Computer chess books are annoyingly rare. There's an abundancy of chess books on the market, ranging from simple moves for kids to full strategic treatises written by the best GMs and chess analysts of the past. There are a lot of AI books that touch the issue of chess, but *real* computer chess books are a precious few. This is peculiar. Chess is the fruit fly of AI (a beautiful analogy borrowed from this book) - there can be no doubt about that. I would expect more material on this topic, at least during the 80s and 90s when chess-machine development still didn't reach a point of saturation. Anywayz, to the review... This is an old book, dated 1975 or so. When this book was written, for example, there was still no Kasparov as a total world champ, no Deep Blue, no Fritz/Junior/Shredder. In short, no programs that routinely challenged the world champions. It shows in the author's opinion. For example, he makes the utterly wrong prediction that "no computer expert chess player will be developed until we gain an understanding of how the human brain works". Palm-based chess players beat any beginner, desktop programs like Junior beat most masters, and on slightly stronger machines (not the Deep Blue monster, but a measly 4-way Xeon) they tie with world champions. The discovery of how the brain works, and especially its application to programming looks still very, very far away. Being severely outdated, most of this book is good only from a historic view-point really, but it does contain some interesting articles that are relevant today. For instance, a basic analysis of human chess skill, what makes us know how to filter moves, what makes some players better than others, what makes GMs tick. There's also a quite understandable treatment of computer chess basics that could be interesting for beginners. The description of the best US chess program of that time - CHESS 4.5 (I say US, because a Soviet KAISSA was comparable in strength) is also interesting, though it leaves some brows raised if you're a programmer. The authors of the program wrote it all in about 2-3 functions that closely intertwine - god help the one who'd have to grok the code. All in all, part of this book make an interesting read if you're a computer chess enthusiast, but don't expect too much from it. I even can't recommend another book, since I'm not aware of good modern books on this topic (maybe it's time to write one ?)

Comments

comments powered by Disqus