How can AIs know what we want if *we* don't even know? (with Geoffrey Irving)
Clearer Thinking with Spencer Greenberg
00:00
Explaining Complex Processes vs Doing Them
Sometimes we find it challenging to explain a simple task, like tying a shoe, even though we can easily do it ourselves. Similarly, neural nets perform intuitive leaps to process information, which may be different from how our brains work. The challenge lies in asking these systems to explain their processes and understanding the relationship between intuitive leaps and the linear algebra that produces the answers.
Transcript
Play full episode
