Appears On
Episode Appearances
Dwarkesh Podcast · Sep 26, 2025
Richard Sutton – Father of RL thinks LLMs are a dead end
“Developed TD-Gammon using temporal difference learning to beat world backgammon champions; precursor to AlphaGo”
Reinforcement Learning FundamentalsLarge Language Models vs. Experience-Based LearningWorld Models and Transition Models
View Analysis