Greg Brockman (Part 2)
Tetragrammaton with Rick Rubin
00:00
Sources of training data and reinforcement learning
He discusses using public internet data and increasingly model-generated and experiment-driven data for learning.
Play episode from 01:22:30
Transcript
