> reasonable evidence that their influence isn't prevalent because they're right, and that they would have ulterior motives for continuing to preach that the world is coming to an end even if evidence showed otherwise.
They make a claim about a specific community in San Francisco, that is tied together by a mix of AI doomerism and sexual relations. And it's not at all surprising or some farfetched claim that people who enjoy having sex with someone might adopt their ideas, at least publicly, for fear of losing that sexual connection.
am i naive? in the direction you pose this, i can't imagine it's that common and if that happens i don't feel sorry for our imagined victim. being a bay area guy who attends so many orgies that AI doomerism rubs off on him? and he keeps spreading it to other people and that some substantial bulk of doomerism is due to this mechanism such that it voids the ideological purity of doomerism? i'm not a bay area guy and i haven't attended these orgies but i hope they have other shit to talk about at the orgies than AI
if the relationship is flipped - being genuinely interested in the theory but then being roped into weird sex shit that you didn't ask for, sure that that sounds predatory. a sex cult where AI is instrumental, rather than an AI cult where sex is instrumental. it would be more remarkable imo but i don't know if anyone is saying out loud that's what's happening.
The claim is that these people, in spite of their claims of simply rationally following the evidence wherever it leads, have become trapped in the idea that AI doom is imminent and the only serious concern that anyone in the AI community (at least!) should have. This irrational attachment is then re-enforced not through rational discourse, but through the strong emotional bonds and biases induced by a network of sexual relations.
And it also claim that this phenomenon is happening in a specific group in Berkley, California, and that a large majority of people working in AI safety at the biggest labs are members of this group.
I think this article kind of misses a crucial point. They weren't always AI doomers, that's a thing they picked up along the way. Their core beliefs are 1. That they are the smartest, and 2. That this makes them morally superior because consequentialism (This sort of follows. If you believe in consequentialism, leaving all your decisions to the superior consequence predictor is surely the moral thing to do)
Maybe one day they will stop believing in AI doom, but they will never, ever stop believing they belong to the intellectually and thus morally elect. The day any of them do that they have either broken with the cult or died.
Maybe that’s why they’re so pissed, ChatGPT’s useless verbosity is threatening to put them out of a job.
yeah, it's no coincidence that AIs sound a bit like them -- a lot of the people doing the instruction tuning are affiliated with these people.
And it's also a profound self-identity threat to them if that kind of intelligence is nothing special.