This is the real deal folks
red.anthropic.com/2026/mythos-...

There's a nice #Artemis II mission tracker here: artemis.cdnspace.ca

LOTS of data about its current position, crew activities, and whether or not the toilet is currently "Go"

A graphic showing "Pride Versioning" where the first number is the Proud Version. Then second is the Default Version. And the last number is the Shame Version.

Made very few changes to the latest build and was thinking maybe I'll be able to ship this one without incrementing the shame version (last number in a version number) and 2 minutes into testing, it's already getting bumped.

Astral to join OpenAI Astral has entered into an agreement to join OpenAI as part of the Codex team.

Wow openAI acquired astral
astral.sh/blog/openai

A line graph of the number of NSF awards in fiscal 2026 compared to fiscal years 2021-2025. The fiscal year 2026 is well below the other curves and increasing only very slowly.

NSF Update through March 13, 2026

1/2

How many ways can you say “We used ChatGPT” without actually saying “We used ChatGPT”?

📢 REMINDER: We're hiring a #DigitalHumanities Scientist to join our interdisciplinary DH Lab in Rome!

🔎 Focus: Digital art history, AI & machine learning, cultural heritage data, collaborative research

⏳ April 30, 2026

🔗 www.biblhertz.it/en/opportuni...

MediaFM: The Multimodal AI Foundation for Media Understanding at Netflix Avneesh Saluja, Santiago Castro, Bowei Yan, Ashish Rastogi

Media/ humanities ml folks might find this useful to track — I can think of a project or two ;)

netflixtechblog.com/mediafm-the-...

Our ICML 2025 workshop on Actionable Interpretability drew massive interest. But the same questions kept coming up: What does "actionable" mean? Is it achievable? How?
We're ready to answer.
🧵

This has been long suspected, but I think this is the first official accusation, right? I wonder if OpenAI has also seen distillation by those labs using their models.

AI Safety Meets the War Machine Anthropic doesn’t want its AI used in autonomous weapons or government surveillance. Those carve-outs could cost it a major military contract.

Anthropic doesn’t want its AI used in autonomous weapons or government surveillance. Those carve-outs could cost it a major military contract.

Well, this is a bit of a plot twist.

Ring 🤝 Boston Dynamics
"Put a dog on it"

🎉 Thrilled by the response to @loreslm.bsky.social! We received 79 submissions & 57 papers have been accepted.
Huge thanks to PC & congrats to all accepted authors!

Looking forward to seeing you in Rabat! @eaclmeeting.bsky.social #LoResLM2026 #EACL2026 #NLProc

On my to read list:
annastansbury.github.io/website/Stan...

The 5th Generation, Evaluation, and Metrics (GEM) Workshop will be at #ACL2026!

Call for papers is out. Topics include:
🐟 LMs as evaluators
🐠 Living benchmarks
🍣 Eval with humans
and more

New for 2026: Opinion & Statement Papers!

Full CFP: gem-workshop.com/call-for-pap...

Underneath the name Charlie Hebdo is a pile of bloodied dead bodies representing US citizens killed by ICE. The pile of bodies are collected in one corner to represent the stars of the US flag and the streaks of blood caused by them being dragged across the ground form the red and white stripes. In the foreground an armed skinhead officer with an ICE vest is dragging his latest victim with multiple gunshots to join the pile. The world is watching in horror.

Wow.

The front page of tomorrow’s Charlie Hebdo magazine in France.

Edited by Taylor Arnold, Margherita Fantoli, and Ruben Ros

📢 The #CHR2025 proceedings are out!

97 papers, ~1600 pages of computational humanities 🔥 Now published via the new Anthology of Computers and the Humanities, with DOIs for every paper.

🔗 anthology.ach.org/volumes/vol0...

And don’t forget: registration closes tomorrow (20 Nov)!

Getting aligned LLMs to help with misaligned tasks via context engineering, otherwise known as “Ender’s gaming”

New Report Calls for Faculty Control in AI Decisions Today, the AAUP released a new report, Artificial Intelligence and the Academic Professions, sharing survey findings and calling for the establishment of policies in colleges and universities that pri...

Today, the AAUP released a new report, "Artificial Intelligence & Academic Professions." The report calls for policies that prioritize economic security, faculty working & student learning conditions as AI tech accelerates.

www.aaup.org/news/new-rep...
🧵1/9

Just to get back to where we were as of January 2025, we will need Manhattan Projects across multiple agencies. This particular loss will produce exponentially increasing human injury and economic damage to the U.S. as the ability to investigate and to regulate falls apart.

Japan has started rice rationing in stores as another harvest gets decimated by extreme heat. The future is grim for island nations that are directly exposed to sea surface heat spike.

Derrick Van Orden on the restricted Capitol grounds on J6. He denies being there.

This is Derrick Van Orden on the restricted Capitol grounds on J6. He denies being there.

Now out in Nature: we analyzed 19,000 computer vision papers and the 20,000+ patents citing them and found that most CV papers are used in surveillance enabling CV patents. www.nature.com/articles/d41...

mmigration and Naturalization Law The principles of national constitutionalism have a relatively straightforward application to immigration and naturalization law. If the government may not destroy or replace the People, or otherwise usurp their sovereignty (or permit their sovereignty to be usurped), and if the People are an identifiable, substantive, and organic legal entity (and not merely the statistical collection of all people within the United States), then government power to permit immigration or naturalization is not plenary. National constitutionalism therefore advocates subjecting to judicial review—with the highest level of scrutiny—all attempts by the federal government to “dissolve the people and elect another’”” through immigration and naturalization policies which have the effect of altering the traditional, racial demographic balance of the United States and thereby degrading the People’s power and privileges. Accordingly, Taney’s dictum in Dred Scott that Congress “may, if they think proper, authorize the naturalization of anyone, of any color, who was born under allegiance to another Government”? is anathema to the implicit limitations, entailed by popular sovereignty, placed upon the constituted power of the nation-state and should be rejected as a matter of law. The effect of this would simply be a return to the pre-1870 status quo whereby only Whites could typically immigrate and seek naturalization.79 Non-Whites would 76 Id. at 323 (citing ROBERT BARNHART, THE BARNHART CONCISE DICTIONARY OF ETYMOLOGY: THE ORIGINS OF AMERICAN ENGLISH WORDS 397 (1995)). 77 Bertolt Brecht, The Solution (1959), reprinted in BERTOLT BRECHT POEMS 1913-1956, 440 (John Willett and Ralph Manheim, eds., Methuen 1976). 7860 U.S. at 419. 79 This is a historically workable standard. See generally IAN HANEY LOPEZ, WHITE By LAW 2-7 (2006) (describing various court cases in American history aimed at determining if the petitioner was White, and explaining how “[t]hough …

Back to Damsky's paper which, again, WON AN AWARD FROM A FEDERAL JUDGE, I did want to highlight the passage in which he wrote that Justice Taney wasn't racist enough in Dred Scott, that immigration should be limited to whites only, and that we could use AI to figure out who the white people are. 3/

I'm going through UF Law book award winner Preston Damsky's twitter account, and it's worse than what the NYT reported. He thinks that Stephen Miller isn't racist enough. Because he's a Jew, and the "Jew clique ... is bringing America to ruin."

x.com/preston_terr...
x.com/preston_terr... 1/

How often do LLMs snitch? Recreating Theo’s SnitchBench with LLM A fun new benchmark just dropped! Inspired by the Claude 4 system card—which showed that Claude 4 might just rat you out to the authorities if you told it to …

New benchmark just dropped: SnitchBench by Theo Browne tests if LLMs will snitch on you to the authorities if you feed them incriminating documents and a tool that lets them send email, as seen in the Claude 4 System Card

Turns out they pretty much all will! simonwillison.net/2025/May/31/...