Dylan Hadfield-Menell, UC Berkeley/MIT: The value alignment problem in AI
Generally Intelligent
00:00
Imitative Learning
One of your favorite papers is called, on the folly of rewarding a and hoping for b. A paper in imitation learning that i just think it's a fantastic idea captures a lot of what's going on. Data set aggregation helps explain how failure mining systems can go wrong when they're deployed.
Play episode from 01:25:43
Transcript
