With its Alpha series of game-playing AIs, Google’s DeepMind group seemed to have found a way for its AIs to tackle any game, mastering games like chess and Go by repeatedly playing itself during ...