close observations of my most traditionally successful peers suggests much academic productivity is driven by a desire to be seen as really smart this has put me at a profound career disadvantage, because, more like a normal person, I just want people to think I'm hot and a good cook

The ACM Digital Library, where a LOT of computing-related research is published (I'd say at least 75% of my own publications), is now not only providing (without consent of the authors and without opt-in by readers) AI-generated summaries of papers, but they appear as the *default* over abstracts.

It's astonishing how a technology called artificial intelligence has revealed the actual stupidity of so many people. This tweet hasn't been bettered

'whether LLMs can make “scientific discoveries” is no more interesting a question than whether Galileo’s telescope “really saw.” The better question is whether LLMs can be usefully integrated in the processes by which scientific knowledge gets made, validated, and disseminated.'

Princeton's AI Lab is advertising positions for AI Postdoctoral Fellows in two areas: studying natural and artificial minds, and designing, understanding or engineering large AI models. We are also searching for a Lead Research Software Engineer! ai.princeton.edu/ai-lab/emplo...

The Pennsylvania Department of Transportation (PennDOT) does not keep the trains and trams running. It will not retitle my vehicle. It does not plow the streets of Pennsylvania. If PennDOT maintains neither the public transit nor the vehicular transit infrastructure, does it actually exist?

Important post: my colleague documents an “official, nationwide policy of intimidating and threatening people who attempt to observe and record DHS operations,” or who peacefully and without interference follow and protest them, even though the First Amendment broadly protects doing those things. /1

another and slightly less stupid reality is possible

If you have to read anything about the prospect of “automating scientific discovery,” “agents for science,” or integrating LLMs into scientific pipelines, please let it be this essay by Kevin T. Baker: artificialbureaucracy.substack.com/p/context-wi...

A nightmarish development. And flagrantly unconstitutional.

This whole being an AI expert thing is exhausting, and for what payoff? Until now, none. I’m glad that I can finally say it’s all been worth it: finally, I can rub the AI into my eyeballs.

I created these Charles Darwin Greeting Cards for All Occasions (or for 4 very specific occasions) 9 years ago today

@bayesianboy.bsky.social An adult spherical cow!

This just an admission that you are biased against associating with extremely annoying and delusional women

To anyone who may have been present at my talk today in which I intimated that you can just, you know, do this: you can just, you know, do this.

I see your “Journal of Artificial Intelligence and Consciousness” and raise you my “Journal of Earth and Flatness”

It would be impossible for me to stress the following enough: a statistical distillation of internet text is not an appropriate tool for military decision-making.

Sec. Hegseth rolled out the U.S. military’s new AI platform, GenAi.Mil, with a link to an empty website. trib.al/tbCJAkT Predictably, the platform can’t actually be accessed from external networks, but the wonky rollout triggered eyerolls across the internet.

manuscript central wrapped (you can only see it after resetting your password)

A few years ago research in philosophy of machine learning was almost exclusively rubbish. That’s still the majority of what I am asked to review or see in print but I cannot express enough how refreshing it is that a modest percent of PhilML is now interesting and diligent work.

What strikes me over and over again about the reality we are currently creating with so-called “generative AI” is that it’s always pitched as saving human time and effort. But this is only true if you are a parasitic grifter. For anyone with principles, anyone who cares about reality

Shameless self promotion, but hopefully forgivable, given the moment we’re in right now. Military ethics is under grave threat. SecDef Hegseth is perverting the “warrior ethos” and putting the mental, moral, and spiritual health of all U.S. service members in jeopardy.

Very much honored to have made the list.

Very happy to see realistic appraisals of language models and uptake of these accounts in legal deliberation contexts.

teaching students through misspellings provocatively: inciting provocation provacatively: promoting bovine supremacy

I had the privilege of delivering my first talk at @princetoncitp.bsky.social today.

Some percentage of the population consistently reports that they are reincarnations of Jesus, that the CIA are monitoring us with subcutaneously injected spy drones, and that the moon landing is a hoax. Is it time to consider whether they’re onto something?

can’t fucking catch a breath make it stop

I’ve been running around asking tech execs and academics if language was the same as intelligence for over a year now - and, well, it isn’t. @benjaminjriley.bsky.social explains how the bubble is built on ignoring cutting-edge research into the science of thought www.theverge.com/ai-artificia...

I’m co-organizing an “AI for Science: Algorithms to Atoms” social event during #NeurIPS2025 with Yann LeCun, Anima Anandkumar, Bill Dally, and Max Welling. If you want to talk about AI Scientist, World Models, the future of AI-driven discovery, come by on Dec 5 3:30pm PT! luma.com/AI-for-Scien...

RIP to the farmer’s almanac. astrology for corn is dead.

I love it when boys come together online to celebrate the accomplishments of women. What an honor to have earned a PhD in 2025 and to receive such positive support from the online community.

Alternatively, according to this account, women’s norms of epistemic deference are innate, while men are essentially… poultry.

Discourse surrounding “AI” these days often feels like a guy showing you his shovel and you say, “that’s a nice shovel” and he says “I am going to perform surgery on my mother with this shovel” and you say “Oh God please don’t that is not what a shovel is for.”

Sometimes I look around me and think: I don’t think any of us, not a one of us, has the right heuristics for when to cede epistemic authority in an information ecology such as this.

Antisemitism is a lot like Qanon child abuse conspiracies in that there’s this real and bad phenomenon and, parallel to that, what can only be described as a large scale mass psychosis about a mostly fictional version of the same phenomenon

Don't be evil? "Google is hosting a CBP app that uses facial recognition to identify immigrants, and tell local cops whether to contact ICE about the person, while simultaneously removing apps designed to warn local communities about the presence of ICE officers..." www.404media.co/google-has-c...

There is a lot of fuss today over whether chatbots can replace human participants in social sciences research when the solution is obvious: ask chatbots to simulate the views of social scientists and survey them on attitudes towards chatbots as substitutes for human subjects.

OpenAI is planting the seeds for a pre-bailout, trying to dupe the government into guaranteeing its future. @ryanlcooper.com explains why this makes no sense given the fantastically absurd business model of the company: prospect.org/2025/11/07/o...

I can’t believe Zohran is transing the trains already www.mta.info/article/f-m-...

"[S]tall, manipulate and deceive." "As far as I have experienced, [ceasefire] simply means that Israel is able to do whatever it wants. We aren't." "This is the 'peace' we are offered after two years of suffering the crime of crimes." #Gaza #Palestine #Israel #USA #Genocide #Military

ML based facial recognition is not reliable in the wild; like many “AI” tools sold to law enforcement, it is thinly-veiled pseudoscience and will harm people by lending law enforcement both a felt sense of epistemic authority and legal impunity.

Surprise: AI based facial recognition is still pseudoscience and will harm people.

rookie mistake. that’s \pagenumbering{gobblegobble}.

I went all the way to Mexico to give a keynote and all I got was this lousy t-shirt

"Plan for a future where you can buy GPUs for ten cents on the dollar, where there's a buyer's market for hiring skilled applied statisticians, and where there's a ton of extremely promising open source models that have barely been optimized & have vast potential for improvement." Yup: Local & open

Nature suggests you use their "Manuscript Adviser" bot to get advice before submitting I uploaded the classic Watson & Crick paper about DNA structure, and the Adviser had this to say about one of the greatest paper endings of the century:

“The sooner the bubble pops, the better” is also true if what we care about is continued financing of AI/ML R&D.