Comment by RivieraKid
4 hours ago
I'm not that young anymore but when I hear Dario Amodei speak and gesticulate enthusiastically about how AI will soon make my job redundant, from a positio of someone who can retire now and who benefits from AI possibly replacing workers, it's very hard to like him.
Lets suppose everyone in the world miraculously loses their jobs to AI, who is paying for it? How? What are they selling? What stops the rest of humanity from turning data centers into a pile of ash?
These people are so disconnected from reality.
And, idk what is the end goal here. Even if 20% of people lose their job the crime and prostitution will shoot up....
More free time for proper rest, hobby, and well-considered expression?
I’m not saying Anthropic and OpenAI are going to be the ones to deliver it - but…you really don’t see the ubiquitous need for these?
1 reply →
You also just explained how people fearing that AI will take all jobs are equally disconnected from reality. Although in less joyful manner.
what kind of bargaining power will you have with "well someone must consume!"?
Talk about shooting the messenger…
A CEO of a large AI company is hardly just a "messenger" here. Come on now.
He is just the poor little head of the snake.
I… yes, he is. He wasn’t born when this was put in motion by Turing in 1950, and he quit OpenAI to try to stick to their original mission, which was exclusively and ambitiously to prevent what most (who cared to look at it) assumed was inevitable.
AI isn’t a new iPhone moment, it’s a fire moment. Dario Amodei didn’t bring it into existence, physics did — we’re just slowly working out the cognitive science for the first time after Crysis machines ended up being really good at something that our type of intelligence relies on (intuitive framing via huge neural networks — see The Frame Problem).
Also this doesn’t really matter in anthropic’s case, but FWIW: we should try to blame CEOs a bit less, and shareholders (capitalists) a lot more. Our system isn’t named ‘officerism’, after all!
The CEO of a major AI company is not “the messenger”
He's the guy doing the thing that he is delivering the message about...
Man, I see this stuff, and I think, "my god! think of the opportunity we have to replace the current economic system with something where we're not all in a fierce knife fight to try to stay employed so we don't starve!"
We have to stop fearing the machines and start fearing the firmware running on our brains! It's capitalism that's the problem, not the robots... if anything the robots can liberate us from capital's death grip!
If Dario was up there preaching about replacing today with Star Trek I think people wouldn’t hate it. Everything those at the head of this project say amounts to the current state except abject poverty for everyone but them. When Dario donates his share in Anthropic to fight hunger I’ll consider being optimistic that capital is going anywhere.
>replace the current economic system with something where we're not all in a fierce knife fight to try to stay employed
Jobs will become even more scarce, so how does that follow?
>robots can liberate us from capital's death grip
We are already live in a world sufficiently technologically advanced to feed and house everyone, and yet we don't.
It’s amusing to imagine the oligarchs are going to anymore charitable than currently.
They don't have to be charitable.
We can literally take it from them. I get why people don't like to think about that sort of thing, but... yeah. We could make AI a sort of public good and do something like the Permanent Fund Dividend or something. There's lots of options even in the "reformist" way of thinking about it. But we shouldn't be naive.
It's not their choice, no matter how much they'd like it to be.
The Chinese models are able to distill outputs into a model of slightly worse performance within a few weeks of each release, and the same copyright exemptions which protect these guys scraping also protect China.
As long as their lobbyists don't win (and the deck is stacked against them, make no mistake) the future of AI is not going to be owned by the frontier labs.