Beautiful Chess played by Humans and Machines07-10-2010
Rebecca Thomas has a long and thoughtful response to my post on Gary Kasparov's article on computers, chess, and humanity. The whole comment is worth reading. But here is how she begins:
Regarding the first of the three comments, I have to take issue with the idea that a chess game played by a computer is necessarily less beautiful than one played by a human. There are various kinds of beauty, and mathematical beauty is a very real thing. Some proofs are more elegant than others, for instance. Some x-y curves are quite beautiful, and often these are captured by particularly compact mathematical expressions. One could wonder why: is this preference for simplicity inherent in our idea of beauty, or have we preferentially developed (mathematical) language to describe things we find beautiful?
Surely, there is beauty to math. And there are various kinds of beauty. An efficient, powerful, and unstoppable game of chess played by a computer may be truly beautiful in its reduction of complexity to simplicity.
The point Kasparov makes is not that rationality is not beautiful in some way, but that it changes the idea of beauty in chess. A bold move, a risky move, a daring move has been valued in the world of chess. Chess, despite its rational reputation, has had an emotional and adventurous side. However, against computers--or even against humans aided by computers--such risks rarely succeed and thus they are devalued.
Chess changes. It becomes less quirky, less risky, and more rational. I don't think it wrong to say that chess becomes less human. Chess, in the age of computer chess, loses the valuation of a particularly human beauty, even if it might reflect an impeccably beautiful mathematical rationality.
It need not be that mathematical beauty is inferior to human beauty, as Rebecca suggests I must mean, but simply that the elimination of the human beauty as a meaningful option in chess is to be regretted.
II. Rebecca's second related point is to concede that chess players are internalizing the values of computer chess and thus playing more and more like computers. But, she argues this is not so new or so bad. Chess has changed before. New theories of chess emerge all the time. Why is this different? Why is this change, she might ask, the tipping point that makes chess less human?
I think the answer is the one given above. The values and approach to chess this particular change inaugurates take one element of chess--its rationality--and elevate it to the only relevant element of chess. All competing theories are judged by their ability to succeed over a hyper-rational strategy, and they will eventually be found wanting. Those who play chess (as opposed to making art with chess pieces) will succumb to the values of computerized chess. While earlier theories of chess may have aspired for complete dominance, only a purely rational computer chess can achieve that aim.