“This is fine. You can do this. And this probably does…” — Dwarkesh Podcast at 56:04
“This is fine. You can do this. And this probably does work. But in practice, having a single forward pass that gives you a pretty good guess is how the breadth is is is is proved out. The there is a sort of duality here. Like it would be weird if let's say the policy recommended an action that”
0:00
2:37:29