Anthropic was $42B in the hole for 2025, reuters reports. Apparently this puts their $-spent-to-$-earned ratio slightly over the one calculated by Zitron for OpenAI, making Anthropic in fact the shittier business.

For comparison, the GDP of Cyprus for 2025 was $41B.

I lean towards things being unfathomably worse that they appear a the surface, there’s just so much smoke, except I think the problem is structural and not the result of deliberate machinations by evil genius pedophiles.

The nazi bar analogy works for more than ideology, and the burgeoning CNC scene they got going seems to be exacerbating the if-you-build-it-they-will-come aspect of the whole thing, because good luck trying to explain to authorities how you got real raped at the fake rape party you are supposed to have willingly attended.

This is all third hand at best, but I live on the other side of the world so whatever. As far as personal experiences go @ivyastrix just posted a new blog about her victimization at rationalist events where she describes some of the systemic pathologies in rationalist organisations trying to deal with SA allegations, and it’s basically the same principles that make them believe that vertical workplace romance is not a minefield of power imbalance and coercion but actually fine and ethical when the right people are doing it, just fundamentally unserious people biting off more than they can chew and acting defensive and victim blaming and trying to pass it off as drama when it blows up in their faces.

It goes off the speculative rails for most of the second half, but I have to admit the rate at which Αela compliments people on her socials on being good rapists could be the most off-putting thing I’ve ever encountered in relation to rationalism.

From the more out there claims I thought the part where he advocates that Irregular (the people with rat/EA ties who made the swiss cheese LLM sandboxes) where employing weaponized incompetence in order to let a big incident happen that would force the government to go through with pausing AI research was interesting. though obviously of dubious likelihood. The whole doomer rationalists sabotaging slightly less doomer EAs in a low key sort of civil war over pausing vs slowing down AI research wasn’t very convincing in general.

There’s also a bit where he thinks the EAs have infiltrated all levels of government and are planning to eliminate the competition (non rationalist controlled AI) via a ‘pivotal act’ that involves bioweapons. Basically doing the rationalist rhetorical staple of smuggling an extreme position as a given and then just piling on the speculation.

Dedicating a paragraph and a half to the fact that the usually unspoken second part of the pause AI doomer project is to just do eugenics for a few generations until we get an upper class of genetically engineered naturally galaxy brained people that can finally do AI properly isn’t nothing, but at some point you need to segue into how siskind rolled out a red carpet for every semi-literate blood and soil gutter fascist and spurned nice guy to make a home in the ratsphere, you wrote the scott alexander email explainer ffs.

There’s also a weirdly grandstanding section where he writes like he’ll get the rationalists like FBI got Capone for tax-evasion, except it’s Lighthaven committing massive fraud against its investors and donors by knowingly funding events that lead to solicited sex acts in the compound’s premises, which is by far the least objectionable sex related issue in a writeup that’s leading up to the claim that the rationalism in incidental and the real goal is funneling vulnerable minors to sfba based pedophiles.

What the actual fuck? Imagine if Hermann Göring defended himself at Nuremberg by saying, “At least I had fun and got rich and worked on ‘cool tech’ with friends.”

Once the rockets are up, who cares where they come down, that’s not my department says Wernher von Braun.

Ooh cryonics, classy.

Wish how the american AI industry is thoroughly permeated by Yud’s views on AI, and how Yud knows fuck all about how AI or computers work beyond as a sci-fi plot device, would also catch on.

Turns out the centrist subscribes to transhumanism (hipster edition)

the overtone window can’t be that fucked, can it?

Had to google Kurzweilian h+, apparently h+ is kurzweil’s brand of transhumanism.

This is very anticlimactic.

if the doomers were really trying to win, they should be looking at the party that at least has a chance of agreeing to more regulations.

It’s possible for a gaggle of self important dollar store machiavellis to land on Trump being the surer bet because he openly takes bribes and they have money, forgetting that when he takes the money their leverage ends because they are not the mafia.

People have been calling them a cult incubator for ages, not sure who said it first.

Entryism in the sneersphere

Absolutely stealing this.

51-post long thread on why rats are a cult,

Jjust finished reading it.

woke:

Everyone is polyamorous, everyone attends the same events, everyone lives in group homes together, everyone venerates the same few authors, and everyone is very worried about AI doom. They are all very contrarian together.

Rationalism as a whole is a religious movement. Not everyone takes marching orders. The core people who mostly live together and attend events and especially orgies in Berkeley are a sex cult, and the ones that aren’t having any sex are part of the wider movement and are its recruiting funnel.

Any time you get more than a few people who are deeply concerned with AI doom together, it seems like their natural form of organization looks like a cult; rationalism as a whole is less of a cult than it is a cult factory. Some of these cults have gone very badly. One of them is the Zizians.

broke:

I feel about [CEOs talking about AI progress slowdown] roughly the same way I would feel about all the pharmaceutical companies getting together to say they wanted to waive anti-trust laws so they could form a cartel to ban medical research:

these people are the enemies of humanity.

Also: Much of AI research IS medical research. They are very literally proposing to stop a pretty good chunk of medical research in its tracks so that they don’t have to fear the paperclipper.

It’s possible he has a similar affliction to Doctorow, where he thinks this tech is neutral on its own and can be sufficiently decoupled from its hypercapitalist pushers who would much prefer democracy be swapped for modernised feudalism, or who at least think it’s fine to collapse global systems of knowledge transfer for a quick buck.

Still, attributing medical research to AI without discriminating between LLMs and vintage orthodox machine learning that doesn’t actually need infinity GPUs inside a climate regulated warehouse the size of the Peloponnese that also needs to be powered by two active volcanos to work is very TESCREAL coded, LLMs doing Great Things Worth Ruining The World Over in medical research only works under the assumption of some singularity type event being just around the corner. And also the part where he thinks it’s normal to declare people who fuck with AI as enemies of humanity.

Basically segyges feels like controlled opposition, just not by EAs, perhaps other people who don’t mind that he never brings up the pervasive eugenics angle? He certainly posts (including in LessWrong) like it’s his day job.

Or maybe he just really wants the basilisk to like him.

Nice to see he’s been keeping productive now that he apparently got banned from bsky.

Just in time for TESCREAL (“the weirdo sf harry potter sex cult that runs AI”) to go viral too.

Yud isn’t necessarily inconsistent here: he really really wants the robot god to become real so it can invent the blowjob machine and win utilitarianism, but he only trusts people he’s met personally to do it, and certainly not those shifty celestials.

Please don’t think about what you as an individual should do about the problems facing the world; you’re just an NPC after all.

Or as the AI2027 blog put it:

We may only have 2 years left before humanity’s fate is sealed!

Despite the urgency, please do not pursue extreme uncooperative actions. If something seems very bad on common-sense ethical views, don’t do it.

Burn it down and salt the earth then. What’s the point of whinging about it?

D) Could be they finally read the room and realized we’re not that many news cycles before experts warn AI might destroy your livelihood and also incarnate as a dark god of genocide turns to burn it.

Anthropic Is Building a Predictive Surveillance System to Monitor Activists

Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitoring system to keep tabs on activists who oppose the rapid development of artificial intelligence.

In addition to monitoring activists in the vicinity of Anthropic executives and keeping tabs on protests near physical Anthropic assets, the firm is also implementing a “pre-crime” approach, attempting to predict incidents before they happen. In some cases, that also means reporting suspects to police before a crime occurs. Anthropic did not respond to the Prospect’s request for comment.

Apparently they can also sick the police on you on a trust-me-bro basis:

Last month, The San Francisco Standard reported that Anthropic had reported a man to San Francisco police for telling Claude that he had bought an AR-15 semiautomatic rifle and had CEO Dario Amodei “in his sights.” When the Standard contacted the man in question, he told the newspaper he was “just fucking around.”

But while Anthropic was fast to call the cops on a frustrated Claude user, the Standard also reported a key detail: Anthropic refused to show police the actual messages, citing Anthropic’s internal policy. In short, Anthropic reported a user for in-platform speech, and then refused to provide police with evidence of actual wrongdoing.

image shamelessly stolen from https://bsky.app/profile/timhenke.bsky.social/post/3mv5puo47ns2z

I’m kind of surprised that this made the news in Greece, unlike the other 128 times something very similar happened. I wonder if anthropic marketing is pushing the story harder than usual, or if it’s the opposite and they’re asleep on the wheel or simply don’t think there’s anything to gain from trying to contain such stories anymore.

Running a game for any reasonable amount of time this way will bankrupt you may times over from token overspend. This is just to make investors horny, it’s not getting released in any meaningful way.

wtf is deal of that segyges person

He does worthless little-birthday-boy AI boosting while pretending to be the serious face of bluesky AI bootlicking fair and balanced anti-anti-AI bluesky.

Occasionally has decent sneer content, though it has a much more evident muh datacenters tint these days. Writer of the siskind email explainer.