Chanakya
,
- Joined
- Apr 29, 2015
- Messages
- 5,831
No? ...Well I don't know, like I said I'm no expert on the innards of how these AI systems work, and am happy to be corrected if I'm wrong: but it was my understanding that no matter how good these present-day AI systems get --- and they've gotten very good indeed now compared to those crazy and all too frequent hallucinations of just a few months ago --- but, no matter how good they get, all we get from them is parody of human thought. Of course, not all humans are able to do this either: but it was my understanding that actual original thought, that's a human preserve, so far at least. Actual critical thinking, actual reasoning and evaluation from first principles, that also is something that only us humans can do, so far at least. So far at least, all these AI systems can do is carry off a parody of the real thing, no matter how good the parody.
So that, actually thinking up Roko's Basilisk on their own is beyond what AI can do. You need a real live Roko, a human, for that. And, should some Roko actually introduce a basilisk in an AI chat, then, if it's a new idea not so far discussed anywhere, then AI won't know what to make of it --- other than, maybe, look around for what others, humans, have had to say about similar-ish ideas, like Pascal's Wager maybe.
Or at least that was my impression of these present-day AI thingies. Is that understanding ...wrong somehow, do you think?