Yet another log has been tossed on the fire of computer chess related artificial intelligence writing. Yawn. This facile piece is a summary of human-machine matches with a few usual stabs about whether or not chess-playing machines are "thinking." Yaaaaaawn. Almost none of it will be new to you, most of it repeats the basic facts, although prefacing many with "rumored to.." in case he gets it wrong.
This doesn't help eliminate a dozen or so factual errors, but we're used to those by now. Even when non-chess writers bother to ask experts (not the case here) they often get it wrong before it makes it to the page. It's hard to write about something as technical as chess when you don't have the background. Knights become bishops, as in this article's description of Kramnik-Fritz match game six. The description of Kasparov-Junior game three is farcical (also game 5). Even when 90% of the information is rehashed, the wrong assumptions are made in the remaining 10%.
The rest of the article is occasionally correct history of chess programming, inaccurately summarized in most cases. Bizarre things like, "It’s the optimization of a chess program rather than the evaluation algorithm that affects the playing manner" are aplenty. And nonsense like, "Having played for a while against chess programs, I came to my own recipe: try to make the best move possible in every situation. When you just make a move that looks like good, without any plan in mind, it may bring you to trouble against the computer." Huh?
And to make his points about how computers don't understand some positions as well as humans he uses a purely tactical example. (Heissler-Kasimdzhanov, 1999). True, it's a deep combination that some programs take a while to find, but others, like Junior 7, find ..Re4 in a few minutes on my machine. Bad example.
