Tuesday, June 17, 20082:00 PM - 3:00 PMCNLS Conference Room (TA-3, Bldg 1690)|
Learning By Learning To Communicate
Jake BEAL, LANL Director's Postdoc AwardeeMIT, Computer Science and Artificial Intelligence Laboratory (CSAIL)
Human intelligence appears to be a product of cooperation among many specialists. I show how a system of specialists can capture cross-specialist knowledge through a struggle to agree on signals for communicating with one another. This process of "learning by learning to communicate" may help to explain how humans develop our unique capacity for the "high-level" agile cooperation that permeates our daily lives. I have created mechanisms for signal agreement that exploit the phenomenon of "communication bootstrapping," in which shared experiences form a basis for agreement on a system of signals. These mechanisms are demonstrated using a vision specialist and a hearing specialist that jointly observe a simulated four-way intersection. As they agree on signals, the two specialists capture some dynamics of the simulation as differences in how they interpret these signals.
Jacob Beal is a postdoctoral associate in the Computer Science and Artificial Intelligence Laboratory at MIT, where he recently completed his Ph.D. under Prof. Gerald Jay Sussman. His research interests center on the engineering of robust adaptive systems, with a focus on problems of system integration for human-level intelligence and on problems of modelling and control for spatially-distributed networks like sensor networks, robotic swarms, and cells during morphogenesis.
Host: Christof TEUSCHER, CCS-3 and Reid PORTER, ISR-2