How do Chess Engines Work - Looking at Stockfish and AlphaZero

Explore the inner workings of chess engines in this comprehensive conference talk from MLCon. Delve into the classic AI discipline of game playing, starting with the breakthrough of Deep Blue defeating Kasparov in the 1990s. Examine the algorithms that made this success possible, including Minimax and AlphaBeta pruning, which are still utilized in leading chess engines like Stockfish. Focus on the advanced use of Monte Carlo Tree Search in AlphaZero, which operates without human heuristics or opening libraries. Learn about AlphaZero's self-play training process using a convolutional ResNet architecture. Conclude with an analysis of a notable game between Stockfish and AlphaZero, discussing the potential end of the classic chess engine era.