The very first key conquest of artificial intelligence was chess. The sport has a shocking number of potential combinations, however, it had been relatively tractable as it was organised by means of a set of rules that are clear. An algorithm can always have great understanding of the condition of this game and understand each possible movement which it and its competitor might make. The condition of this game can be assessed just by taking a look at the boardgame.
But many different games are not that easy. If you choose something similar to Pac-Man, then figuring out how the perfect move would entail considering the form of the maze, then the place of these ghosts, the place of some extra areas to clean, the access to power-ups, etc., and also the very ideal strategy may wind up in tragedy if Blinky or even Clyde creates an abrupt movement. We have developed AIs that could handle these matches, also, but they have needed to choose a very different strategy to those who defeated Go and chess.
At least till today. These days, however, Google’s DeepMind branch published a paper describing the arrangement of a AI that could handle both baseball and Atari classics.
See 12 staying paragraphs | Remarks
