DeepMind's self-play reinforcement-learning system, cited in the Chess for Programmers lightning talk as a milestone in AI tackling chess, following Deep Blue's 1997 victory over Kasparov.