There is a long and thoughtful essay (that is masquerading as a book review) by Garry Kasparov in NYRB, where the former world champion takes a winding personal detour discussing the meteoritic rise of computer chess between 1994 and 2004 from its humble origins to the present state. He makes several very intriguing and interesting points along the way that have much broader ramifications than simply chess. I am not going to recap them here and will instead suggest that you go and read the entire thing.
By a strange coincidence, I have lately picked up an old hobby of mine that I had sorely neglected for several years, mostly in favor of playing poker. I have been fumbling my way around playing the game of go again, and it has been a lot of fun despite of my being incredibly rusty. When Kasparov lost to Deep Blue and the final triumph of computers in chess was imminent, it was go that was often brought as an example of a game that couldn’t be crushed by brute computing power the way chess had been. Enthusiasts of go, probably even more so than chess players, have always been very susceptible to making a great deal out of the game’s inherent beauty, an aesthetic dimension that computers simply cannot grasp. This elegiac tone is also present in Kasparov’s article. One thing that caught my attention was his gloom over the failed promised of computer chess, which has approach-wise stagnated to brute force search of billions positions plus some game-tree pruning algorithms and opening/endgame databases for standard situations. Kasparov laments that the makers of computer chess programs are content with their creations playing better than their strongest human opponents simply by overpowering any strategic advantage that a man may have by superior tactic ability; and is clearly implying that the true challenge would be to make a computer play chess the way we humans do—and that subsequently, humans are increasingly playing like computers. However, what if “the human way” is simply a reflection of our inherent limitations? What if “strategy” really is just a shorthand for our limited tactical horizon? In this case, writing a program that would play chess like human would make as little sense as trying to create a car that walks on two legs.
The article also reminded me of a book—Yasunari Kawabata’s Meijin (“The Master of Go”)—that is a similar elegy for the beauty of the lost world. It talks about a game of go that was played in Japan in 1938 between the old master and a young contender, which Kawabata actually did report on for a newspaper at the time. The novel casts the game as a metaphor for the transition that the very Japanese society was going through and refers in a very subtle way to some of the rather neglected consequences of Japan’s drive to modernize.
“From the way of Go the beauty of Japan and the Orient had fled”, says Kawabata. “Everything had become science and regulation. The road to advancement in rank, which controlled the life of a player, had become a meticulous point system”.
Kawabata himself considered Meijin his best work. It was first published one year before Alan Turing conceived an idea of a machine that would play chess, and yet the issue at the heart of the book is very much the same that Kasparov was faced with in 1997 when he was forced to resign to the “$10 million alarm clock” designed by a team of engineers at IBM. Both Master’s loss to a challenger in Meijin and Kasparov’s loss to Deep Blue marked the beginning of a new era, and an end of an old.
In this sense, Kasparov’s piece can be read as an elegy for the beautiful world of human thought. Of course, humans will go on thinking, and probably thinking better in the instrumental sense of the word than they have been ever before. However, it may be that precisely because of our limitations, the human mind has been forced to create shortcuts, devise strategies, and, for the lack of more precise criteria, judge them on an aesthetic basis. And removing those limits would then render obsolete something that not only has been held as deeply beautiful, but also something that has made us truly human.