A related chapter on artificial intelligence by Roger Schank highlights both connectionist (neural network) models as well as algorithmic (symbolic) computer simulations. Artificial intelligence researchers are forced to leave the realm of abstract concepts and wrestle with the myriad of practical challenges associated with forcing a computer not to function like a computer (inflexible and intolerant of ambiguity and error) but with the fluidity and flexibility of human intelligence. I hoped for a more extended discussion of chess-playing computers, the most obvious computer-based challenge to human intelligence. The world champion Garry Kasparov has been defeated by Deep Blue, a supercomputer develop by a team of IBM engineers. Does Deep Blue exhibit "true" intelligence in view of the fact that it can only play chess and has no capacity to perform other tasks? My bias favors a "yes" response, but even if the verdict is "no," a closer examination of the functional architecture of Deep Blue—it is composed of a very large number of computational modules functioning cooperatively and in parallel rather than a single, standard, sequential processor—may provide key insights into the architecture of human intelligence.