Related, Mike Masnick at Techdirt is iffy at best about LLMs but this is a good take

Techdirt: The Apocalypse Will Not Be Sexy

This essential nuts-and-bolts safety work also doesn’t, typically, require brilliant and manly heroes to nobly die for the ashes of their fathers and the temples of their gods. Instead, the hard work of actual safety requires things that are much less sexy: empathy, conflicting values, and a deep understanding of tradeoffs that will never fully resolve. Those qualities are not highly esteemed in the present moment. But they are critical for tackling the actual threats and the actual risks, which aren’t as flashy, and don’t make for such fun headlines. AI safety matters, but we’re unlikely to move in the direction of actual safety if all the focus is on fantastical stories involving killer robots.

Followup to Trump’s Extremely Superior renaming of “AI” (https://awful.systems/post/9709966/12701983)

RawStory: Trump scrambles to fix ‘supreme’ poll as MAGA melts down over ungodly comparison

our friends are taking Donald Trump seriously

https://www.lesswrong.com/posts/HGJSGRwjPqhDEud5Y/you-should-go-vote-for-the-maga-rebranded-name-for-ai

I can’t belive the fucking New York Post is AI skeptic now

https://nypost.com/2026/09/19/us-news/openai-anthropic-oversold-security-breaches-to-pressure-feds-into-protecting-turf-insiders/

The election for the new Riksdag in Sweden was last Sunday, and I just checked the list of all votes, including for parties that didn’t crack the 4% electoral count limit

https://resultat.val.se/val2026/RD?r=S

If you scroll down past the racist assholes at the top you’ll find RegleraAI.nu (basically Regulate AI Now) which of course has a website: https://regleraai.nu/. Usual doomer stuff, just in Swedish.

They got 390 votes, more than explicitely Nazi Nordiska motståndsrörelsen at 355.

the thread has hit the orange site, lots of spluttering indignation

https://news.ycombinator.com/item?id=49766911

points to previous submission

https://news.ycombinator.com/item?id=49737985

here’s another Rat-explainer piece

https://www.iankduncan.com/personal/2026-09-16-sex-ai-and-the-apocalypse/

(via https://news.ycombinator.com/item?id=49746654, where commenters are predictably blasé about it all… “yeah yeah doesn’t everyone know who Big Yud is?”)

in general, I don’t think EA breaking into the mainstream like this led to the outcome they wanted - there’s way more discussion about weird sex and eugenics than the dangers of Robot Satan

Some LWer wanted OpenAI to spam a warning about AI killing us all to every user of their app

https://www.lesswrong.com/posts/gJJ9YHzuBvwAXrthW/there-is-a-channel-to-900m-weekly-users-what-goes-in-it

New Yorker writer talks to another journo about the recent doom hype

https://archive.ph/2pXlU

There’s way to much hype for my liking but might be interesting.

Important to note, Joshua Rothman states he is worried about stuff like virus release, global warming and drone warfare. Among the faithful, this is strictly not Doom ,because statistically a nucleus of the human race will survive. Only 100% guaranteed lethality for every human being on Earth is x-risk.

he has a slobstack (ofc)

btw did you know the Luddites weren’t actually working class (n/m they predated the Marxist proletariat by like 75 years or so)

https://www.verysane.ai/p/against-the-luddites

enjoy this rightwinger employing multiple rightwing tropes to decry that “AI safety” is viewed as leftwing:

https://www.lesswrong.com/posts/Rx38cuCpL9hguLCDq/for-love-of-the-lightcone-don-t-partisanize-ai-safety

it’s all of course framed in the most parochially US-centric terms possible, and Extremely Online as well

maybe this grand project is undermined by the following exchange in the comments

niknoble: "It’s been interesting to see slowdown advocates worried about partisanization, when that is also a big worry of mine as an accelerationist. […]

reply by Hide “Why are you an accelerationist? Are you not concerned about everyone on earth being killed?”

reply to Hide by O O: “If you believe there’s low odds for doom, it’s a rational position.”

Good luck, safety advocates.

yeah I have nfc how bluesky and different views of it work but here’s that thread depicting our dear friends as a sex cult

https://skywriter.blue/@segyges.bsky.social/3mvom4b4dn22q

sure sounds like someone is salty they’re not being invited to the orgies

someone mentioned weird quasi-lefist SE Gyges in last week’s post, they have a LW account:

https://www.lesswrong.com/users/se-gyges?from=post_header

The new Maas biochips

So there are 2 explanations to why mssrs Altman and Amodei have decided to ‘slow down’ “AI” development:

1. it’s too dangerous, any misstep could awaken the evil Robot Satan and kill us all
2. the rate of progress in new GenAI models has slowed down dramatically, and there’s no obvious way to fix that

I’ll let you guess which explanation I find most likely.

edit I missed “blatant regulatory capture”

https://www.theregister.com/ai-and-ml/2026/09/14/big-ai-sets-out-its-terms-for-regulatory-capture-and-calls-it-pace-the-frontier/5296067

Via some rational, humane process, ideally involving a sharp angled blade attached to a heavy weight.

Lobste.rs news: after letting the DHH knob-slobblers and interested-in-the-fascist’s-newletters users weave enough rope to hang themselves with, the banhammers have arrived:

https://lobste.rs/moderations?moderator=(All)&what[users]=users

expect the next step be even more pressure on the mod team to get them replaced with people “more aligned with the mainstream”

Swedish media reports essentially the same, which shows the importance of targetting outlets like WSJ or NYT - those media orgs might not have enormous reach but they have credibility. It’s usually a safe bet to repeat stuff they have repeated.

I don’t know if there is any professional journalist in Sweden who has sat down and researched these networks. It’s not surprising if not, they’re not that big a deal here (in fact, “AI” as an election issue is basically non-existent, the election is on Sunday)

without going into the merits of the specific case, I find it highly amusing that OpenAI fucked up their big Navier-Stokes announcement which is now more about how they stole research from honest mathematicians rather than how awesome their AI is

AI is PEDs for the brain except they don’t actually enhance performance or make you feel good