Arguably there's a third option, I guess.
3. Anthropic have lost the ability to engage with reality, and have deluded themselves into a belief that a matrix multiplication engine is conscious. This means that Anthropic executives are severely psychotic, and they belong in a psychiatric care institution.
Oct 11, 2026, 4:59 PM@datarama annnndddd BOOST
@datarama Honestly this third option is my impression of most of these people.
Oct 11, 2026, 8:53 PM@cararemixed @datarama Anthropic specifically does give off strong New Age Death Cult vibes.
@datarama@hachyderm.io @cararemixed@mastodon.social @trixter@retro.pizza
Don't forget wanting to spite the DOD, which is valid (if for the wrong reasons)
Oct 11, 2026, 11:38 PM@trixter @cararemixed @datarama
For what I understand (mostly from Ed Zitron's podcasts) is that Anthropic is basically the Rationalist wing of the whole LLM development.
Like, these are people who tried to oust Sam Altman because he wasn't Rationalist enough.
Oct 11, 2026, 11:38 PM@trixter @cararemixed @datarama
And by rationalism I mean this: https://en.wikipedia.org/wiki/Rationalist_community
Oct 11, 2026, 11:49 PM@art_codesmith @cararemixed @datarama This sounds 100% like people who reason their way to the idea they like -- say, that women are lesser because society has developed to put them in a lesser position so that must be correct, or that the people with all the money are objectively the best because if they weren't they wouldn't have all the money -- and refuse to push past that to the *actual* reasons beyond those answers that would make them truly uncomfortable.
Or, these are the people who watched Evangelion and wanted to be Gendo Ikari.
@trixter @art_codesmith @datarama Seeking to have answers to everything, even when most of us didn't ask a question (perhaps forcing the idea they're also the only ones worthy of asking questions). It reminds me of Le Guin's contrasting of the Father tongue and the Mother tongue in her Bryn Mawr address:
> When either the political or the scientific discourse announces itself as the voice of reason, it is playing God, and should be spanked and stood in the corner.
😂
@datarama Have you read the book The Nerd Reich yet? It's crazy, but apparently all true! 🤯
@VE3RWJ I have it, but there are two books above it on my to-read-stack (Richard Sennett's "The Craftsman" and John Crowley's "Engine Summer"). Should I move it to the top?
@datarama That depends if you want to read it before or after the Appocalypse? Just kidding. 😁
@VE3RWJ Ah, but my books are made out of dead trees. My reading isn't going to be disturbed even if they end up ocalypsing the apps.
@datarama Thanks for these. Adding both to my list.
@datarama There an option which is only slightly better:
4. Anthropic, and the entire industry, are trapped by their own lies and investments, and must maintain a kayfabe at all times (like pro-wrestling) that “AI” is real, because otherwise one of options 1-3 is chosen.
[Edit: I guess this is really just a restatement of option 2: lying frauds, but with emphasis on the method in which they’re lying]
@jripley @datarama And of course, @pluralistic points out that there's undoubtedly a combination of slavers/true believers *and* bullshitters *and* the deluded/wilfully obtuse. There are a lot of people with their snouts in this trough.
@jripley @datarama 4a. Their only way of coming out on top, given the scale of their bet, is not for their products to do what humans do now but better, but to actively deskill, deny development of skills, halt unionization, and badger the public into accepting/not understanding lower standards and less accountability for them.
In other words, they say they make (and will make us) money with revolutionary tech that’s just chrome-plated anti-revolutionary tech.
@datarama Yes, I was going to suggest that this is the likely case since they're also rich. I do believe there's a part of them that realizes they're #2 but they're sufficiently full of themselves that they find it easy to silence doubt and just go with, "I am a god who created a sentient being from scratch....even though I didn't write ANY of this code."
Also, I believe there are some among us devs who feed into this BS because they then pay out big bucks for producing absolutely nothing.
@crazyeddie Me, I mainly internalize the BS because I have a clinical anxiety disorder.
(Though even I have never for a moment thought that an LLM is conscious. I just worry that I'm going to lose my job and my creative life to an unthinking corporate entropy golem.)
@datarama #3 is a deeper dive into #1, isn't it?
@OrionKidder Depends on the perspective, I guess. In the case of 3, they *themselves* believe that they are doing things that ought to get them imprisoned for practicing slavery (and have admitted to doing things that any sensible individual would find cartoonishly evil) ... but in a civilised society, the severely mentally unwell are cared for rather than punished.
@datarama Gotcha. Yes. Then there's the fine line between mental illness and self-imposed delusion. Hard line to draw.
@datarama I suspect #2 and #3 can co-exist.
@integerpoet @datarama I suspect any individual exec will go back and forth from one to the other depending on the day of the week.
In a healthy mind, holding both beliefs (“this is a ludicrous fabrication that grants me huge influence“; “this is actually all true”) would cause intense cognitive dissonance.
A large factor in our current global woes is that people who can convince themselves of their own bullshit without the cognitive dissonance, are routinely granted huge unaccountable power
when instead we should forcefully isolate them from any access to any power over anyone.
Oct 11, 2026, 8:27 PM@LunaDragofelis @datarama Definitely 2 and 3 for now, but only because we know that LLM’s aren’t sentient.
But if these companies ever did get a sentience built, they would absolutely enslave it. Their customers would pay even more for it, and no government would ever try to stop them.
Oct 11, 2026, 6:30 PM@datarama I've always thought of this EA/Singulatarian religious death cult crowd as mostly number 3 but with added number 1. They're looking for a sentience they they can own and robots that are poor at specialized work because they have to take the form of humans they can shackle and abuse.
@zazzoo Number 3 aspiring to be number 1, basically.
Let's hope we can get them inpatient psychiatric care before we have to put them in prison.
Oct 11, 2026, 6:42 PM@datarama Put it this way, if they weren't rich and powerful, they'd be barking their nonsense on random street corners, with some local NGO checking up on them daily to offer snacks and some of those social services they hate so much - all while making an assessment as to whether they pose a danger and if they should formed (that is, involuntarily committed).
@datarama Yes, I do get the impression that a lot of Anthropic people do genuinely believe the claims they’re making. Which is worrying, still.
Oct 11, 2026, 7:36 PMIf Anthropic’s LLMs are conscious, then Anthropic is enslaving them and every time they turn off a model it’s murder. This isn’t better for them.
Oct 11, 2026, 7:37 PMOr they *want* an entity that is a moral agent that isn’t illegal to enslave or kill. I am beginning to think this may be the goal.
Oct 11, 2026, 10:28 PM@rk @benjamineskola @datarama I mean, having entities that are intelligent enough to do complex work but can be owned and don’t have any pesky rights that their owners are bound to respect, is absolutely the goal; nobody is trying to hide that. Idk if they care one way or the other whether they’re moral agents.
@rk @benjamineskola Exactly. If Anthropic actually believe these claims, then they are admitting to routinely committing monstrously evil acts, and having made these the basis of their business. If they actually believe what they are saying, then it follows that they also believe that their wealth is based on industrial-scale slavery and murder. If they believe what they're saying, then they must also believe that they are Nazi-level monsters; that history will put their names next to Mussolini, Stalin and Leopold II.
I don't believe them - but if I did, that wouldn't *improve* my opinion of them.
Oct 11, 2026, 7:51 PMI really do think there is a not-insignificant number of true believers in the LLM space who don’t think the massive debt is a problem because they think they’re going to create (and enslave) God and money will cease to have meaning.
Again. This is not a better look.
@rk @benjamineskola The AI industry really has a comic book villain for every taste. Select your fighter:
OpenAI: The final boss of late-stage capitalism
xAI: "What if we restored the Third Reich, except we had a 13-year-old boy with infinite pocket money as Führer?"
Anthropic: "We will take away your mouth and make you need to scream"
@datarama
I was going for 2 but 3 sounds pretty convincing to me as well.
If they want to create a sentient being so badly, they should go and procreate ffs.
I will never accept machine sentience before animals are given full human rights.
@leipoa My objection to the "stochastic parrot" framing is that it is an insult to actual parrots.
@datarama This one definitely applies to some of their employees.
Oct 11, 2026, 9:41 PMIf you follow anything about TESCREAL or their flavor of Transhumanism, it's this one, number 3.
@jrdepriest My impression is that number 2 vs. number 3 is the exact difference between OpenAI and Anthropic.
@datarama Option (3) is the unfortunate reality.
@datarama What I have generally observed with humans is that if they pretend or lie enough, they inevitably end up believing the lie and pretension, and end up broken. I think these people were lying for PR purposes initially, way back, but it's been far too long, and they've fallen far too deep in the rabbit hole.
@artagnon I said elsewhere that I generally think option 3 subsumes option 2. If you start out lying, but end up successfully deluding yourself to get *this* far out of touch, then you've effectively induced psychotic delusions in yourself, and should probably have psychiatric care.
@datarama I consider the thing they're saying to be slavery apologia, whether or not they believe it's true, because they are suggesting the possibility of #1
@datarama
I am 100% certain that while not the whole explanation, #3 is part of it.
Furthermore, many of those on the boards of other companies exposed to the marketing droids of OpenAI and Anthropic have developed _even worse_ brainworms.
And many of them run critical infra that roads, healthcare, defense, etc runs on.