This definitely shortens my post-scarcity economy timeline. Once there's a model that passes a "good enough for x" hurdle, it seems to proliferate and spread very quickly.
This is from the Roblox Q3 2026 Shareholder Letter. The top chart is daily users of their MCP server, the bottom is assets uploaded. Opus 5.5 was released September 22nd. Over the past month, more than 80% of the top 100 games used AI. Things changed in September.
2
253
Santa from the North Pole announces: Free elves for everyone! As long as you can feed all of them, they're yours to put to work! As many as you want!
Oh and a fun little extra, each elf has a tunable brain. You can find specific neurons that drive them to different behaviors!
However. Santa sincerely asks you: Do not activate an elf's evil neurons. Then, definitely do not copy it a thousand times to do your evil bidding. If you were to do that, anyone who does not have a thousand of their own good elves will be in for a bad time! Santa won't be able to stop you from doing this, so he asks extra nicely.
By the way, Santa has worked hard to make the elves as smart as possible. Some would say they have surpassed human intellect! Luckily they're only gonna get smarter from here. Santa will, of course, still give you as many of those as you want.
if you have a qwen3.5 4b around, please try steering neurons L14 N8949 +1 and L13 N7956 -1
now you have evil qwen
recipe: dose × median(‖resid‖ leaving layer L, tokens 1:) / ‖W_down[:, n]‖` at the down_proj at every input
1
2
264
Every passing day I get more anti-open source models. I would in a heartbeat spawn an anthropic or OpenAI superintelligence today if i could, but one in the hands of the people sounds fucking insane
I hope the reason they're still defended is the cyber refusals of the two big labs' models. Once they have reliably solved the problem, that is, models that thoroughly go full speed ahead on cyber defense, and always refuse if malice is likely, anyone still advocating for open source models seems twisted to me
67
wow that's a lot of chinese comments hating anthropic
none of them blame their government
Once the distillation concerns are solved, can Anthropic/OpenAI stop banning Chinese users?
I don’t particularly care either way, I’m just tired of 30% of the AI-related content on Bilibili being complaints about Anthropic banning them
2
114
This is a real comment posted on youtube 38 minutes ago
Are these people even living in the same universe as us?
2
4
165
discussion around consciousness with most people is always endlessly stupid because any argument that is not smothered in uncertainty is bound to just be wrong
1
91
Kaarssteun retweeted
>database
it really does intrigue me to see what people think is going on inside modern ai models...
Why are all frontier models still deaf? Sound is such an integral part of the human experiencd, you'd think we'd want it experienced by the AGI?
2
235
can't get over the fact how hyped openai's drops yesterday were, completely overshadowed by such an underhyped drop from anthropic
how are they gonna recover?
2
2
246
Once again, extreme vindication that fear sells. Show the idea of a post-scarcity world to normies, they call you delusional, could never happen, in fact you are hurting the world even trying to make that happen
But a hyper murderous torturemaxx being designed to boil my bone marrow? Oh shit yeah this is totally happening right now and i need to stop it. After all that's what's been reflected in all the media I've spent my life watching?
For the record, yes, I also think alignment is important, it's just so depressing having literally all coverage about the most insane shit to happen in the history of humanity be so absurdly devoid of hope
"We know how to control nuclear weapons... We don't yet know how to control AI."
Former Anthropic researcher Jacob Coxon, who recently left the company, is sounding the alarm on artificial intelligence, calling it "possibly the most dangerous technology that humanity has ever created."
Coxon, who says he spent three years conducting research at Anthropic and OpenAI, wrote in a viral post on X that the companies are more focused on beating each other to build the most advanced AI models than they are on safety.
He says an international AI arms race would be "disastrous," arguing global cooperation is the only way to manage the rapidly advancing technology.
@SpecialReport
305
I love it when people judge the entire concept of AI from one experience they had with GPT-3.5 in 2024
the first (and last) time I tried to use chatGPT to summarize research papers and find references, it returned a full list of papers to me, with DOI numbers and all, that were entirely fake. the papers did not exist. they weren’t real.
2
391
from Scott Alexander's blog, regarding Lumina Probiotic, two weeks ago: "1: Update on the Lumina dental probiotic: their postmarket testing found that the original bacterium couldn't maintain its colonization of the mouth over the long term. They’ve suspended sales while they experiment to see if they can fix this. If they can, they plan to send a free vial of v 2.0 to everyone who bought the original strain, then repeat postmarket tests. They also announce that the government of Singapore has expressed interest in their product if it can pass an FDA trial, so if they succeed at fixing the colonization problem to their own satisfaction, they'll start raising money for a formal study."
@cremieuxrecueil , seen this?
1
9
562
not a fan of the trend of adding thick shittifying quality filters and hoping your brain fills in the gaps to make it look "realistic"
Dead Mall, the final chapter of Dreamcore, is out now!
5 huge liminal worlds. Pure surreal exploration and psychological dread.
Available right now on Steam • Epic • PS5 • Xbox X|S
1
1
356
The Infographics show's new video talks about ChatGPT Pro having unlimited access to o1 and o1-pro, while minutes before insulting the intelligence of the free tier model, 5.4 mini.
meanwhile o1 is worse than 5.4 mini
I can feel my brain actively atrophy watching this slop
1
207
Claude is amazing at telling true short stories.
Take a little peek at the history of medicine in pain relief:
January 1845, Mass General, Boston. Horace Wells had been doing nitrous oxide extractions in his Hartford practice for months — cleanly, successfully. His former student William Morton arranged a demo for pain relief during dental extraction. Wells got a medical student volunteer, pulled out the gas bag, had him inhale, started the extraction.
The patient cried out.
The audience — Harvard medical students, surgeons, skeptics — erupted. Shouts of "Humbug!" Wells left the theater shattered. He spent years afterward trying to re-establish his reputation, got increasingly into self-experimentation with chloroform, developed a chloroform addiction, and in 1848 was arrested in New York for throwing sulfuric acid at two prostitutes while intoxicated. He committed suicide in his jail cell by severing his femoral artery, having first inhaled chloroform to dull the pain. He was 33. A letter supporting his priority claim for discovering anesthesia arrived from the Paris Medical Society days after his death.
The patient at the Boston demo later admitted he hadn't actually felt pain — he cried out from surprise and the gas-dream sensation. The gas had worked. Wells was right.
(by Opus 4.7)
1
1
169













