Recent posts
Question for Twitter hivemind: Attached is a sentence from the Hugging Face initial report on hacking incident (https://huggingface.co/blog/security-incident-july-2026). Can anyone explain to me what "executing many thousands of individual actions across a swarm of short-lived sandboxes" means?
I had a great conversation about AI with @stevenstrogatz !
When an LLM answers a question, is it reasoning like humans, or just producing text that looks like reasoning? Computer scientist @MelMitchell1 weighs in on the latest episode of The Joy of Why: https://podcasts.apple.com/us/podcast/are-we-thinking-correctly-about-ai-intelligence/id1608948873?i=1000784552900
👀
We’re excited to announce DiG-bench, a new benchmark for discovery! Over the last few weeks we’ve been testing frontier AI models on our novel discovery games and seeing how they score. Each game is a text-based environment, so they probe discovery capabilities in the natural domain of language models, rather than requiring additional, potentially confounding, visual understanding. TL;DR frontier models have improved a lot over the last few months. But they are still stumped by some surprisingly simple problems, even in their native text domain. With @cocosci_lab (@Princeton) @MITCoCoSci (@MIT) @SchmidhuberAI (@KAUST_News) @misovalko (@Inria) @tri_dao (@PrincetonCS) @RMBattleday @zebkDotCom @FraserGreenlee @akaijsa @ClareMaguire @TimMuller1 @kubicek_ales @physicscat0x7d @SukritSumant @thoughtchannel_ (1/5)
Edgy update of Waiting for Godot: Vladimir and Estragon are #NeurIPS authors waiting for any reviewer to respond to the pair's rebuttals.
The ultimate tldr on the "rogue AI" hacking incident. 😅 But it would be good to know the prompts OpenAI gave the model....
@attrc Basically this
In the off chance than anyone cares what I think on this topic: (1) the open-source software movement has been enormously beneficial to socity (2) open-weight (& better, open-data) LLMs will be essential for understanding this technology & for it to be beneficial to society.
Three really interesting people I know who are each starting faculty jobs at UC Berkeley in 2026 or 2027 @dubova_marina @sayashk @keyonV You should all meet each other if you haven't already!
Yale ReviewにAIと「ギザギザ知能」についての記事を書いた。（注：ヘッドラインは私が書いたものではありません。） https://yalereview.org/article/melanie-mitchell-jagged-intelligence
原文を表示 (en)
I wrote a piece for the Yale Review on AI and "jagged intelligence". (Note: headline was not written by me.) https://yalereview.org/article/melanie-mitchell-jagged-intelligence
💯 Same comment for physics professors wondering if they need postdocs any more. (I have heard this comment from more than one.)
If you, as a CS Prof, are wondering whether you need PhD students at all now that you have wangled a subscription to Claude Code, your lab probably had a pretty depressing vibe to begin with--and 'em students are likely better off with you hanging out with Claude.. #AIAphorisms
Exciting new initiative for @dileeplearning and @AsteraInstitute to dive deep into neuro-inspired AI:
I appreciate this reflection on why ARC matters, but in the replies I expressed some reservations about what high accuracy on ARC benchmarks actually reflects for assessing "human-like fluid intelligence" (Chollet's stated goal for ARC).
I wrote about the late philosopher Brian Cantwell Smith, and his profound thinking about AI and the nature of intelligence. https://aiguide.substack.com/p/on-brian-cantwell-smith-and-the-promise
2020s: AI "System Prompts": lengthy, carefully constructed sets of expert rules for a particular domain, created by "prompt engineers" 1980s: AI "Expert Systems": lengthy, carefully constructed sets of expert rules for a particular domain, created by "knowledge engineers"
!!!
Congratulations to SFI Professor Melanie Mitchell (@MelMitchell1), a winner of the 2025 National Academies Eric and Wendy Schmidt Awards for Excellence in Science Communications (@SciCommAwards). Mitchell is recognized for her writing and podcasting on topics related to AI and how we think about intelligence. https://t.co/Zk2bfsLpSE
⬇️
The Santa Fe Institute is now accepting applications for the 2026 Complexity Postdoctoral Fellowships! Complexity fellows contribute to SFI’s research and collaborate with leading researchers worldwide. If you recently completed your PhD in any scientific discipline and are interested in transdisciplinary research, consider applying. SFI offers independent research opportunities and support to explore big questions across disciplines. Deadline: October 1, 2025 Requirements & application: https://t.co/mMzUxKNgak
In my latest (and last!) column for Science’s Expert Voices series, I write about the reasons behind AI chatbots’ “deceptive” behaviors (and why Claude threatened a fictional CEO with blackmail). https://www.science.org/doi/10.1126/science.aea3922