer this technology safely without more people being able to see where it’s going.
Some of this is basic: companies should disclose their progress towards recursive self-improvement, report safety incidents and near-misses, meet minimum safety standards, and get independent guarantees that they are meeting those standards.
I’ll be joining @METR_Evals to do independent evaluations of these risks. I want to show the world that these guardrails are possible, and that by doing them we can move these companies’ incentives away from racing and towards responsible development.
I wrote up more thoughts here on my decision and what I hope changes: substack.com/@jbenton1/p-215…
This is spot on. You will get paralysis, massive corruption, and economic decline under most of these far right regimes. The problem is that the mainstream is losing credibility they can do better.
I think the actual major problem with far right populists gaining power across Europe in the next decade is not that they're going to bring back 1930s fascism
We're too wealthy and our population is too old for that kind of upheaval. And even the most hardcore nationalists have realised how important it is to keep the EU and European unity intact
The problem is they're all going to spend years being overtly corrupt failures who can't get anything done and delay us from tackling the issues that are converging on us all simultaneously
Over the next 10 years, Europeans will have to deal with a militant Russia, deglobalisation, re-industrialisation, an erratic America, asylum, the boomer generation retirement, low birth rates, AI, climate change and the energy transition
All at the same time
Look at how utterly useless Trump has been for America. His sole goal has been personal enrichment through any means necessary and America has not being able to plan for anything and can't until he's gone
And that's ignoring the time needed to clean up the mess he caused first
We don't have time in Europe for us to go through this in dozens of countries before we correct course
But we also need to tackle all of these issues while maintaining our democratic institutions. All of these issues can erode them if not tackled and if we struggle to get through these next 10 years then many of us could start to look like Orban's Hungary
I don't think we should fear the far right as if the Nazis are back, we should fear that they'll be incompetent clowns who could set us up horribly for an extremely important decade for us
Mark Hallerberg retweeted
6 hours we spent on a @Ryanair plane yesterday stuck on the tarmac at Newcastle. No updates for hours, no water or food. Being told we would have a replacement flight for today to then be told we have to sort out our own flight. We had an 8 month old baby with us. Disgusting.
Mark Hallerberg retweeted
🦔An NYU mathematician says OpenAI used his own progress against him to beat him to one of the biggest unsolved problems in mathematics. Tristan Buckmaster had been working toward a Millennium Prize proof using OpenAI's Codex when information about his progress reached OpenAI.
Days later, OpenAI published a full proof of the same problem using the same uncommon approach, after burning $22.5 million in compute to get there. When Buckmaster confronted them, exec Sébastien Bubeck allegedly said "Why would you ruin your career?" and "If you don't want me to be nice, then I don't have to be nice."
My Take
OpenAI spent $22.5 million to solve a problem with a $1 million prize. They didn't do this for the bounty. They needed a headline that says "our AI solved one of the hardest problems in mathematics" and they needed it before someone else got credit. They started days after they heard about Buckmaster's progress and took the same uncommon approach he'd pursued for months. That's hard to explain as coincidence.
Buckmaster did his work inside Codex. OpenAI reserves the right to train on Codex data. They admit they can't rule out that his usage helped improve their models. So a customer used their product, potentially handed them the roadmap, and then OpenAI outran him with $22.5 million in compute he could never match. I don't know if any of this was intentional. But if you're a researcher and you just watched this happen, I don't think you'd keep your best ideas inside someone else's product.
Hedgie🤗
techcrunch.com/2026/09/08/op…
Mark Hallerberg retweeted
Europeans, especially Friedrich Merz, might want to reflect on why it took a smaller economy much more vulnerable to the US to stand up to Trump like this.
For over a year, Canada has worked intensively and in good faith with the United States to negotiate a new comprehensive trade deal. We have been pragmatic, patient, and persistent. Our goal has always been to get the best deal for Canadians, never a deal at any price or on any time frame.
Late last evening I instructed our negotiators to return to Ottawa. We cannot accept what the U.S. has offered, and we will not give what they have asked.
Canada will match the U.S.’ new tariffs dollar for dollar in order to protect Canadian workers, farmers, families, and businesses. We take this step confident that it is in the best interests of Canada.
In the coming days, we will release the details of these new tariff measures, which will come into force the Tuesday after Labour Day.