Ok, maybe that's being a little over-dramatic, but it's kind of true. A computer system named Chinook that was finished in Canada in April can never lose at checkers. The reason for this is that it can look at the state of the game and from that point, predict all possible outcomes based on all possible moves (there are 500 billion billion possible moves, according to iafrica.com). From there it can create a kind of tree graph that shows the clearest way for it to win. Meanwhile, its human counterparts can, at best, look a few moves ahead and will probably miss a number of possibilities along the way. It is impossible to win, but one can create a draw if they are good.
I decided to try out this mind-destroying mechanical masterpiece for my own. I went on down to Chinook's homepage and tried to play against it. At first it simply told me how much of an advantage it had over me. Then, as my terribleness at playing checkers became obvious, it just came out and suggested we play again (as it had already determined I had no possibility of winning). That was depressing.
That being said, it does raise an interesting point about how things we consider "games," really aren't games when your mind is as large as a computer's. It would be like being shown a rack of tubes into which one can place balls and, if the tube is chosen correctly, the player moves on to the next round with less tubes. If one could see all of the possible moves and their outcomes, it would be like playing the tube game with only correct tubes open to have balls dropped into them. You can't lose! Is it really a game at that point?
Blogged with Flock