I have a very rough blog draft on when it is and isn't useful to dispute terms and why we do it: e.g. whether AI is intelligent, what AGI is, whether machines know/understand/learn. But it might just turn into a weird love letter to Carnap.
Conversation
Replying to
Are you as annoyed as I am about the 'AGI doesn't make sense as a concept' people
1
5
Replying to
I haven't encountered this much, to be honest. Is a charitable reading of it "AGI is a bad/unhelpful concept"?
2
1
Replying to
One example: youtube.com/watch?v=kWsHS7
E.g. that intelligence is context-dependent and free lunch theorem means one can't have a truly general agent etc. Also the Chollet position (at least as of a few years ago)
