2 ms·
I don't really agree with the other commenter, I don't think EAs are intentionally sacrificing humanity for AI, but the fact it's unintentional doesn't really c
by postboxaerobics 2mo ago
I don't really agree with the other commenter, I don't think EAs are intentionally sacrificing humanity for AI, but the fact it's unintentional doesn't really change the fact this is what they are doing.
Anthropic are explicitly building ASI, using recursive self improvement, with no knowledge of how to control it. They spend only a minimal fraction of their funds on safety work, none of which relates to controlling ASI.
I am open to being wrong about Anthropic, but it will be hard to convince me that a company developing ASI is prioritising safety. Do you really believe what they are doing is safe?
- urams 2mo agoI truly don't see any major problem with what they're doing. For any current significant risk (security, bio risks), their models are notoriously useless.
- pixl97 2mo agoThis is an odd take. OpenAI models you use are pretty useless for hacking. The models they have seem to be pretty damned good at it. It's kind of like saying you can't kill everyone with weapons when you're looking at what you can buy at the local Walmart and not what's in the 1000 foot deep missile silo. Also, it's kind of a once you build it, it's too late kind of issue. No model will be good enough, the problem is still in the future, we're all dead.