Footage of me at eight months pregnant right after the heartburn meds kick in

I am a big fan of a much more conceptual approach. Eg, press.princeton.edu/books/paperb...

Alright I'm gonna give this a full watch, live commentary thread below. youtu.be/87DyyMV0kCY

Black Hat USA 2026: The 'Breaking' News: The OpenAI–Hugging Face Incident

YouTube video by Black Hat

youtu.be

Is all code becoming the same? On one hand, 95% of Kaggle submissions that set a random seed now use 42 (a Hitchhiker's Guide joke LLMs love). But, it turns out that while coding syntax is converging, approaches to problems are not converging. Human prompters drive real variety in solutions.

Europe needs to do more to harness its AI ambitions, writes @marietjeschaake.bsky.social giftarticle.ft.com/giftarticle/...

Europe must do more to harness its AI ambitions

The US and China use the technology as an instrument of state power, while the EU is still considering its course

giftarticle.ft.com

Wondering if there would be a measurable difference in outputs if I'd end every Claude prompt with "Help a gal out!"

Maybe I'm late to the party but: Desrosiéres' "The Politics of Large Numbers" . . BOOM! What a book 😵🥇

ya i don't think people have thought about this fully solutions to problems becoming more bespoke, and private, isolated, and illegible to others does seem to imply problems that show up at scale (beyond the level where any individual with an agent could solve) will be harder to solve, not easier

@wavebidder.bsky.social · 4d ago

The whole point of open source software is to cache common solutions to common problems. Expecting users to customize *without explicitly asking them for PRs* is leaving a ton of value on the floor

This paper by researchers from MIT & Stanford finds that most people would be financially better off if they followed the advice of LLMs (GPT-5.2 & Gemini 3 Flash) But some people get a bit better advice than others, which largely depends on the questions they ask tahachoukhmane.com/wp-content/u...

I think I have to use my cat’s nonplussed stare of judgement as a reaction meme

🚩 Please take all of this seriously. On one hand this exhibits how AI can easily be weaponized in PSYOPs. On the other we see a correlation between need (according to Maslow's Hierarchy) for love & belonging coupled with escalatory suicidal/homicidal ideation.

Mark Follman@markfollman.bsky.social · 5d ago

My new investigation into a mass shooter’s harrowing history with ChatGPT reveals numerous red flags—and some very disturbing ChatGPT replies—from long before his attack on Florida State University www.motherjones.com/media/2026/0... 1x 🧵

This was one of several points that @siridahl.com and I discussed on a recent podcast about AI and sexuality: plus.flux.community/p/culture-da...

Siri Dahl on how new labor dynamics and AI have changed sex, porn, and dating

Women have remade adult media and some people are very upset about it

plus.flux.community

I don't even know why some discussion about closing pools in some place during a heatwave was in my feed and I have no educated opinion on that matter, but welcome to social media where everyone eventually dies a death by rhetorical question

Fascinating experiment: current AI systems lack creativity to reliably pursue research arxiv.org/abs/2607.27191 - poor judgment about the bar for publishable research - uncreative responses in research design - ineffective backtracking from dead ends - poor resource awareness - instruction drift

Can AI agents conduct open-ended AI research? Early evidence from two case studies

Forecasts of explosive AI progress hinge on AI agents automating AI research. But evidence on whether agents can carry out open-ended AI research is thin. Current evaluations either test agents on nar...

arxiv.org

Many don't see how much AI has changed recently. Coding agents aren't just chatbots. We can’t deal with the challenges of AI if we don’t understand it. Here's a post to provide some evidence (Claude doing small ML experiments) and form intuitions. davidpreichert.substack.com/p/if-you-hav... 🧵..

If you haven’t recently used Claude Code*, you might not understand where AI is at

A report on a series of mini machine learning projects, executed with, and mostly by, Claude

davidpreichert.substack.com

Plausible AI-generated data may not support counterfactual analysis. Different histories can leave the economy looking the same today while implying different mechanisms. When data cannot distinguish those histories, counterfactual analysis can fail. An illustration： medium.com/@ivyyangxueq...

hell fuckin' yeah, great hack mattbeton.com/blog/bitnet-...

I'm just now reading @randomwalker.bsky.social's writeup of his ICML keynote. It is very insightful. Everyone thinking about the future of work in the age of AI should read this. www.normaltech.ai/p/what-will-...

What will be left for us to work on?

My keynote at ICML 2026

normaltech.ai

“The AI assistant, which Kinney Drugs named Burt, after its founder, was built by a third-party AI pharmaceutical company, Synerio. In the pharmacy’s terms, Kinney Drugs states that ‘Burt may not be accurate, complete or up-to-date and may be misleading or contain errors and omissions.’”

Vermont pharmacy AI struggles with medical accuracy - Valley News

Vermont and New York pharmacy Kinney Drugs faces criticism after implementing an AI assistant that reportedly causes prescription errors.

vnews.com

As I said before, AI proving and disproving conjectures left and right just means value allocation in math needs to change. It doesn't mean that math as a human discipline is over. I would love for the community to celebrate being wrong in a productive way as much or even more than a proof.

Alice Schwarze@aliceschwarze.bsky.social · 6d ago

Just because math has celebrated proofs more than other contributions doesn't mean we don't all know and have known since way before AI that conjecture statement (and hypothesis formulation in applied math) and proof contextualization are important contributions to humanity's quest for knowledge

Just because math has celebrated proofs more than other contributions doesn't mean we don't all know and have known since way before AI that conjecture statement (and hypothesis formulation in applied math) and proof contextualization are important contributions to humanity's quest for knowledge

Itai Yanai@itaiyanai.bsky.social · 7d ago

I actually don't mind being wrong.. we said this is what's happening on the Millennium Bridge.. what we proposed doesn't work. But it was interesting. We really did stick our neck out and unfortunately got our head chopped off. –Steven Strogatz, Cornell Professor @stevenstrogatz.com

Basically my two cents on OpenAI challenging the Leiden declaration

Axe Ghost. On Steam!@axeghostgame.bsky.social · 7d ago

yes. if I say 'this is my code' about AI gen stuff, that's just shorthand for 'this exists because of my decisions and i take responsibility for it' - and sensible people understand that. this ego thrashing about authorship is undignified and irrelevant. get over yourself.

"We're all worried," as what it means to do research (in my field, Theoretical CS) seems to be shifting, and shifting fast. What to do? Senior researchers must lead by example, knowing that not everything will pan out. What I'm suggesting below may not work everywhere, but here's my own advice: 1/

Charlotte's web, cyberpunk edition

@wavebidder.bsky.social · last wk.

this is my new favorite diagram

whole thread is very much worth reading, to get a sense of the human texture of what we’re going through, including the difference between proof and explanation

Henry Yuen@henryyuen.bsky.social · last wk.

Some initial thoughts, and a complicated mix of feelings. Wow. I mean, Erdos problems are cool (I genuinely mean that), I didn't know about the Jacobian conjecture before it got disproved. But this newest batch from OpenAI hits home in a way the previous announcements did not.

There's lots of stuff LLMs can't do, but if you want to disbelieve all the recent math, you have to believe a bunch of Fields Medalists with long track records are just straightforwardly lying about the thing they care most about. Here's a recent blog by Gowers: gowers.wordpress.com/2026/05/08/a...

A recent experience with ChatGPT 5.5 Pro

We are all having to keep revising upwards our assessments of the mathematical capabilities of large language models. I have just made a fairly large revision as a result of ChatGPT 5.5 Pro, to whi…

gowers.wordpress.com

Every workplace expects employees to put there work responsibilities first for at least 40 hrs/wk. Society and all its institutions expect mothers to put their childrens' needs first 24/7. The only way to survive this is to ruthlessly reject expectations on all fronts and live with the blame.

Alice Schwarze@aliceschwarze.bsky.social · last wk.

Colleagues and friends have tried to be understanding of me as a working mom. But if you have never been a full-time working primary care giver, you won't ever be able to understand what it's like because (1) it is genuinely insane but (2) complaining about any part of it is not socially acceptable

Colleagues and friends have tried to be understanding of me as a working mom. But if you have never been a full-time working primary care giver, you won't ever be able to understand what it's like because (1) it is genuinely insane but (2) complaining about any part of it is not socially acceptable

Sarah Tuttle@niais.bsky.social · last wk.

All of this and more. I wouldn't trade my children for the world. But this country has meant that I've traded a lot to be their mom. www.seattletimes.com/opinion/wond...

I recommend this article about AI reasoning, where the author lets us in on his struggles w/ AI cognitive dissonance. Plus some priceless quotes from @rao2z.bsky.social. (My recommendation has *nothing* to do with the fact that I'm quoted in it too 😇) www.quantamagazine.org/is-ai-reason...

Is AI Reasoning Right for the Wrong Reasons? | Quanta Magazine

The idea that artificial intelligence can “reason” is more intuitive than ever. But intuitions can be wrong, and the science is far from settled.

quantamagazine.org