Met with Stuart Russell to talk about the future of AI and his new book, Human Compatible:https://www.vox.com/future-perfect/2019/10/26/20932289/ai-stuart-russell-human-compatible …
-
Show this thread
-
Russell's key point: right now, we design AIs with objective functions that don't encompass what we actually want. That means that the "smarter" they get - the better they optimize for something that's not what we want - the worse off we are.
1 reply 2 retweets 11 likesShow this thread
He argues that if we instead design AIs whose objective is to get a better model of humans and do what we want, then there's still a lot which can go wrong - but more competent AI means better outcomes instead of meaning worse ones. That's a safer paradigm to be working in.
9:28 AM - 26 Oct 2019
0 replies
2 retweets
13 likes
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.