It is a good time for moonshots. AI has reached a level where there are transformative projects that could result in huge social good, but require public R&D, consensus & transparency to pull off. Examples of such projects: universal tutors, co-scientist/replication systems, and remote medical help

This is exactly the kind of stuff we would love to see written and submitted to the Protopian Prize! protopianprize.com

Rio's municipal IT cooking up frontier models, heck yeah. Need more of that energy.

First time US govt blocks access to a model and it is on flimsy grounds against the company with the temerity to refuse them and the most safeguards on users. AI will not be a tech of liberation or empowerment unless there are local models and many actors at the frontier. Our access is so fragile.

There’s so much to be disappointed/disgusted/outraged by with how the AI industry has been operating. But we don’t have to accept this as the only way of doing things. We can imagine other AI futures—thriving, democratic futures. Write #PublicAI futures for the Protopian Prize! protopianprize.com

CrankGPT. “Just a hand crank, a little computer, and a small stack of speech and language models running locally. Provided the electronics are kept dry and at a reasonable temperature, there’s no reason this thing won’t still work in a thousand years.” [squeezlabs.github.io]

Birds in Ukraine getting gigabit before rural America

Lol. Hey AI checkers, you're not taking my freaking —'s away. Before AI existed I had set up keyboard combinations to make them, because that's how I write. So this is why they say it's 100% AI. Because 6 emdashes, 1 triad, and 1 ai phrase ("relentless"). Fucking hell.

We all know that there is a bunch of illegal stuff happening in this admin. We we’ve seen ICE unlawfully arresting & brutalizing people in broad daylight. So many crimes are being committed. We know this But what if we can’t PROVE it? It may sound strange, but it may happen if we’re not careful 🧵

The new Obernolte-Trahan bill does frontier catastrophic risk evaluation quite well... but the preemption clause is far too broad. Overall, I think it is more like the "Pretty Good American AI Act" instead of the "Great American AI Act." publicknowledge.org/public-knowl...

what i mean is that something needs to get done about AI. the thing itself. not the buildings that the models are served out of.

someone made a fork of opencode that routes through the unsecured ai endpoints from chipotle

Claude now uses the qualifier "honest" in almost every output, sometimes more than once. Sounds like Anthropic has a gun pointed at its server rack constantly.

People don't long for "inconvenience." They long for permanence. And though streaming services may be convenient, people are waking up to the fact that we are paying rent-seeking middle men month after month for things we will never actually own.

when i need to justify my anti-IP position in the future i will tell people it is because i am catholic

AI is merely the natural conclusion of tech, so you can predict a person’s AI fears by taking their unspoken fears about tech for the last 20 years and 100x’ing them

🎉 Today, @lujan.senate.gov and @markey.senate.gov introduced the “Digital Opportunity Foundation Act” to establish a nonprofit foundation to help close the #digitaldivide by leveraging public and private investments. @dorismatsui.bsky.social introduced a companion bill in the U.S. House.

R FCC chairs and embarrassing coffee mugs, name a more iconic duo

We've become numb to this stuff, but a president accusing a journalist of treason -- a capital crime -- for reporting on an ongoing war is not normal, and we should never accept it as normal.

🧵 Wayback Machine director Mark Graham is thanking the hundreds of journalists speaking out in support of the #WaybackMachine and the importance of preserving the online historical record. 🌐📰 Read Mark's letter ➡️ blog.archive.org/2026/05/11/a... @mark.bsky.social #WaybackMachine

Thoughtful thread on a topic that is too often simplistic and reactionary.

Glory, glory, hallelujah

Toxicity dressed up as personal positivity/affirmation. We have shifted into an era of vice signaling. It may seem small given the scope and scale of issues in the world, but the ubiquity of selfishness and casual cruelty has infected everything from personal relationships up through politics.

It’s World Press Freedom Day, and it’s important to make this point: the freedom of journalists isn’t only the freedom to write. This is the first WPFD in 30 years that the news at major media outlets isn’t being preserved. savethearchive.com/newsleaders

I didn't need a stronger reason to support UBI, but creating an explosion of Spinoza scholarship is certainly a plus for me.

World's strongest ML engineer vs world's weakest LLM whisperer

HAPPENING NOW: Preemption Of State AI Laws Is Not The Path Forward. americanprogress-org.zoom.us/j/86315035925

AI data centers are increasingly reshaping local economies, infrastructure, and environments. But do they actually provide anything of value to the public? Read the latest from @natpurser.bsky.social: publicknowledge.org/what-to-ask-...

My faith in the legal system will be greatly bolstered if plaintiffs take this opportunity to completely lay waste to Mr. Beast.

🎁

Any EA, AI accelerationist, or AI for Good person that thinks working with this admin is distasteful but smart, must make a clear account for how that concept works with such fundamentally anti-scientific and regressive elements like Kennedy in place.

Even better: stop treating tech monopolies as if they're an immovable part of the landscape which we can only accept, ban, or regulate. We have the means to replace toxic social apps with better ones, including youth-designed and youth-governed alternatives.

Good analysis from @natpurser.bsky.social on the WH Concept of a Plan for AI legislation.

*chuckles smugly* all it does is next token prediction

Yes...ha ha ha...yes!

"hot, sexy, dangerous boys are not covered by copyright" Beautiful. 🥲

Happy St. Patrick's Day, New York.

In all sincerity: Because these are public documents and anyone trying to limit the availability or use of public documents should not be in journalism.

I'd love to see someone dig even further into the linguistic theory here, but I have to say that this post from @segyges.bsky.social reflects my skepticism of the "Stochastic Parrot" paper very well.

What are the best AI/LLM tools for collaboration? A lot of things seem to lend themselves to individual use, but I’m curious to explore more what is built to support teams and collective empowerment, decision making, etc

✋ Former USG war crimes lawyer here. Apropos of SecDef's remarks this morning: Denial of quarter—even the declaration of no quarter—is a war crime. And recognized as such by the US Government. From DoD's Manual for Military Commissions.

If LLM consciousness is interesting to fight enough it should be interesting enough to read seriously about, right?

This is a common misconception, but it's incorrect, on two counts. 1: SCOTUS' denial of a cert petition leaves a lower court ruling undisturbed, but that's not the same as the Court affirming the decision below - either technically or colloquially. 🧵

This is a really incredible thread.

This is Trump's former AI adviser

⚠️ Anthropic: “We cannot in good conscience accede” to the demands of the Department of War. Concerns about the Department’s potential use of AI for mass surveillance and fully autonomous weapons have not been resolved.

Ah yes, the preferred alignment method of Robert Edwin House.

Yesterday morning kids were sent running from their bus stop in panic because ICE showed up. This guy at today’s ICE Out of Lindenwold protest is a must watch 😭. @maddow.bsky.social

When a sitting representative refers to the Super Bowl halftime show as “pornography,” people should use that to reflect on how they and other representatives are simultaneously working to ban, censor, and restrict “pornography” and what that actually means

This project is a truly great resource. Aspen Digital's Emerging Tech team is the best around for insightful and accessible information about AI and this handbook has an amazing line up of expert authors. Check it out!