He was actually quite gracious about the defeat and his readiness for the challenge, but acerbic on IBM not playing particularly fair and doing what they could to stack things against him psychologically, i.e. increase their chances of winning beyond the pure merits of Deep Blue.
Google imho did something similar to Lee Sedol. If he had understood that AG had no book it learnt, wasn’t optimised to play against him, but was a ‘generalist’ application, he could have approached the game very differently, playing freely in a style that he was most comfortable with, and fuseki he knew best. Instead he tried to play ‘unusually’ from the off, which seems to have meant playing not just weak moves, but moves that played against his comfort zone.
Kasparov ranges quite freely about the power to learn from AI and the good that comes of it. One passage that struck me was how he evaluates a position, a bit like someone who likes cakes and walks into a cake shop. They instantly know what they think will taste nice, but then should spend time carefully reading ingredients to know there’s no surprises. It was a neat evocation of intuitive understanding…
