(I get the feeling we’re close to this bubble’s end.)
Anyways, welcome to the Stubsack, your first port of call for learning fresh Awful you’ll near-instantly regret.
Any awful.systems sub may be subsneered in this subthread, techtakes or no.
If your sneer seems higher quality than you thought, feel free to cut’n’paste it into its own post — there’s no quota for posting and the bar really isn’t that high.
The post Xitter web has spawned soo many “esoteric” right wing freaks, but there’s no appropriate sneer-space for them. I’m talking redscare-ish, reality challenged “culture critics” who write about everything but understand nothing. I’m talking about reply-guys who make the same 6 tweets about the same 3 subjects. They’re inescapable at this point, yet I don’t see them mocked (as much as they should be)
Like, there was one dude a while back who insisted that women couldn’t be surgeons because they didn’t believe in the moon or in stars? I think each and every one of these guys is uniquely fucked up and if I can’t escape them, I would love to sneer at them.
(Credit and/or blame to David Gerard for starting this.)
Jensen Huang discussing a comeback to Nvidia with his supporters
Mate, YOU RESIGNED. None of this ‘uhh actually I wanna go back’ bullshit. I’m sure Nvidia’s board would love to take back the CEO who’s getting investigated for violating export controls
“Don’t quote corporate governance norms to MEN WITH GPUS!”
Completely forgot about the interactive AI model forecast graph from Kokotajlo and friends
Nothing’s changed aside from a small disclaimer linking to their most recent blog post (discussed here)
Only thing noteworthy is that Kokotajlo’s model has an AI Coder that “can fully automate an AGI project’s coding work, replacing the project’s entire software engineering staff”, so get ready!
Peeked back at their blog to see if I could find anything about it. Text dump incoming:
The model with Daniel’s latest median parameters predicted that when an Automated Coder (AC) was developed, AI software R&D uplift would be ~2.7x and the pace of progress would be accelerated from the 2024.0-2025.5 level by ~1.25x.
So naively, if we’re at a 1.7x AI-attributed pace of progress multiplier, we might think that AC has been achieved (as of Model 4).
Based on the METR report and Anthropic’s Fable 5.2 system card, it seems unlikely but not totally implausible.
In the Fable 5.2 system card: Anthropic says: “we have been using Claude Mythos 5.2 extensively in our own day-to-day research and engineering during the pre-release period, and it does not seem close to being able to substitute for our Research Scientists and Research Engineers.” This is the strongest evidence against AC already. On the other hand, METR surprisingly doesn’t rule out full AI R&D automation: “We are uncertain, but think it’s unlikely that [Claude Mythos 5.2] can fully automate the AI R&D process at Anthropic.” Fable 5.2, based on its AECI, likely represents “Model 3,” which is the second-most-capable internally deployed Anthropic model.
Idk man that seems like that pretty solidly means AC did not happen
Relevant interview quotes from the METR report include: Being “no longer bottlenecked on code,” “SWE acceleration… is extremely substantial.” “When people present experiments, they most often just point to an [agent scaffold] thread. Experimentation overhead is driven towards zero super fast.” There’s also the report’s assessments on automating researchers beyond coding:
“We were told that R&D is accelerated primarily via AI assistance rather than fully autonomous contributions, but there are some signs that fully autonomous contributions will play a larger role soon”. “Simply put, Anthropic tries hard to substitute researchers out for models, but researchers are still required.” Note that humans could still contribute some alongside an AC: the criterion is that only the AIs could do all the coding as well as only the humans.
ok and?? if only the AIs could do the work then they wouldn’t be having trouble substituting out researchers for AI.
In the report’s chart showing “Model 4 achieves >50% on all of its benchmarks”, there is a software engineering benchmark on which Model 4 gets ~52%, and another around ~75%. Surprisingly, on the 52% one, scores have been getting worse since Mythos 5. This makes me discount it substantially, but still it’s informative that Anthropic chose to include it in the AECI. The progression for the 75% one looks more reasonable; Eli thinks these benchmarks provide weak evidence against AC. (Not strong evidence, because humans might also score well under 100%.)
My take is that if there was an AI model that could automate all the coding on Anthropic’s AI projects, it would be absolutely plastered all over the news. Especially with their IPO looking shakier and shakier by the day. Plus if they’re apparently trying to replace researchers but can’t that’s a pretty strong indicator that AC isn’t here
I am sick and tired of all this benchmark business. It really feels like people staring at tea leaves and bird patterns and sheep guts to see if they line up with the prophecy written in the book some random shady oracle sold them 200 years ago. If they wanted a clear picture, why don’t they just ask a practitioner in the field? Ask a programmer if software has become very reliable and user-friendly, and if codebases are well-structured and easy to maintain and not spaghetti nightmares. Or they could just open their eyes and see Microsoft and the FAA shitting their pants on two consecutive days. But I guess it’s not Rational if it doesn’t let them make up 37 numbers.
It’s like they’re auditioning for chief jargon guy “If only we could modulate the tachyon flux by disharging the deflector dish, we could form a rudimentary turboencabulator. We benchmarked the dilithium crystal matrix, which bottlenecked our capacity to inhibit exponential takeoff, but indicators are negative.”
if we word salad hard enough we’ll manifest the Robot God
“I like the idea of trying to integrate AI components into guitars, where you can prompt a guitar to sound a certain way. I’d like to explore that.“
Matt Bellamy, frontman of band Muse
I see that my unawareness of a certain aspect of pop culture has spared me from disappointment yet again.
not muse pls
unfortunately he is a fair bit of a twat. I was also disappointed when I learned about it a few years ago
My face when reading awful this week (positive):

The funniest part of Jensengate is seeing all the techbros being really sad and petitioning to restore Jensen to his rightful position, thinking that there is some analogy to Sam Altman. They seem to not realize that the reason Saltman was fired was because the board didn’t like his stupid face. Meanwhile, Jensen has committed three affairs, two with employees, and somehow that is the less important thing, considering he smuggled tens of billions of dollars of chips illegally into China. I wouldn’t be surprised if he ends up in jail. If Jensen thought he had a reasonable chance of surviving, why would he resign and start whining like a teenager?
They even built a vibe coded petition website with zero user validation whatsoever:

And also finally the circular financing stuff is making headlines, probably because they think they can pin it all on the one guy. Even the executive CEO is explicitly described as someone who was apprehensive with the whole propping up startups long enough for them to order a gazillion GPUs thing.
I think the board drama at OpenAI was the EA cultists (led by Ilya Sutskever) realising that Sama wasn’t a true believer but a cynical capitalist - the real life embodiment of the Simpsons monorail salesman. They realised he wasn’t going to take their fairytale about an omnipotent digital sneky seriously and tried to fire him. He was able to get back in due to support from his tech industry allies who are also uber capitalists (like Nadella)
the reason Saltman was fired was because the board didn’t like his stupid face.
He did lie to them and tell different people different stories to try to manipulate them. But you are right among the things he lied about was AI “safety” related stuff, and (as samvines describes) he lied because he wasn’t a true believer like they were. (And thus tried to bypass their oversight in releasing ChatGPT.)
Lukewarm take, considering that LLM models could use a lot more thought and care in their development and release (although obviously not for the reasons the doomers want), the EA cultists were actually the (relatively) good guys in that whole saga of Sam Altman’s firing and rehiring.
I don’t know, I’m not gonna root for an arsonist on the Hindenburg even if it is full of Nazis.
under no circumstances you “gotta hand it to them”
full article about it here apparently over 6800 Nvidia employees have signed, but then again no user validation so we have no idea what the real number is
Nvidia deliberately kept GPU prices high through circular financing deals, report claims
pretty sure we all knew this one, but what the hell more evidence doesnt hurt
Why not simply make Claude the ceo of Nvidia?
#chatCEO already!
why can’t Rats get laid, reason #12,906
Almost everybody attending a rationalist event is going to be some brand of cosmopolitan internationalist nerd, whether they accept it or not.
Words fail me. A few CFAR people had contacts in Czechia and Germany but most of these people can’t dare to explore exotic subcultures in their own city, and race pseudoscience is central to their culture.
I think there’s an inbuilt unexamined assumption that they are automatically part of a cosmopolitan elite, thanks to their paths through early life, and they are now transcending beyond that. Standing on the shoulders of David Brooks, perhaps
At best, I think “cosmopolitan” means “wants to stay at a Hilton on three continents, and gets off to Japanese and American porn.”
Someone diagnosed the Post-Rationalists as bring resentful that their educations and class backgrounds did not bring them the wealth, power, and status they felt they deserved.
…Standing on the shoulders of David Brooks, perhaps
Old: Taco Tuesday
Bold: Savage Saturday
Rationalists blurt out racist stuff
P(rationalists are racist) ~ epsilon and since P(say racists things) * P(I observe rationalists say racist things) is very high we must still deeper.
P(say racist) = P(are racist) * P(other reasons) ergo P(other reasons) must be very high. So they must be impressing women.
I have pretended to do science. Please clap.
One comment at the moment:
Or maybe they’re just racists, and trying to impress a woman is a convenient excuse?
With, shockingly, three out of three upvotes (but one agree balancing one disagree)
Possibly the modal jarring experience is that they’re talking alone with a nice man who seems pretty normal (if a little awkward), and that guy will just inject a conspicuously racist comment into their conversation. Not like a casual comment that seems unfortunate in context, or an edgy comment in a discussion that was already dancing around race, but something completely out of the blue.
How could such epistemically healthy individuals possibly be racists, literally unthinkable.
I know what actual chuds are like. That being the case, when I see this behavior from rationalists firsthand (again, usually in front of a woman), it often strikes me as bizarre and performative.
Then a few paragraphs later he starts delineating the types of racism he finds acceptable:
people making these comments are usually not “actually” racist. Almost everybody attending a rationalist event is going to be some brand of cosmopolitan internationalist nerd, whether they accept it or not. And if they’re not, they’re still not going to be racist in the way that a guy from San Antonio whose relatives used the N word at the dinner table is racist. Their “racism” was internalized as a result of a passive and NPC-like relationship with social media, and is fundamentally uninteresting to outside observers.
Basically if you were being racist (charismatically) I don’t think OP would have a problem.
Went for “classy” racism but got classist instead
🎶 Can’t shake the Devil’s hand and say you’re only kidding 🎶
A Quality Shitpost™ from fedi regarding recent events, courtesy of Jared White:

Alt-text, to maximise shitpost transmissibility
A screenshot of a Simpsons episode (don’t know which one), featuring Kwik-E-Mart owner and controversy magnet Apu (labelled “Weird Nerds”) jumping in front of a would-be shooter (labelled “Ethical Framework”) to save an employee (labelled “Freedom 0”)
mutual aid but for your agents :^)
Proudhon would never
This is the exact opposite of mutual aid
Just remembered I had this piece of rationalist word salad saved for later

the context is that dude was arguing that anyone who understands the ‘potential of the tech’ knows that it is an existential risk and that you only need a layman’s understanding to know that. (he specifically said that ‘a sufficiently intelligent five year old’ could figure it out) and when I argued against him this is what I got in reply
i don’t understand, are there trees or are there not? should we plant trees? is it futile? should we dig a ditch?
also am i a tree?
Quite clearly they have understood that planting trees (as in planting trees) is in direct opposition to the rolling boulder (as in slopbots). Therefore we should stop planting trees so that we may be more effectively crushed by the boulder (slopbot).
A story of two halves
Yesterday

Today

Nvidia Employees Are Pushing for a Sam Altman 2.0
Before Nvidia’s board of directors announced Thursday morning that Huang would be stepping aside, to be replaced by longtime company CFO Colette Kress, a veteran engineer at Nvidia submitted an internal post urging the board to take employee sentiment into account before making a final decision, Wired reported Thursday evening. It has since turned into a kind of petition for the disgraced ex-CEO’s return, garnering more than 1,600 comments, many of which are reportedly supportive of Huang’s reinstatement as CEO.
Not at Sam Altman return levels of backlash yet
why don’t they make AI Jensen in the digital factory so it’s like he never left and then they can kiss him each night and tuck him into bed.
Seriously, what the fuck is melting people’s brains out there. Is everybody in the bay area poisoned by ergot?
the commerce secretary is also testifying about Nvidia’s chip exports before the banking committee
Commerce Secretary Howard Lutnick has agreed to testify before the Senate banking committee on December 10 over questions involving Nvidia’s chip exports to China.
Outlook was down for an entire day yesterday, which affected me since my school email uses it. I don’t remember any Outlook outage this bad before, and from what I’ve heard, Microsoft was cosplaying as a headless chicken in their attempts to fix it. Maybe getting rid of 60,000 employees to embrace the magic of AI isn’t working out so well.
Meanwhile I saw this AI panic post on Linkedin (warning: psychic damage). Apparently all the doomers are vindicated and all the critics are completely discredited because of … you guessed it, math problems! You see, one mathematician was very sad in an interview! (Let’s also ignore the many nuanced statements like this that mathematicians have made about their own field.)
But somehow he hasn’t noticed AI affecting anything he actually understands. Well, you see, that’s because in the real world, the magic of AI does need to wait a few years before the puny humans understand how to wield its power (but of course the doomsday is coming on schedule in a few years). The argument sounds nice until you look at of software, where there should be no physical obstacles for adopting AI. Now, coding has been solved by AI on the planet of Earth Prime where he lives, where OpenAI and Anthropic have long monopolized every software market with products of much higher quality and reliability than what came before. Unfortunately, I live on Earth, where I have seen two major vibe-coding failures in the last two days (the Outlook outage and the Humble Store rewrite).
Won’tcha look at that. Now that MAGA isn’t hot shit anymore, Alex Karp goes full “How do you do, fellow Democrats” and says Palantir has basically always been a progressive company.
https://www.businessinsider.com/alex-karp-palantir-democrat-progressive-midterms-2026-11

In the style of “claims to be pro-life, dies anyway”: claims to be a Democrat, dismantles democracy anyway
Alex Karp will say anything that will let him take coke and murder people.
https://www.lesswrong.com/posts/7xJeaAKTKxbHL6F9a/a-new-kind-of-prediction-market-for-science
Got a problem of randos shitting AI-generated ARXIV preprints into existence faster than the scientific community can deal with it? SHOEHORN IN PREDICTION MARKETS!
(Incidentally, my boss has been complaining about the onslaught of low quality preprints that are shat out with the usual low-quality signifiers that let you filter things being harder to immediately see, and his conclusion is that he is sad to say science is going back towards and has to go back towards systems of reputation and patronage where you have to look at who lent their reputation to a work and who newcomers are being vouched for by to determine if it is worth looking at it at all.)
(And I am again reminded of our electronic information sphere becoming like the Earth’s genetic information sphere in that it is full of selfish parasites looking to hijack you to use their nonproductive information, with productive cellular life jealously guarding its genomic nucleic acids and not letting it interact with the froth of selfish viral and mobile elements bouncing off it at all times.)
Math 3.0: exploiting Lean bugs to scam bettors on Kalshi












