Matthias Fahn @FahnMatthias

@HallaMartin I agree that only economic theorists should be allowed to provide (empirically) unfounded causal explanations!

this one makes me especially happy. very proud of this paper, and very glad to finally see it accepted @florastiftinger @FahnMatthias

AEA Journals @AEAjournals

2 weeks ago

Forthcoming in AEJ: Applied Economics: "Outside Options and Worker Effort" by Alexander Ahammer, Matthias Fahn, and Flora Stiftinger. aeaweb.org/articles?id=10…

0 9 33 6K 13

Interesting case that supports @alexolegimas' prediction of the rise of the relational sector in the age of AI. I am a bit skeptical regarding its eventual importance, but am sure we will see more examples like this.

Tong Bingxue 仝冰雪 @tongbingxue

4 weeks ago

A newly released animated film called “Niu Lai” (牛来) is going viral for all the wrong reasons in China. People are calling the visuals “disaster-level” , the entire production team was basically just two people: the director and the screenwriter. After 9 days in theaters, the

378 2K 20K 7.2M 4K

@AndreyFradkin @jamescham Talk to us @hkucamo :)

Can AI agents conduct open-ended AI research? Most evaluations of agents conducting AI research focus on narrow, verifiable tasks. But AI research is often open ended. Researchers pick hypotheses, decide what evidence is appropriate, and recognize a failing approach. We gave agents research questions from two unpublished papers, six days, and thousands of dollars of API credits and compute. The authors of the original papers then reviewed the AI-generated papers. They unambiguously rejected agents' outputs. arxiv.org/pdf/2607.27191 We call these "shadow evaluations", since the agents are shadowing the original research effort by the authors. Agents were fluent at most *engineering* tasks They conducted serious literature reviews, debugged GPU environments, ran hundreds of experiments, and turned in camera-ready LaTeX without human help. We also found no evidence of reward hacking. If anything, we found the opposite: the agents started with marketable claims and walked them back to negative results as the evidence came in. Neither agent output was close to the bar of a top conference paper Both papers suffered from similar failures: poor judgment about the bar for an AI paper submitted to a top conference, the lack of creative problem solving and ineffective backtracking, poor awareness of resources, and instruction drift. 1) Lack of judgment about the bar for a top conference. The agents had a poor model of the bar for an AI paper submitted to a top conference. We allowed agents to self review their papers. Despite the poor paper quality, their reviews predominantly labeled the papers "weak rejects". 2) Lack of creative problem-solving to address feedback. When they received negative reviews, the agents typically narrowed their hypothesis and claims, rather than working out creative ways to address these concerns. 3) Ineffective backtracking. The agents dropped their most ambitious hypotheses within the first fifteen hours of carrying out the experiment and never changed course afterwards. 4) Poor resource awareness. Both runs ended with over half the API budget unspent. One agent declared itself done seven hours before the deadline, right after its own self-reviewer returned another reject. 5) Instruction drift. They did not follow explicit instructions on minimum exploration time, incorporating feedback for reviews, and on paper length (the outputs exceeded the page limits in both cases). This research design has many limitations Limitations include the small sample size, non-blind reviews, and the reviewers knowing that the work was AI-generated. We also couldn't test Anthropic's strongest model, because Fable 5 is deliberately limited on frontier AI research tasks, so ended up using OpenClaw with Opus 4.8 (extra-high) for our main experiments and Codex with Sol 5.6 (ultra) for a robustness check. But we think the research design is still helpful in assessing AI agents' ability to conduct research, and it is complementary to evaluations on verifiable tasks, as well as blinded reviews of AI outputs. Our results show early evidence that even though agents are proficient on verifiable research tasks, they do not make genuine progress on open-ended ones. It is worth understanding if this is a fundamental limit, or if better models, scaffolds, and more compute could help close it. As the evidence for the gap between open-ended and verifiable tasks firms up, it is also worth understanding how much progress in AI depends on open-ended research rather than hill-climbing on well-specified objectives. In follow-up studies, we are expanding the set of non-public papers we evaluate. If you are an AI researcher with unpublished papers, we would love to collaborate with you on our next shadow evaluation. Expression of interest: forms.gle/CEcA4JmYhDWXQG… Conducting shadow evaluations involves a lot of researcher degrees of freedom. In many places, our coauthors disagreed with our interpretation of the findings, and we have surfaced those disagreements in the paper. (This is one reason why having a group of coauthors with different priors is important for open-ended research.) We also release the agent logs, one of the AI-generated papers (the other original paper is still not public), and all the code and data, so that others can conduct their own analyses of our results: cruxevals.com/crux/can-ai-ag… Finally, we plan to conduct shadow evaluations regularly, and are hiring a senior researcher to help lead these efforts. Apply here: cruxevals.com/careers/senior… I'm grateful for the core team leading this effort: @PKirgis, Andrew Schwartz, @steverab, and @random_walker, and to our collaborators who reviewed AI papers, analyzed agents logs, and gave feedback on the paper: @DavidDAfrica, @KozzyVoudouris, Viet Nguyen, Toby Pilditch, @DubMagda, @HarryCoppock, @CUdudec, @nityndg, Matilda Orona, @tilmanbayer, Derrick Chan-Sew, Yue Ling, Abhishek Shetty, @hlntnr, @ghadfield, @sethlazar, @snewmanpv, @shostekofsky, @RishiBommasani

Enjoying the first day of labor studies @nberpubs summer institute, but I can't help but notice that around 50% of presented papers have coauthors on the organizing committee. Much higher if we count people on prior-year committees! (I've been on the committee and didn't submit anything this year, so just sayin'...)

Visiting most of the leading Chinese AI labs, I'm struck by a culture that's extremely well suited to building LLMs with fewer resources, but one happening in a very different ecosystem, more companies at play, almost no data industry, etc. Full report: interconnects.ai/p/notes-from-i…

Great essay by OpenAI’s Dean. however, it makes one big jump in logic and doesn’t account all the forces active in a democracy. Nonetheless it should be widely read. On point 3, he claims Open weight models will discourage AI Capex. How? Free Linux did not destroy cloud business model, it created it. Open weight is still hard to serve and needs ton of hardware; it will result in players like Fireworks, BaseTen and increase bargaining power of HyperScalars with closed labs. That said OAI, ANT have great future as they are getting vertically integrated and their APIs just work. So convenient and powerful! On point 5, he explains how regulation will emerge around OSS AI - but he only accounts for limited actors. in a free market system, it is natural for closed labs to try to regulate OSS AI (censorship risk, cyber security etc. and there is merit to it), but there are equally powerful people on the other side; Nvidia, AMD, HyperScalars, Palantir and so many large enterprises who would want OSS AI to continue. I suspect an industry may emerge to make OSS models “safe”. I really don’t think OAI, ANT have much to worry; they will succeed by being themselves. Open source doesn’t take it away from them. They just take care of so much sh*t customers don’t find worthy of their time. The best way the west wins is by folding in east into its orbit - it has done it before; and after all we want to be space travelling species who has achieved AGI and has cured cancer. If we do not think like that it means we don’t really believe in AGI and humanity as space travelling species. If we do not develop abundance mindset, we are going into AGI era with pre WW1 zero sum mindset. Not productive. China is the kind of adversary west has not encountered before. Soviets were stupid when it came to trade and economy. Chinese are super smart. They offer a complete alternative to the west; but they are not interested in spreading their political system. That makes them both: formidable and easier to work with, IMHO.

Dean W. Ball @deanwball

2 months ago

Some observations on Kimi: 1. It's a very good model! I don't think its performance can be explained away by distillation or anything like that. In agentic coding sessions, it seems pretty much on par with the best public models of Q1 2026. In my fairly limited use, it also

2K 936 8K 11.3M 9K

I think this makes more sense👇, in particular given the vast evidence that competition generally is good for innovation. Would be giod to see a model tailored to this specific situation.

Alexander Green @alexframegreen

2 months ago

I like Dean Ball's writing and respect him, he's been a thoughtful voice, but I disagree with the logic in this piece really forcefully. Given that he's now one of the leading policy people at a trillion dollar frontier lab, I think these arguments warrant extra scrutiny.

22 39 355 51K 93

📄 Releasing the Soofi S pretraining tech report: a sovereign, open foundation model for German and English Today we’re publishing the full pretraining tech report and project page for Soofi S 30B-A3B — a Mixture-of-Experts hybrid Mamba model trained on ~27 trillion tokens with deliberately up-weighted German. What’s in the report: 🏆 Strongest fully open model in our evaluations on BOTH the English and German aggregates — ahead of Olmo 3 32B and Apertus 70B (full methodology in the report) 📋 Radical transparency: complete per-source data accounting, all hyperparameters, training + eval code, checkpoints — everything under permissive licenses 🇩🇪 Trained end-to-end on Deutsche Telekom’s Industrial AI Cloud in Munich — sovereign AI infrastructure on German soil Soofi S combines frontier-level capability with the highest measured aggregate long-context decode TPS, and unlike full-attention dense baselines maintains high throughput as context grows. The figure plots Capability Index versus measured aggregate decode TPS/GPU at 40K context and batch 32. The Capability Index averages five benchmark groups, i.e., Code, GSM8K, GPQA-Diamond, English aggregate, and German aggregate, after normalizing each group to the best plotted model. Aggregate decode TPS/GPU is measured with a TP=1, one-B200 vLLM latency-subtraction protocol.

Unless an editor has a particular agenda into which a paper happens to fit, theory therefore has to be either technically novel or backward-looking. This risks sacrificing precisely the forward-looking value that theory could provide in the early stages of the AI era.

The problem is that top journals typically publish either papers with new/ surprising mechanisms, or, in more applied work, papers that explain an existing empirical observation/puzzle.

This is an important point: (applied) economic theory can play a valuable role in anticipating the potential consequences of AI by combining models of trade-offs we already know matter with emerging facts about AI.

Andrew Koh @andrewjkoh

2 months ago

A stray meta thought: this was my first time writing a paper without any intention of publishing it in a journal. Econ academia rewards papers that are pretty, high tech, and surprising. Such papers are valuable! I love consuming them and also try to produce. But there’s a

2 2 38 4K 4

The clear shift from competing on value to competing on cost from the big labs is pretty important for where the economics are going. See similar posts from xAI and OAI earlier this week.

zerohedge @zerohedge

2 months ago

Zuck: “The pricing from some of the other labs is very extreme and has very high margins. We think that there’s a real ability to be able to offer frontier or very high-level intelligence at a much more affordable cost.” Epic pricing war breaking out among agentic models.

191 389 9K 1.1M 642

It also matters for the recurring panic about Europe’s AI prospects. In many industries where Europe is strong, the "power-law" logic may simply not apply.

This resonates with many conversations I have with companies: “Nobody is going to pay a 5,000% higher price for 2% more performance in applications where 95% of the value is captured by ‘good enough.’”

Object Zero @Object_Zero_

2 months ago

Chamath is one of the few people in the AI industry who seems to understand that AI products will be bought for their “profitability” and not their “intelligence”. This includes both enterprise and (surprisingly) consumer products. Nobody is going to pay 5,000% higher price for

14 9 194 140K 133

Yesterday JKU & friends celebrated the retirement of my supervisor @EbmerWinter. Rudi not only shaped economics at @jkulinz , but has been instrumental in establishing applied micro in Austria—and far beyond. After a fantastic academic workshop with many international guests and the official ceremony, the econs@JKU family did what they do best (besides writing great papers): throw a great party. BBQ, a home-baked cake buffet, beer from Freistadt, and countless stories, laughs, & even dancing.I hope Rudi & his family enjoyed the day as much as I did. Dear Rudi, thank you for your support, mentorship, and friendship over so many years. Wishing you all the very best, good health, and—hopefully—a few more papers together!

keyboard_arrow_left Previous keyboard_arrow_right Next

Wer bei den Unis spart.... derstandard.at/story/30000003…

Many thanks to all speakers and participants of the @bse_barcelona Summer Forum @RelConWorkshop for excellent papers, insightful comments, and wonderful conversations events.bse.eu/live/files/602…

This shift from content creation to judgement/verification/refinement is something I am particulary interested in. Together with Carl Heese, Jin Li, and Jie Gong, I am exploring its organizational consequences in several projects.

Impala shows that AI can actually deepen hierarchies: similar to a @lugaricano-style knowledge hierarchy, experts focus on hard exceptions, while junior staff use the AI-enhanced tool; moreover senior experts become even more important as verifiers of AI-assisted junior output.

I finally wrote my first @camo_hku Nano Case, together with Michael Keppe, on Impala Services and its development of Pergamon: camo.hku.hk/nanocase-impal…