Dynamic Models for Intention (Goal-Directedness) are Required by Truly Intelligent Robots

Walter J Freeman
Department of Molecular and Cell Biology
University of California
Berkeley CA 94720-3206 USA
wfreeman@sulcus.berkeley.edu

Abstract
Intelligent behavior is characterized by flexible and creative pursuit of endogenously defined goals. Intentionality is a key concept by which to link brain dynamics to goal-directed behavior, and to design mechanisms for intentional adaptations by machines. Evidence from vertebrate brain evolution and clinical neurology points to the limbic system as the key forebrain structure that creates the neural activity which formulate goals as images of desired future states. The behavior patterns created by the mesoscopic dynamics of the forebrain take the form of hypothesis testing. Predicted information is sought by use of sense organs. Synaptic connectivity of the brain changes by learning from the consequences of actions taken. Software and hardware systems using coupled nonlinear differential equations with chaotic attractor landscapes simulate these functions in free-roving machines learning to operate in unstructured environments.

Speaker
Walter J Freeman studied electronics in the US Naval Reserve 1944-46, physics and mathematics at MIT, medicine at Yale (MD cum laude 1954), internal medicine at Johns Hopkins, and neuropsychiatry at UCLA. He has taught neuroscience in the University of California at Berkeley since 1959, now as Professor of the Graduate School. He has received a Guggenheim Fellowship, a MERIT Award from NIH in 1990, a Pioneer Award from the Neural Networks Council in 1992, was President of the International Neural Network Society in 1994, and is a Life Fellow of the IEEE. He has published over 400 articles and 5 books on brain dynamics.