This is a book from which I am learning very little. There is a lot of fluff and very little real information.



Although it does cover all the main genres, the treatment is not very thorough. In particular, there are several popular methods and key ideas that are completely missing from the book. For instance, Behavior Trees and utility functions are not even mentioned.



The chapter on classic games devotes less than two pages to alpha-beta search, which means it's useless. The author doesn't seem to know about Monte Carlo methods applied to board games (something that has revolutionized computer go in the last 3 years and which could be applied to a large class of games).



On the other hand, Artificial Neural Networks and Genetic Algorithms get their own chapters, even though it's hard to find any applications for them in Game AI.



In my opinion the book misses most of the things it should talk about, and this is particularly unforgivable in such a thick book.