Board games
fromArs Technica
1 week agoFiguring out why AIs get flummoxed by some games
DeepMind's AlphaGo and AlphaChess fail at impartial games like Nim where winning depends on intuiting mathematical functions rather than pattern recognition from self-play training.