Paul Christiano
Person on 2 sites
Alignment researcher; developed reinforcement learning from human feedback at OpenAI, founded the Alignment Research Center, and now works at the US AI Safety…
-
1 channel on ppll (opens ppll.app)
-
6 beliefs on korrents (opens korrents.com)
-
A slow takeoff is significantly more likely than a fast one: AI that is nearly as powerful will have transformed the world before the incredibly powerful kind arrives. (opens korrents.com)
“I currently think a slow takeoff is significantly more likely.”
-
There will be a complete four-year doubling of world output before there is ever a one-year doubling. (opens korrents.com)
“There will be a complete 4 year interval in which world output doubles, before the first 1 year interval in which world output doubles.”
-
Expecting a slow takeoff means expecting AI to matter more and sooner, not less; the disagreement is about what happens before superintelligence, not after. (opens korrents.com)
“The main disagreement is not about what will happen once we have a superintelligent AI, it’s about what will happen before we have a superintelligent AI.”
-
A slow takeoff is significantly more likely than a fast one: AI that is nearly as powerful will have transformed the world before the incredibly powerful kind arrives. (opens korrents.com)