Triangulating General Intelligence, Cognition, and Cortical Microcircuits
Dileep George, Co-founder of Vicarious AI
Thursday 5 November 2020
Building artificial general intelligence, and understanding the brain are grand challenges. Can trying to solve both these challenges simultaneously be easier than just tackling one? In this talk I’ll present a systematic approach that could kill two birds with one stone. In this approach, observations from cognitive science and neuroscience are translated to their computational counterparts to build artificial systems that generalize like the brain. Working back from these artificial systems then help to understand cortical circuits. This approach is used to derive a detailed functional model for visual cortical microcircuits, by working back from a generative model for vision. Efficient inference and generalization guided the representational choices in the original computational model. The cortical circuit model is derived by systematically comparing the computational requirements of this model with known anatomical constraints. The derived model suggests precise functional roles for the feed-forward, feedback, and lateral connections observed in different laminae and columns, assigns a computational role for the path through the thalamus, predicts the interactions between blobs and inter-blobs, and offers an algorithmic explanation for the innate inter-laminar connectivity between clonal neurons within a cortical column. I will also describe how we have used this overall approach in learning concepts as cognitive programs, and in learning cognitive maps as structured graphs.
Dileep George, PhD | Dileep is co-founder & CTO of Vicarious AI, where he leads the research and development of a general intelligence layer for robotics. Vicarious is an AI+robotics startup with backing from several well known investors that include Mark Zuckerberg, Jeff Bezos, Founders Fund, and Khosla Ventures. Vicarious has pioneered brain-inspired and data-efficient AI algorithms, and they gained world-wide attention for fundamentally breaking text-based CAPTCHAs with very little training data. Dileep has authored several influential papers, published in Science, NIPS, ICML and CVPR and his research has been featured in NYT, WSJ and NPR. Before cofounding Vicarious, Dileep was CTO of Numenta, an AI company he cofounded with Jeff Hawkins and Donna Dubinsky during his graduate studies at Stanford. Before Numenta, Dileep was a Research Fellow at the Redwood Neuroscience Institute. Dileep’s research on hierarchical models of the brain earned him a PhD in Electrical Engineering from Stanford University. He has an MS in Electrical Engineering from Stanford and a B.Tech in EE from IIT Mumbai.