It took me 1 try to guess who @andymasley.bsky.social was @mystery-mutual.bsky.social mystery-mutual.fly.dev

It took me 1 try to guess who @kabirkumar.bsky.social was @mystery-mutual.bsky.social mystery-mutual.fly.dev

It took me 3 tries to guess who @scedmunds.bsky.social was @mystery-mutual.bsky.social mystery-mutual.fly.dev

It took me 2 tries to guess who @eugenevinitsky.bsky.social was @mystery-mutual.bsky.social mystery-mutual.fly.dev

It took me 4 tries to guess who @why.bsky.team was @mystery-mutual.bsky.social mystery-mutual.fly.dev

It took me 1 try to guess who @moskov.goodventures.org was @mystery-mutual.bsky.social mystery-mutual.fly.dev

That’s just some of what these massive corporations and billionaires are getting for cutting a check for Trump’s ballroom. If it looks like pay-to-play corruption, that's because it is. My committee has been investigating this for months, and we will hold everyone accountable.

Have my AI folks been following what's happened to biomedical research funding? diabetesjournals.org/care/article...

Contest for building a detector for AI lying.

After two years of forced MacOS usage, I'm back on a Thinkpad. Native Linux is great, but WSL makes it a non-issue. Docker vs venv, basically, and you get right-click, plus a file manager that's designed to manage files.

We made CNN, y'all!

In case you haven't heard, the NIH, prompted by Vought (architect of Project 2025), wants to get rid of peer review and put grant funding decisions under federal control. The stated rationale is to reduce political influence on scientific funding. 🤔 lite.cnn.com/2026/06/04/c...

It’s not “do good work and be successful.” It’s doing good work, talking about the work, figuring out how to get your work to the right people, accepting that good work is slower than bad work and finding a path through that anyways

🤔

On OMB's proposal to make all grants political: "Speak up! Now’s the time! Standing by the side of the room, glancing around nervously and hoping that somebody else does something will lead us to disaster. These are not normal times, so stop acting as if they are." www.science.org/content/blog...

Upset about the unlawful Vought/OMB proposed rule? Don’t think we should kiss peer review goodbye & let politicians fund science based on patronage & whims? Beyond commenting & calling congress, you can also add your voice to the call to remove Russell Vought. actionnetwork.org/forms/civil-...

Some political person from HHS who clearly doesn’t know which way is up, is reviewing NIH grants and making absurd suggested changes. This is on top of a set of discriminatory changes required by NIH. www.science.org/content/arti...

OMB just published 412 pages in the Federal Register (Docket OMB-2026-0034) giving political appointees authority to terminate any NIH grant, override peer review scores, and block you from publishing with federal funds. Comment period closes ~July 13. Thread 🧵

and they called it a waste of tax dollars back then, too

The mindset that turns one guy writing a sardonic editorial into "these men running around" and "fucking resign" is what allowed Vought and Co to gain power. The absolute humorlessness playbook has failed harder than I've ever seen anything fail, and I've been to a few little league sporting events.

This is why a change in "agency priorities" should not terminate an awarded grant. Nobody knows where the next transposon, GFP, or Taq polymerase kind of discovery is going to come from! Science is not engineering!

Nothing is as unscientific and harmful to innovation as killing the messenger, and that's what these OMB rules are doing. Use this template to explain why the proposed rules must be scrapped.

It was great working on this with AWS aws.amazon.com/blogs/public.... I’ll be speaking a bit about it at FORCE 2026 this week. #force2026.

Kudos to Perplexity on this move. OpenAI and Anthropic will need to do something like this too, or they'll lose the researcher demographic, and that will have an outsized impact on adoption throughout industry. www.stm-publishing.com/ebsco-inform...

Ten places where Magnifica Humanitas matters for AI. At 42k words long, Pope Leo XIV’s new encyclical has a lot to say. In our most recent Digest, Mitchell Howe outlines the parts which might be the most impactful.

@michelleding.bsky.social, @kamachee.bsky.social, Rui-Jie Yew, and I made an annotated bibliography compiling academic/journalistic work about the mechanics of the AI NCII ecosystem and the roles that companies play in it. It's public. Check it out if you'd like! docs.google.com/document/d/1...

UK MPs were just selected to introduce bills of their choice. And we've published the updated version of our bill to ban superintelligence we presented to Number 10! If introduced, it'd be a world-first. Ask them to introduce it using our contact tool: ukbill.controlai.com

Science Needs 3rd Spaces We envision scientific third spaces as “collaborative hubs where researchers, artists, engineers, and practitioners co-develop methods, share resources, and pursue projects driven by curiosity". authors.elsevier.com/sd/article/S...

I could be convinced that the main reason Grok exists is to bring down onerous regulation on the whole industry, as revenge.

If what's happened to journalism given the reduced barriers to entry is any guide, we should expect AI in film making to result in more fringe productions. The mainstream stuff already has the funding it needs. This could be good or bad - we decide which!

I thought Cassidy was supposed to keep RFK in line? Whoops!

For the EU crowd: The CEN-CENELEC JTC 21 working group has released the “AI Risk Management” standard for public Enquiry. This requires providers of high-risk AI systems to establish, implement, document and maintain a risk management system. adamleonsmith.substack.com/p/pren-18228...

I spent the past semester building interactive teaching apps with AI-assisted coding. I used my personal website to share them with my students, but that makes them hard to find for other educators. So I built Galvoro — galvoro.app

Interesting news from Consensus.app: licensing deals with major publishers. This was a big gap in their offering, and particularly important because they are more focused on the general public, who didn't know that gap existed.

ArXiV has a new LLM policy (Screenshots with alt text so you don’t have to click through to the other place and see all the stupid responses)

Introducing AI StopWatch. It’s easier to speak up when you know what’s going on. It’s easier to see what’s coming when you know what’s already here. (1/9)

Wherein SMBC discovers metarationality.

AI companies have freeloaded off Internet Archive and now publishers are cutting it off. This is what happens when people freeload off the commons. blog.archive.org/2026/05/06/w...

Gotta know the underlying frequency to make any sense of accuracy rates (this goes for lab tests, too.)

NetHack is one of the most complex and longest-lived open source programs ever written, and after 46 years, v5.0 shipped today. www.nethack.org/common/inde... And ... it is a VERY cool large codebase to work with in the LLM era.

The time it would take a peer reviewer to determine that this is nonsense exceeds the time it takes to create it. I don't see how the system holds up much longer.

It's still weird to me that there are such differences between indexes. How have we as a society not solved the issue of just indexing everything that has been published, all together, in one catalog?

I've been warning people in AI safety about dark tactics for a while. Happy to jump on a call to discuss defense & how to get the most out of a media engagement.

In other China news, an employee is told that AI has made his job obsolete, gets laid off, sues the company for wrongful termination, and wins. Now that's the way to rage against the machine! www.npr.org/2026/05/01/n...

PauseAI wants to hear from you! Is AI affecting your work, education, security, creativity or daily life? If so, tell us your story! More information below: #pauseai #ai #AIsafety #artificialintelligence

A recent (short) talk of mine is now up on YouTube. It is about why, if you are working for an AI company on making its systems safer, I think you should consider quitting. www.youtube.com/watch?v=Ugu7...

New paper w/ UK AISI: Millions of people now use AI to help them write and communicate. In three experiments (14k participants, 3m+ human ratings) we show that AI writing assistance systematically distorts writer personas – their perceived beliefs, personality, and identity. 🧵

Safety and innovation are not mutually exclusive: for many companies, especially in high-trust industries, AI’s risks are also a hindrance to adoption. Europe needs to invest in safe, reliable AI to avoid hitting the adoption wall. My op-ed in the @financialtimes.com: www.ft.com/content/bc29...

I’m in Madrid this week for the first in-person meeting of the United Nations Independent International Scientific Panel on AI. I'm thrilled to be part of this distinguished group of 40 international experts appointed to provide an evidence-based scientific assessment of the state of AI.