Cambrian Intelligence: The Early History of the New AI

Will we build intelligent robots from the brain down, or from the legs up? Rodney A. Brooks, the controversial director of MIT's Artificial Intelligence Laboratory, is betting on the latter--and the smart money's following him. Cambrian Intelligence, composed of eight papers written between 1985 and 1991, explores the technical and philosophical aspects of behavior-based robotics, offering much for the interested mind to ponder. Does cognition mediate between perception and action, or is this an illusion? Can a robot be called "intelligent" if it lacks anything we would call a brain? Brooks, simply by asking these questions, launched a new movement in artificial intelligence, and these brash, bold papers show how he laid the work for his eventual conquest of Mars with tiny autonomous robots.

Whether you're new to Brooks, know him from his work, or saw him in the documentary Fast, Cheap, and Out of Control, you'll find a lot to love in Cambrian Intelligence. The four technical papers do inspire moments of anxiety in nontechnical readers, but the four philosophical papers are splendidly thought-provoking and even a bit confrontational. (Reviewers have called his work "inflammatory," but Brooks earns this privilege because he's right.) While the work reported here is several years old, it still feels fresh and new to those of us who grew up reading about top-down artificial intelligence as if it were the only and obvious design choice. The next generation of intelligent robots are growing from the bottom up, and you can get in on the ground floor with Cambrian Intelligence. --Rob Lightner