ABSTRACT: AI technologies are advancing at an ever-increasing pace, and yet we are still far from any real AI: We have no machines with the flexible,general-purpose intelligence that is the hallmark of human intelligence.  I will talk about what might be needed to close this gap, drawing on insights from research in cognitive science and especially the study of cognitive development in human infants.  In particular, I will present recent progress in our group on building AI systems with some of the basic commonsense reasoning, perceiving and planning abilities we see in young children, as well as the ability to learn new concepts from very small amounts of data — often just one example.  The talk will not be overly technical, but I will describe some of the technical innovations that make this work possible: tools from probabilistic programs and program induction, in addition to new kinds of neural networks. BIO:
 Josh Tenenbaum is the Paul E. Newton Career Development Professor of Cognitive Science and Computation in the Department of Brain and Cognitive Sciences, and a member of the Computer Science and Artificial Intelligence Laboratory. He received his Ph.D. from MIT in 1999 and after a brief postdoc with the MIT AI Lab, he joined the Stanford University faculty as Assistant Professor of Psychology and (by courtesy) Computer Science. He returned to MIT as a faculty member in 2002. He currently serves as Associate Editor of the journal Cognitive Science, and he has been active on the program committees of the Neural Information Processing Systems (NIPS) and Cognitive Science (CogSci) conferences.

Hosted by Professor Stefano Soatto

REFRESHMENTS at 3:45 pm, SPEAKER at 4:15 pm

VIDEOTAPED LECTURE: