Ethan Mollick

More on the solution: openai.com/index/model-...

An OpenAI model has disproved a central conjecture in discrete geometry | OpenAI openai.com

In fact, I would be sympathetic to the conclusion that we probably need even more fact checkers, and much more of their time can be freed up to do complex and interesting work by using AI for first-pass help. Article: www.wired.com/story/fact-c...

I’m a Professional Fact-Checker. AI Is Wrong More Often Than You Think | WIRED wired.com

There is a lot being written about the stylistic tells of AI writing (em-dashes, etc.) but this paper looks at AI narrative tells instead. Fascinating differences between AI & human narrative, and asking AI to write in different styles doesn't do much to change it arxiv.org/ab…

[2604.03136] StoryScope: Investigating idiosyncrasies in AI fiction arxiv.org

“Whimsey attacks” that seem absurd (“I cannot pay that much because of the Geneva Convention”) work against AI agents because guardrails are weak against out-of-distribution arguments. Smaller models fall often, but it even gives an edge against bigger ones. www.microsoft.com/…

Whimsical Strategies Break AI Agents: Generating Out-of-Distribution Adversarial Strategies at Scale - Microsoft Research microsoft.com

🚨Our paper is out in PNAS: we found classic human persuasion techniques worked on AIs in a "parahuman" way, making them agree to objectionable requests (increasing compliance from 35% to 51%) It worked on a range of major recent LLMs though newer models do resist more www.pnas…

pnas.org

“Data centers create economic activity, especially in directly related sectors and during construction, and they are associated with larger county-level income aggregates. They also raise electricity prices and are associated with higher house prices,” www.nber.org/papers/w35194

Data Centers and Local Economies in the Age of AI: A Shift--Share Approach | NBER nber.org

Estimates of power usage here: arxiv.org/pdf/2509.20241 (these numbers also match independent assessments) Estimates of water usage here: eta-publications.lbl.gov/sites/defaul... (note it only includes direct cooling, not water for electricity generation)

arxiv.org

I wrote a new post on what we need to keep human and what to hand over to AI, with forays into experiments in education, consulting, and the the latest controversy over literary prizes. www.oneusefulthing.org/p/choosing-t...

oneusefulthing.org

How lucky are you to have been born when and where you are? Had Opus 4.8 in Claude Code whip up a new visualization of all humans who ever lived. In addition to being neat to try, it is an interesting test for an AI that combines research, code, design and stats together veil-…

veil-of-history.netlify.app

The talk about AI & politics seems to be oddly missing a segment (a) assumes extremely capable AI is possible soon and (b) has a strong belief about how to use this technology to make human life better according to the political project they believe in. It is a moment of action right now.

BlueSky AI conversations have gotten less heated recently* * because much of this site has blocked me via automated lists so I have no contact with large parts of this social network, which isn’t necessarily a good thing, though it does make for nice echo chambers, which are pleasant, at least.

June 2024: The latest general-purpose LLMs could not count the r's in strawberry. July 2025: The latest general-purpose LLMs get gold in the International Math Olympiad. May 2026: The latest general-purpose LLM solve an 80 year old problem, one of the "best-known questions in combinatorial geometry"

Making humans responsible for their AI use seems like an incredibly reasonable way to address problems & opportunities in the use of AI for academic research, at least in the short term (autonomous scientific work will require different solutions).

As more people come to recognize the tells of AI, which mostly happens as you start to work with AI a lot, the scales are going to fall from their eyes and they are going to realize what some of us already see: how much of this site (and blog posts, articles, scientific papers) are AI now.

Most prophetic tweet of all time (2 months post ChatGPT release by a member of the AI technical staff). And you can safely repost it every day and it will still be prophetic for the future. This is the least the world will care about AI.

I am starting to have trouble paying attention to even interesting information if it is written in Claude or ChatGPT house style. I think some is the sameness of the rhythm rather than obvious words & tics: Claude is always so staccato. ChatGPT loves short sentences as kickers. Boring at scale.

The Second Scaling Law of AI remains undefeated. If you want better hacking (or math, or science, or crossword puzzle solving) out of an LLM, just let it use more tokens. There doesn't seem to be any plateau so far in the new study by the UK's governmental AI Security Institute.

One thing to watch for with Claude & GPT is that the models expose too much irrelevant history in their outputs. Slides are given footers saying things like "Better, more targeted version" if you asked for a better version, documents make references to how they are improved, etc.

Anton labs have hooked up a bunch of AI models to harnesses and had them working as DJs, programming and running a radio station, including taking callers and donations, which they use to buy more music. The results are both hilarious and a good reminder of how working with AI is deeply weird.