Thursday, December 08, 2005

New Yorker Article on Computer Chess

The latest issue of The New Yorker magazine (December 12, 2005) features an article by Tom Mueller titled "Your Move: How Computer Chess Programs Are Changing the Game." As Mig points out, its information value for serious players is probably rather low. But for the general New Yorker readership, it does a good job of portraying the highest levels of computer chess (including the Hydra-Adams match) in an accurate, understandable, and interesting way. I think it's unfortunate, however, that those highest levels are all that gets portrayed in any public discussion of computer chess. The problem is that the general public can easily get the impression that chess is essentially "solved" (which it is not) or at least that there is not much point in pursuing it since computers are already so much better at it than we are. You might as well play poker, where people can at least still bluff their way to victory against the risk-averse silicon beasts. Moreover, by sticking to the story of the world's best computers we miss out on what to me seems the more interesting story of how readily available GM-strength chess computers have helped to popularize the game like never before.

The focus of Mueller's piece is the computer programmer and self-described average player Chrilly Donninger, who is best known as the man who developed Hydra (with the help of all-important Saudi funding, of course). He is an interesting character who reminds me of the other chess programmers I have met over the years. For one thing, he is much more interested in technology as a problem to be solved than he is in the people it might help. As Mueller tells us, in fact, "Doninger is no longer intrested in man-versus-machine matches" -- nor in any chess games played by those blunder-prone player of flesh and blood. Even the prospect of Kasparov vs. Hydra (the unlikely but exciting dream match-up for many chess fans) leaves him cold. He says: "I'm much more intersted in beating Shredder, Fritz, and the other programs... I learn more from those matches."

Personally, I find computer matches rather ludicrous. But maybe I just haven't been paying attention to them for a few years. Maybe computer vs. computer games are more interesting these days. I assume, however, that they are still more interesting for programmers than they are for chessplayers. We are looking for plans and ideas, counterplans and strategies. Computer chess doesn't really offer that.

It must be said, though, that Mueller has done an excellent job of portraying his subject. I especially enjoyed his clear discussion of the rise of brute force programming and of the issues GM-strength chess programs raise about the definition of intelligence. After reading the article, I feel more convinced than ever that, as philosopher Mark Greenberg is quoted as saying, "there's no reason in principle that a computer couldn't think, have beliefs and oher mental states, [and] be intelligent." We just have not quite reached that level of complexity yet. But we keep getting closer and closer and it is no longer a matter of "if" but "when."

Good articles on chess in major publications are always welcome for promoting the game. I simply wish I'd see more articles about the human side of the game than the "man versus machine" angle so dominant these days. Perhaps "man versus machine" is the new "Cold War" for which chess has become a metaphor. And there is some fun to these articles for the philosophically minded. It's just that I'm more interested in the smaller, local story. And if the story must be about computers, then I wish it would survey the larger world of chessplayers to see "how computers are changing the game."

The wide availability of GM-strength chess programs (some even freely downloadable from the internet) has made chess a more interesting and enjoyable activity for everyone. Chessplaying computer programs give anybody an instant opponent, coach and trainer, analysis partner, and chess-publishing assistant. Computers have raised the level of everyone's game at every level. And computers have transformed the game more generally. Today, anyone can go online and instantly play against someone else anywhere in the world. Or download recent games for analysis. Or gain access to articles about chess history or the latest theory. The list goes on. I'm rather more interested in how people are using computers than what the computers are off doing on their own. They are inevitably part of the game. I just wish the story were less how they are taking it over than how they are helping it grow.

8 comments:

Patrick said...

This is easily the best chess blog/site on the web (IMO).

Very professional, businesslike writing. You could, like, teach business writing or something.

Michael Goeller said...

Funny, I actually do teach business writing... I guess you knew that. But thanks for the amusing note. :-)

Demian Entrekin said...

Michael,

I wonder if perhaps you have missed one of the key points in the article. Or perhaps this is merely a question of my bias vs. your bias. Either way, what I find fascinating in Mueller's article is the idea that there is a new battle ground emerging, and it is not about to subside any time soon. The ground has shifted to that of the better programmer. Chess in this context becomes a worldly problem set for the programmer to attack. Notice how there are different programming styles vying for the superior strategy, and chess is now one step removed from the battle. This actually has immensely profound implications for the world we live in. Think about economics, warfare, nano-technology, robotics, etc. What we are witnissing is the rise of the programmer as the world shaper, and the question I have is whether or not they are wise enough to take on the task...

- Demian

Demian Entrekin said...

I mis-typed my link...see below.

Anonymous said...

Nice recap of Mueller's article. I found it to be fascinating as well. Though with regard to:

"(with the help of all-important Saudi funding, of course)"
Umm... he was funded by a Sheik in the UAE, not exactly Saudis. It would be like comparing Canadians to Americans.

Michael Goeller said...

Thank you for the comments. I especially appreciate the correction regarding "Saudi" vs. "UAE." I was writing from memory and did not double-check the article itself--though, frankly, I have to admit that I had simply assumed he was Saudi.

As regards the comment from "demian" that I have not recognized the ways that programming as competition is parallel to chess, I confess I had not. And no doubt it is a difference of "bias" between us. As I clearly say, in fact, my own bias makes me more interested in the ways that people are using computers to play or study than the ways that programmers are using chess as a model to solve problems. I do agree that this is an interesting aspect of "computers and chess." I only wish that we would occasionally get a story that was a bit more focused on chess than computers, since the typical "computer chess" story, which suggests that chess has been solved by computers or that it is of secondary importance, has the potential to be more detrimental than than supportive of popularizing the game.

Dan Heisman said...

Michael,

Yes, I have made the same observation - vis the public misperception about chess being solved - on my web radio show Thursday night on ICC Webcast (formerly Chess.fm). Regards,
NM Dan Heisman
Novice Nook at Chess Cafe

Anonymous said...

I used to be an avid internet chess player, playing endless games on FICS and having some impressive wins among many woeful losses and humiliations. Then I discovered something which would not be a surprise to most of my net opponents, that Shredder beats me, humiliates me, trounces me, slaughters me every time. It has drained me of all interest in chess. This is the drawback, that silicon chips will demoralise us, thrash us all the time and leave us cold and depressed.
Chess should exist on many levels, not just the cold calculating machine one.
Computers should enhance us, not destroy our will.
Chess is no longer of interest to me directly because of my experience with computers- and to put this up here I have to type a "word", the "word" is "shfynhr"... oh, "01010010100101" !