Newsgroups: comp.ai.games
Path: cantaloupe.srv.cs.cmu.edu!rochester!cornellcs!newsstand.cit.cornell.edu!news.acsu.buffalo.edu!dsinc!spool.mu.edu!howland.reston.ans.net!swrinde!newsfeed.internetmci.com!eanews1!rsvl_ns!unirsvl!news
From: Simon Haines <simonehaines@unn.unisys.com>
Subject: Re: Deep Blue and Kasparov
Sender: news@unirsvl.rsvl.unisys.com (News Administration)
Message-ID: <31D33718.6929@unn.unisys.com>
Cc: petervisser@unn.unisys.com
Date: Fri, 28 Jun 1996 01:36:24 GMT
X-Nntp-Posting-Host: 192.39.129.113
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=us-ascii
References: <4qsb7g$3ql@eagle.freespace.net>
Mime-Version: 1.0
X-Mailer: Mozilla 2.0 (WinNT; I)
Organization: Unisys
Lines: 39

Tim Carter wrote:
> 
> I have a brother who feels that "Deep Blue" would never be able to
> out-wit Gary Kasparov in a chess match.  He feels that the computer
> cannot be programmed to react to an unpredictable move maded by a human.
> He also argues that the computer doesn't have enough of a human element
> to it, therefore not being able to make a totally unpredictable move.
> 
> I feel that Deep Blue will eventually wipe Kasparov out.  Deep Blue has
> to have the most amazing artificial intelligence program to be able to
> even challenge a grand master like Kasparov.
> 
> I'm just looking for other opinions concerning this :)
> 
> Tim Carter
> timc@freespace.net

I too feel that one day Deep Blue (or somesuch other program) 
will annihilate Kasparov. I find it difficult to reconcile the 
idea of an 'unpredictable move' in chess though. Due to the 
nature of the game, all moves could be accounted for. Sure, Deep 
Blue would have a hard time of it if you pulled a knight from 
the board and ate it, but it probably would only be phased for a 
mere nanosecond before it replied 'Illegal move'.

The algorithms to beat a World Chess Champion, I believe, 
already exist, it's just hard to get them to fit the 50 moves in 
2 hours chess rule. Sure, it is a cop-out to decry 'processing 
speed, godammit, give me more processing speed!' for Deep Blue 
has already beaten Kasparov in one game. We have the 
tools, with genetic algorithms (I believe the first successful 
implementation of a GA was in Chess: the alpha-beta window was a 
GA before the name existed), neural-nets and fast lookup 
databases. Perhaps a little fine-tuning, a little 
number-crunching and a whopping database added on would make it 
3 out of 3. I hope this thread develops as I am curious to see 
what others think...

-simon
