X-Message-Number: 8480
Date: Wed, 20 Aug 1997 09:02:05 -0700
From: "Joseph J. Strout" <>
Subject: Computers and Go

In Message #8475, John K Clark <> wrote:

>When a computer became the world Chess champion a few months ago the
>apologists for biological brains were saying there were plenty of other games
>where humans still had a commanding lead, the two most common examples given
>were Go and Othello. Last week a computer beat Takeshi Murakami, the world
>Othello champion, 6 games in a row. Any bets on how long Go will hold out?

I think Othello only took so long because it's not a very popular game
(comparatively speaking), and so not much effort was put into it.  It's a
very simple game with a narrow breadth and a trivial evaluation function; I
wrote one on the 1MHz Apple II which could routinely beat me (given a
minute or so to think per move).

Go is another matter entirely.  It's a much more complex game, with
difficult evaluation even at the very end, and an extremely broad search
tree.  Moreover, there have been million-dollar prizes for decades for any
computer program which can beat a Go champion, and yet the very best
programs play at the level of a relative novice (e.g., amateur with a few
years' practice).  Eventually it must fall to brute force, in principle,
but it's going to take a LOT of brute force.  Or something much more like
human intelligence instead.

The two traditional challanges to AI put forward by someone-or-other were
Chess and symbolic integration.  These were ridiculous challanges; by that
measure, 99.99% of the population is not intelligent, either.  A far better
challenge to AI would be something we all do easily: pick out our mother in
a crowd, drive while holding a conversation, walk while chewing gum...

,------------------------------------------------------------------.
|    Joseph J. Strout           Department of Neuroscience, UCSD   |
|               http://www-acs.ucsd.edu/~jstrout/  |
`------------------------------------------------------------------'

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=8480