One program to rule them all A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play | Science http://science.sciencemag.org/content/362/6419/1140