Sun, Aug 06, 2006 - Page 9 News List

An old Oriental game of strategy still confounds programmers

By Charles Arthur  /  THE GUARDIAN , LONDON

However Charles Matthews, a 3-dan Go player also based in Cambridge, is unconvinced.

"If you can't beat your computer Go opponent, you are going to be one of the rabbits at the local Go club. But further progress has been by rather small increments. The top programs used to be about 8 kyu; they might have crept up to 7 kyu, but that should mean that I, as a 3 dan, would have a hard time giving them a nine-stone handicap (a nine-turn head start), and that isn't my experience," he said.

While it's hard to know how well humans are playing, Matthews reckons that amateur shodans play at about 90 percent efficiency, and top professionals at about 98 percent efficiency.

"That is with a purely hypothetical notion of `perfect play' being a few handicap stones beyond top pro," he says.

Grapel accepts that gaps remain. But he also thinks there may be a deeper reason why computers remain bad at Go, and humans good.

"I believe Go requires certain human characteristics -- visual recognition, matching shapes and logical reasoning. You have to do spatial reasoning about which direction you should play. It's all about predator and prey, hunting and chasing, and territory. All these are very basic yet complex human facilities. They're all useful in our natural environment for other things," he says.

The implication is that computers are bad at Go because they're still bad at being human. Which might come as some relief.

In 2002, David Levy, one of the earliest drivers of computer chess, wrote: "Perhaps Go will be the final bastion in man's attempts to stave off his inevitable intellectual defeat at the hands of the machine."

Despite humanity's own best efforts to undermine it, the bastion still looks remarkably solid.

This story has been viewed 5509 times.
TOP top