Here's Looking at You; January 1999; Scientific American Magazine; by Beardsley; 2 Page(s)

Parties have a way of generating outrageous ideas. Most don't survive the night, but a scheme that bubbled to the surface at a 1992 event held by Rodney A. Brooks of the Massachusetts Institute of Technology is changing the way researchers think about thinking. Brooks, the head of M.I.T.'s artificial intelligence laboratory, was celebrating the switch-on date of the fictitious Hal 9000 computer, which appeared in the movie 2001: A Space Odyssey. As he reflected that no silicon brain could yet rival Hal's slick mendacity, he was seized by the notion of building a humanoid robot based on biological principles, rather than on conventional approaches to robot design.

The robot, known as Cog, started to take shape in the summer of 1993. The project, which was initially to last five years, is intended to reveal problems that emerge in trying to design a humanoid machine and thereby elucidate principles of human cognition. Instead of being programmed with detailed information about its environment and then calculating how to achieve a set goal-the modus operandi of industrial robots-Cog learns about itself and its environment by trial and error. Brooks says that although there are no near-term practical goals for Cog technology, it has stimulated "a bunch" of papers.