Last summer, the chess-playing supercomputer Deep Blue sent Garry Kasparov, the world's best human player, crashing to defeat. But the invention of a truly intelligent machine, a computer that simulates human reasoning, is still a long way off. A team of scientists at CalTech in Pasadena, however, are working to bring the dream of a computerized "thinking machine" one step closer to reality with the creation of the Bozeman Brain, named after the Montana town where they plan to do the major work. Toward that end, James Bower, a computational neurobiologist, and his colleagues, are using computer modeling techniques to understand how our neural circuitry -- the so-called wetwiring of the brain -- actually works. Over the last several years, they've developed a general purpose neural network simulation system, GENESIS, to support the construction of structural simulations of real neural networks.

By observing and analyzing computer simulations modeled on real brain structures, they hope to understand how the brain learns, remembers and recalls. What they glean from these experiments will be used to build the scaffolding of the artificial intelligence community's long sought holy grail: a computer that can mimic the intricate processing of the human brain.

--Linda Marsa