The title of his lecture is Advances in modeling neocortex and its impact on machine intelligence.
Hawkins has a background that would make him an ideal faculty member at a place like Beckman where interdisciplinary research is done with an eye toward applications. He has an undergraduate degree in electrical engineering with post-graduate work in biophysics and an interest in pattern recognition for speech and text input. Hawkins co-founded PDA companies Palm Computing and Handspring, as well as his current company Numenta and the Redwood Center for Theoretical Neuroscience at the University of California, Berkeley.
Numenta is a computer technology company that has the goal of developing a biologically consistent model of the neocortex, called Hierarchical Temporal Memory (HTM), and applying it to practical problems. The company’s website states that its founders believe “biological principles will drive the next generation of intelligent computing.”
Hawkins also co-founded a research center for studying neocortical activity, the Redwood Neuroscience Institute in 2002, later gifting it to the Berkley campus in 2005 as the Redwood Center for Theoretical Neuroscience. The goal of the Institute is to “develop mathematical and computational models of the underlying neurobiological mechanisms involved in perception, cognition, learning, and motor function.”
Both of Hawkins’s ventures depend on understanding how the brain works when it comes to higher functions such as language and sensory perception. His talk at the SmithGroup Lecture will focus on his theories about brain function and on the ways he and other scientists are trying to meet the challenge of creating computing systems that can reproduce the kinds of capabilities found in the brain.
In the abstract for the talk, Hawkins writes that “Our belief is that to solve many problems of machine intelligence we first need to understand the principles by which the brain works and then build machines that work on those principles.” He adds that the new algorithms being created at Numenta “appear to be a leap forward in understanding what layers of neurons are doing in neocortex and we believe they will form a foundation on which many machine intelligence applications will be built.”
Hawkins’ theory of brain function, outlined in his book On Intelligence, is called the memory-prediction framework, which he describes as a “real intelligence’ approach to the study of intelligence to distinguish it from past approaches for creating truly functional artificial intelligence systems.
Hawkins writes in his prologue for the book that the “best way to solve this problem is to use the detailed biology of the brain as a constraint and as a guide, yet think about intelligence as a computational problem – a position somewhere between biology and computer science.”
The SmithGroup Lecture Series brings nationally-known scientists to campus for talks at Beckman. The SmithGroup architectural firm designed the Beckman Institute for Advanced Science and Technology building.
Hawkins will speak at 11 a.m. in the Beckman Institute auditorium.