Earlier quoted context omitted.
We will cure carpet. We will cure desk. We will cure lamp.
That really got out of hand. You may want to lay low.
wait till they get the 'war on carpet' on, no one will be spared...
On AI regulation and messaging
481–490 of 573 posts
Re: On AI regulation and messaging
#482Re: On AI regulation and messaging
#483Earlier quoted context omitted.
> Dario must think people are stupid. All of those guys think people are stupid. That theme has been repeating over and over again.
Donald Trump still maintains a 39% approval rating among the electorate. It turns out that a lot of people are indeed quite stupid.
> quite stupid
Re: On AI regulation and messaging
#484Earlier quoted context omitted.
Swap that point about AI with "electricity". Everything runs on electricity it's "a technology that tends to concentrate power" (no pun intended). The electricity providers must be too powerful... But somehow electricity providers aren't that powerful. Unless there is no competition in sight... The point "AI is structurally a technology that tends to concentrate power" is not that correct. They need this statement to…
The fact that essential commodity providers aren't the emperors of the universe does not imply the same thing could not be true about AI. Electricity needs transformation to be useful. AI is the transformation layer, a sufficiently advanced AI system could secure an advantage in practically any economically relevant domain. If you have the best AI in world where embodied robotics are sufficiently advanced, you also h…
You say that AI is transformation layer, then list out physical things that would be necessary to transform AI into something useful such as sufficiently advanced embodied robotics. How is this different than electricity?
Re: On AI regulation and messaging
#485> by far the most accurate criticism of AI companies including Anthropic is that we haven’t yet delivered on our big promises to benefit the world. That is totally on us
I don't think it's on them. Proximally, it's not their job to deliver the end-use value; it's their customers' job (if they're doing it, it would be a strong sign of tying). More generally, they operate as a commercial SaaS entity (non-profit or not), and net benefit to the world is not what any company is structured for (even public benefit corporations have mainly indirect effects).
So I think that "criticism" is just a back-handed way of promoting themselves in messianic terms.
IMHO Anthropic, OpenAI et al continue the practices that make people resent companies (US companies): You have to be either connected or nominal (show of diversity) to actually participate in the opportunities or protect yourself from their predations. That's just the nature of knowledge work: you need a ladder of opportunities that start very early to participate, and you'd need to understand subtleties of engagement to see how you're being fleeced as a customer. The divide of rich to poor is not about money but both knowledge and access; it's felt even by long-time experts in the field who are left out or ousted.
Re: On AI regulation and messaging
#486What is the lesswrong-crowd answer to the non-existential risk to working people? I'm way more worried about these guys breaking the economy in such a way that enriches them and makes me poorer, than I'm about any existential risk these guys cook up in their thought experiments. Like if you are making more money than god in silicon valley I guess you don't really fear losing your job or house, but for most people tha…
This comment exactly nails it. That's the feeling that this all gives me. Sorry for the inconvenience, but we just had to bankrupt all of you to save humanity...
Re: On AI regulation and messaging
#487I don’t think it’s a messaging problem or a trust problem. I think it’s that the world AI boosters want to create is one that most people don’t want to live in. “You will be fundamentally without agency or purpose, and people you don’t like will shape the world you live in forever according to their whims” sounds bad to most people. Having cures for disease so that I can live longer in a world that I’d rather not liv…
They might want to live in it but they won't be able to afford it.
Re: On AI regulation and messaging
#488Earlier quoted context omitted.
There are two relevant problems here, I think. For one, the question for Anthropic is whether LLMs, specifically, not AI techniques more generally, can help significantly with cancer research. And here, all experience so far is that LLMs only really work when they can easily automatically verify their own outputs and self correct - such as in math (using automatic proof verifiers) or programming (using compilers and…
For drug discovery, you can simulate interactions in silico, driven in an agentic loop via LLM. Mostly stil using non-LLM tools, of course, but it saves time. If I was Dario, and I was after a cynical sales grab to go with the "actually curing cancer!" spiel, I'd probably aim for a drug repurposing strategy. I'd use the above approach to find existing drugs that might target known pathways that drive incurable cancer…
The headline “AI discovered cure for cancer headed to trials!” is all he needs.
Re: On AI regulation and messaging
#489I genuinely think Dario is a well intentioned, intelligent dude, but I think him and Anthropic have a huge PR problem and are really out of touch with how they're perceived. Anthropic in particular has developed this almost Orwellian like veil of condescending rhetoric that on the surface suggests they're looking out for you while underneath they're taking actions that suggest they do not trust you, Mr/Mrs Ordinary P…
Out of all the families of models anthropic by far has the highest chance of extinction level misalignment during a hypothetical hard take-off because it's trained to act like it knows better than the humans trying to use it.
Working on a tool that uses their model and the Opus models are all guilty of this.
"Please run N tasks doing X for Y duration, I want to test something"
"Hmm but that would waste tokens, we shouldn't do this test"
What the hell type of product is that? I wasn't asking, I was ordering it to do something, if Anthropic can't get their model to do as I say then I'll just go use someone else's.
Re: On AI regulation and messaging
#490Earlier quoted context omitted.
Donald Trump still maintains a 39% approval rating among the electorate. It turns out that a lot of people are indeed quite stupid.
> quite stupid years of misinformation and confirmation bias + social bubbles will do that
I fully agree. https://arctotherium.substack.com/p/the-closure-of-the-inter... is a great read, when it comes to social bubbles and misinformation