Author: Terry Giles
Date: 05:37:01 04/15/05
Go up one level in this thread
On April 13, 2005 at 21:52:48, Mike Byrne wrote: >On April 13, 2005 at 20:07:18, Walter Faxon wrote: > >>A brief article in "New Scientist" about 5 years ago suggested that future >>quantum computers will be able to attack chess by doing the quantum equivalent >>of a full-width search 300 ply or more deep. I googled today to find the >>following material: >> > > >I think the word quantum understates the jump required to do a full width 300 >plyu search. In the near future (next trm years), will continue to see a >doubling of processor speed every two to three years. I also see a patternm of >diminishing returns for the faster and speeds at these higher levels. Doubling >the speed at 1800 might have provided us with 100 elo points (or more), >Doubling the speed at 2300 ELO might have provided 50 elo points or so. Perhaps >doubling the speed at 2800 might provide an ELO of 25 point gain (against >humans). It is possible that computers 10 years from now will not be >convincingly better against the top humans than they are today. Why is that ? >Top human players bring a feel to the game that is very hard to quantify, >program and measure that is beyond calculating abilities of even top programs. > >Naturally if there is a shift in the processor speed doubling every 2 to 3 years >paradigm -- say from every 2 to 3 years to every 2 to 3 days - things may end >upa lot different. But without that shift, my prediction that in the next 40 >years - Chess ELO on a home PC <= Top 10 GM Elo under tournament time controls. >Not too much different from where are today imo. > >I am probably in the minority with this opinion - and feel free to tell me how >wrong I am! > >Waiting for Rebutal Regards, > >Michael Hi Mike, You may have seen this rather tongue-in-cheek mail I posted here a while back, if not I hope you like it... (A slightly modified original post by : Terry Giles on June 20, 2003) The perfect chess-playing machine 'Quantum-Chess' amazingly resigned its first game after only twenty moves in its match against the current human world champion Iris Zhine. The English GM opened the game with her favourite g4 and after reaching an interesting and apparently equal position ‘Quantum-Chess' promptly resigned. Question: Assuming the machine really can play perfect chess, and Iris Zhine cannot, why did ‘Quantum-Chess’ resign? ANSWER v v v v v v v v v v v v v v v v v v v v v v v v ‘Quantum-Chess’ plays perfect chess and as such saw that after Iris Zhine's 20th move its position was lost (white mates in 47 moves against any defence) and it assumed, almost certainly erroneously, that its human opponent would find the correct continuation to win the game. Iris Zhine had been partly lucky in finding the first twenty moves of a perfect chess game. If ‘Quantum-Chess’ had been unable to utilise its opening book, which in this particular opening line was twenty moves deep, it would have resigned immediately after Iris Zhine's opening move of g4 (white mates in 66 moves). Note: ‘Quantum-Chess’ as its name implies plays chess by utilising massively parallel quantum computation and therefore is able to play chess perfectly ;-) However lacking as it does any psychological dimension it failed to realise that its opponent would almost certainly be unable to find the winning continuation from the resigned position. ‘Quantum-Chess’ should have played the opponent and not the board. THERE’S HOPE FOR US ALL!
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.