Letter: Deep Blue is not that clever

Sir: While the victory of Deep Blue over Garry Kasparov demonstrates that the machine certainly does play a mean game of chess, the implications for artificial intelligence are less clear.

Intelligence should be a measure of quality of thought, not speed of processing. A very simple program could be written to beat Kasparov provided it could run on a machine of sufficient (currently unobtainable) speed - just evaluate every possible move.

Kasparov's brain is said to process moves at two per second. The quality of his "program" - ie, intelligence - is such that it almost matches Deep Blue's program executed on a machine running at 200 million moves per second. Roughly speaking, that makes Kasparov 100 million times more "intelligent" than Deep Blue.

To make it a fair test of intelligence - not speed - slow down Deep Blue to two moves per second, run the same program and see who wins.