Author: Steven Edwards
Date: 18:35:25 04/18/05
Go up one level in this thread
On April 18, 2005 at 20:50:52, Mark Ryan wrote: >http://news.bbc.co.uk/2/hi/science/nature/4449711.stm > >"But when Moore's Law is effectively slowed down in about 10 to 20 years' time >..." > >A few years ago, Grandmaster Lev Alburt stated that chess computers would never >be stronger than the strongest humans. If there is a practical (or asymptotic) >limit to computer speed, maybe he was right. "For computers, over time the price per kilogram is roughly a constant." Alburt is far too presumptuous. Many proposed algorithms for chessplaying have not yet been well explored. Classical problems like N-queens and the Travelling Salesperson were once thought to be intractable for large sizes; now they have trivial solutions or near solutions, and this done on modest hardware. Could a near trivial algorithm be discovered for chess? I doubt it, but it's not an impossibility. My personal effort towards a cognitive planning program, if successful, could significantly change the expectation of playing strength for a particular hardware platform. If it fails, then there are still other alternatives to A/B and eventually these will be more fully examined.
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.