Something incredible about the A.I. of chess computers
Aug 17, 2007 by Albert Frank
I gave the following position to the best chess software, without their tablebase ( = knowledge by memory of all positions with maximum six pieces on the board)
- White (to move) : Kg7, Rb1; h7
- Black : Ke6, Ra8; e4
A good chess player (minimum « first category ») sees immediately that 1.Rb5 wins (Black will soon have no valid move) and that the promotion 1. h8Q gives only a draw.
Computers, after two hours (on a 3 GHz Pentium 4 with 1 Go RAM) even don’t look at the winning move, and just stick at 1. h8Q, so seeing (or thinking?) in an incredible poor way.
Once more, these computers are sometimes like “great players, strongest than any human”, and sometimes “absolutely stupid”.
I don’t know if any conclusion can be taken out of this for the moment.
Addendum: After 8 hours, the computer finds the correct solution.