Dylan Hadfield-Menell, UC Berkeley/MIT: The value alignment problem in AI
Generally Intelligent
00:00
How Do You Filter Resumes?
I think one of the big outcomes of my work is thi split between objective and subjective. I think that our fields chooses to focus on the wrong things here, almost directly. One example that we use in an m l systems is bias and hiring. And people are sort of surprised by the fact that this bias exists, or what are we going to do about it? We've got this bias data. What options do we have? It's amazing that we ent stopped to think very much.
Play episode from 47:45
Transcript
