See the problem of Friendly AI; that is, if humans are going to make a powerful AI, we should make sure it doesn’t do something to wreck our shit, like turn the whole universe into paperclips or some other crazy thing—i.e. it should be Friendly.
RichardKennaway was putting a jokey spin on the idea by suggesting that we solve the problem of designing Friendly Human Intelligence, by analogy to the problem of designing Friendly Artificial Intelligence. (Edited last sentence for accuracy.)
Exactly. Well, not instead of FAI, but FHI is an important problem, as old as humanity: how to bring up your kids right and stop them wrecking the place.
See the problem of Friendly AI; that is, if humans are going to make a powerful AI, we should make sure it doesn’t do something to wreck our shit, like turn the whole universe into paperclips or some other crazy thing—i.e. it should be Friendly.
RichardKennaway was putting a jokey spin on the idea by suggesting that we solve the problem of designing Friendly Human Intelligence, by analogy to the problem of designing Friendly Artificial Intelligence. (Edited last sentence for accuracy.)
Exactly. Well, not instead of FAI, but FHI is an important problem, as old as humanity: how to bring up your kids right and stop them wrecking the place.
I’ll take this to the open thread.
You’re right, ‘instead of’ was sloppy phrasing on my side, I’ve edited my comment.