“Exactly. But you don't know which ones let you win because…” — Dwarkesh Podcast at 1:36:13
“Exactly. But you don't know which ones let you win because they were truly better versus winning on accident. How would you design a baseline where it's truly better? Yeah. So this is where in RL, people use things like TD learning to better approximate the quality function, the cube that”
0:00
2:37:29