ICE ended up returning the man, Saly, after realizing he’s a fucking US citizen with no criminal record, per his sister-in-law. These fucking animals.

Paypals generative ai help desk is genuinely one of the least helpful support options I've ever had the displeasure of using

“It was like all the noise of the rest of the world was gone, and it was super intense,” said Daniel. “Just me, God, the sky, and AI.”‘

I wrote about a fake child care scandal, the connection to the murder of Renee Good, and the right-wing vengeance ecosystem.

i know people want “abolish”, but this would be a massive improvement over the status quo and we can do this and then abolish them later

oh fuck it, free to read www.patreon.com/posts/alrigh...

And Renee Good’s last words appear to be “We’re not mad at you.”

I would really caution anyone against implying that paywalling this feature is a legitimate response to public outcry. The paywall is an endorsement, not a restriction. Musk is saying "this thing you're all talking about is valuable. Pay me for it." The actual material effect is just monetization.

I'm glad to see this pressure on Google and Apple. If any other company was running an active CSAM generator in public you better believe it would have been removed from their app stores by now www.wired.com/story/x-grok...

We live in an era where the powerful appear to have lost all grip with reality - and yes, I do think the dual forces of the Internet and AI have much to do with that. My latest here (and I sure wish current events related to this topic would stop escalating):

Bari will never explicitly say "it's fine for masked agents of the state to murder civilians." But she will say, again and again, that the people who are upset about masked agents of the state murdering civilians are hysterical, illegitimate, censorious liars.

Emily Heller, an eyewitness, is speaking out to set the record straight. Officers were screaming at Renee Nicole Good, “Move, move, move.” So she moved, and they shot her multiple times? Then Noem and Trump branded her a domestic terrorist. Does everyone see how absolutely insane that sounds?

A thread of ICE and CBP sex offenders and credibly accused sex pasts. Content-warning descriptions of abuse Yuma Sector CBP agent Ramon Marquez, 31, was arrested in May 2025 and has been charged with 15 felonies including 14 counts of sexual conduct with a 16-year-old between Dec 2024 and Apr 2025.

Also, look: a sad fact is a lot of Americans cannot extend their empathy to all of humanity but can extend it to some people who are more similar to them. I agree this represents a moral deficiency but it's more urgent for us to take power than it is to correct the moral failing of many Americans.

You do *not* want to be a carrot in the ocean.

If you have access to an llm not for work. Can you ask it what teams the main characters in heated rivalry play for. Can prompt "ilya and Shane"

remember when Grok said some progressive shit and Elon immediately revised the code?

Slay the Princess - The Pristine Cut Collector's Edition (Switch) is $169.19 on Amazon buff.ly/kOf39Zp #ad

also, @reuters.com continues to set an example of irresponsible reporting

*how* grok produced porn and CSAM shouldn’t be of primary importance/concern to regulators. rather, they should treat this as evidence of failure to comply with rules and proceed to enforcement i.e. ban the platform till X shows it complies with regulations

time until listening to "this year" by the mountain goats this year: 73 hours 54 minutes

The degree of ideological capture that LLMs have achieved in EdTech is disappointing but not surprising. More surprising (+ more worrying) to me is the degree to which the tech is being embraced uncritically by the academic institutions themselves.

"I’m going to walk y’all through why this is very bad, and I’ll attempt to make it as clear as possible to outline the unnecessary risks and complications Trump is now forcing upon our country and the people of Venezuela and international stability."

Google “said its AI Overviews, which use generative AI to provide snapshots of essential information about a topic or question, are ‘helpful’ and ‘reliable’”. “…some of the summaries, which appear at the top of search results, served up inaccurate health information and put people at risk of harm.”

I will go to my grave arguing OpenAI rushed GPT 3.5's release without having done adequate safety analysis (not even sure what that would've looked like tbh) and instead offloaded that work onto the public. And that set off an arms race among LLM developers that none of them were prepared for

great journalism means ascribing reliable narrator headline status to an algorithmic parrot built by pedophile white supremacists

The AI-boosting CEOs and the self-declared doomers are all part of the same hypocrisy. Also, the Guardian item takes Eliezer Yudkowsky seriously, which is a thing that no writer should do. awful.systems/post/5677395

MAIHT3k, Ep 69: www.buzzsprout.com/2126417/epis... I thought it would take at least some effort to tear this paper-shaped object apart. Instead, it was so flimsy it disintegrated under the force of @alexhanna.bsky.social & I simply … reading it. Thanks to Ozzy Llinas Goodman for production!

🧵 Our mutual aid hub has been a HUGE draw for our immense volunteer base, many of whom have never wanted to be involved politics before now. This model isn’t just about the immediate impacts: It’s about showing our values, building a community, and reaching voters — without selling our soul!

I've not had time to read this but you might find it relevant: pivot-to-ai.com/2025/11/24/d...

This is such an important point, and it is so sad that it needs spelling out, and still many academics mesmerized by “AI” do not get it irisvanrooijcogsci.com/2022/12/29/a...

This is NGC 2264, also called the "Christmas Tree Cluster." It's a cluster of young stars about 2,500 light years away. They were 'born' during the Pliocene Epoch (2-5 MYA), about the same time as early hominids like Ardipithecus or Homo habilis, so they're barely older than humanity. 🎄💫🔭

mark cuban waking up on christmas tomorrow fully a communist. in response to confused bluesky users inquiring about his change of heart, he merely gives a cryptic answer about 3 ghosts

I'm still able to download it, but just in case I've re-uploaded it here drive.google.com/file/d/1GRcg... and here drive.proton.me/urls/X8C14W3...

It's done.

This is a dumbfounding guideline. Absolutely a destructive and incompetent recommendation that was clearly made by people not qualified to judge such decisions.

good video, by the way, about an incredibly bad (and, a short time after this video, Oscar-winning) documentary

If we are banning cell phones for kids we need to be talking about banning chatbots for boomers www.persuasion.community/p/my-chatgpt...

you can't just "oops, not that one" after they're out! journalists have them! they're numbered sequentially! come on!

NEW: Mystery AI Hype Theater 3000, Episode 68: This is What Algo-cracy Looks Like Tech bros say automation can strengthen democracy. @alexhanna.bsky.social and @emilymbender.bsky.social unpack their flimsy arguments about the democratic process. www.buzzsprout.com/2126417/epis...

i thought i had 206 tabs of christmas shopping but at some point i stopped christmas shopping and started googling weird trains

I wrote this brief talk on why “augmenting diversity” with LLMs is empirically unsubstantiable, conceptually flawed, and epistemically harmful and a nice surprise to see the organisers have made it public synthetic-data-workshop.github.io/papers/13.pdf

Screams

This is the most difficult one btw

🚨 THE BATTLE BEGINS 🚨 Your favorite year-end contest is back. It’s time to choose who will be this year’s Worst Person In Tech for 2025! Each day of this week new matchups will drop until we choose the winner on Friday. 🗳️ Cast your ballot: twsu.forms.app/worst-person...

When we're all standing in the ruins of the AI bubble we'll look around, and at each other, and realize nothing of value remains. No value will come of this bubble. Nothing.

They’re redefining “productivity” to mean “how many messages do you send the ChatGPT,” and calling those who are overly-dependent on an LLM “frontier employees”:

“While the AI industry claims its models can “think,” “reason,” and “learn,” their supposed achievements rest on marketing hype and stolen intellectual labor. In reality, AI erodes academic freedom, weakens critical reading, and subordinates the pursuit of knowledge to corporate interests.”