let a human play chess against a strong computer and (unsurprisingly) lose. analyze the game and determine the losing move. this is, or could be, interesting, representing human thought. mistakes make us human. (though sometimes, perhaps often, the losing move is a blunder, not that interesting).
to keep things simple, the computer should not deliberately give up a winning advantage once it gets it. before that, it could play on an easier mode, deliberately making mistakes, to keep things entertaining for the human. one could also find interesting instances of the human giving up a winning advantage, though, with this initial easy mode, the human might not get punished for it.
(previously, the computer deliberately giving up winning advantages.)
playing out a game with the engine against itself (in strong mode) from a position hypothesized to be lost for the human could estimate ground truth, determine whether a move was a losing move.
should the human play anti-computer chess? the human losing to a "computer move" is not that interesting, but a human playing anti-computer chess is also not that interesting.
No comments:
Post a Comment