Post

Why talk about how nuts the AI rationalist sex cult is, and not just focus on the (very real) risks and harms of AI? Well, when we're discussing "who should get to regulate AI, and how," I think it's VERY relevant to talk about the antisocial, unsettling ways many people running US AI think.

If we want to actually get a chance to regulate AI sensibly, we have to get these crazy millenarian cultist weirdos out of the powerful positions too many currently occupy. To accomplish that, we absolutely need more normal members of the public to realize that they *are* crazy weirdos.

It becomes a LOT easier to argue to the general public that we shouldn't let the AI industry regulate itself when you can ALSO produce clear evidence that much of the AI industry appears to be steered by genuinely crazy weirdos with anti-social and anti-human ideas.

I am intensely in favor of regulating AI (and I have been advocating for and reporting on ethical tech and tech regulations for a long time). But I'm afraid we won't get a chance to do that if the wacko millenarian cultist faction is able to control the regulatory process (as they're trying to do).

Of course, we can't exactly expect too much in the way of actually-good AI regulation under the disastrous Trump regime. But I do think we can fight to hold the line and demand sensible regulations that aren't steered by big tech and big money interests until 2028.

The timing is good for this, because the public appears to be a lot angrier about mass surveillance, the unethical and pervasive use of AI, and general Silicon Valley overreach than they've ever been before. People are a lot more interested in hearing critiques of big tech than I can ever remember.

So no, I just don't think that talking about how weird these people are is some sort of galaxy brain propaganda distraction that actually benefits the same set of unethical big AI pushers. I think we do in fact need to talk about the weirdness if we want to get a chance to regulate AI.

Also, what @queengwenevere.bsky.social said. Emphasizing how creepy and weird a lot of big AI has become is also a key way to convince politicians not to play footsie with them -

This also matters because too many big-name, well-funded "AI safety" proponents are actually linked to the Weird Millenarian AI Sex/Death Cult, and thus focus on galaxy-brain hypothetical AI risks, instead of actually-relevant stuff like "job displacement" and "data center environmental impact."

Imagine if, instead of regulating drones (which we did!), we instead decided to let drone companies self-regulate, and ALSO dedicated tons of money and time to "problems" like "what if, in 1000 years, hyper-intelligent dogs learn to fly drones into mailmen." This is what we are doing with AI.

also imagine if there was a weird cult that wanted to worship the drones, or at least believed they were somewhat sentient little guys like you and me, and we not only let those people run most of the drone companies, we also let many of those people self-regulate the drone industry. weird, right?

I think it was easier to regulate drones because they're physical objects, which means they're less abstract-feeling and more obviously dangerous than AI is, at least to many. But people seem to be rapidly realizing digital-first technologies can be just as dangerous as more tangible technologies.

I don't think you have quite got it re your point 2. The timescale for fully autonomous AI is nearer 10 years than 1000. Consider everything that is now 'in the cloud' or is connected to the internet in some way - your bank, the police and security services, ...🧵