Hacker Newsnew | past | comments | ask | show | jobs | submitlogin
Artificial Intelligence as Alien Intelligence (boundary2.org)
23 points by mgunes on April 5, 2015 | hide | past | favorite | 3 comments


Ok, read it, not worth reading. Go read Stuart Russell instead : https://www.cs.berkeley.edu/~russell/papers/russell-edge14-m...

1) He is an active and substantial participant in the community. 2) It's about 500 words 3) It makes sense 4) It doesn't mention Freud 5) It uses short words


"Any sufficiently capable intelligent system will prefer to ensure its own continued existence and to acquire physical and computational resources – not for their own sake, but to succeed in its assigned task."

I am always surprised when people say this kind of thing. A human-utility designed AI would identify "it's own" to be decidedly human. And thus it's assigned task would be human. Assuming you properly pinned down human utility sufficiently to call it a successful AI.


I agree with you - I think that being frighted of AI's is as rational as being frightened of children, but Prof. Russell is a more authoritative source than me, and I think it would be better for people to read his stuff than the nonsense that the parent links to.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: