Terrence Sejnowski

While growing up in Cleveland, Ohio, Terry Sejnowski developed a deep interest in physics. “It seemed to me that of all the sciences, physics had made the most progress in understanding the universe,” he says. “It had a long history of successfully unravelling many mysteries, and I found that very attractive.” Sejnowski discovered he had a natural aptitude for mathematics and physics, taking graduate courses in both subjects while still an undergraduate at Case Western Reserve University. In 1968, he enrolled at Princeton University, where his PhD advisor was the theoretical physicist John Hopfield. While at Princeton, Sejnowski took courses in biology and studied the neuronal activity of macaque monkeys in the laboratory of Charles Gross, one of the founders of cognitive neuroscience. He also spent a summer immersed in a neurobiology course at the Marine Biology Laboratory at Woods Hole, Massachusetts. By the time he left Princeton in 1979—and with the encouragement of Hopfield—Sejnowski had made the transition from theoretical physics to theoretical and computational neurobiology. “I found the mysteries of the brain to be just as exciting as the mysteries of the universe,” he says. “And you could do it in your own lab.”

After two postdoctoral fellowships, first with Princeton biologist Alan Gelperin and then with Harvard neurobiologist Stephen Kuffler, in 1981 Sejnowski joined the faculty in the Thomas C. Jenkins Department of Biophysics at the Johns Hopkins University. A few years later, in 1988, he moved to California, with joint appointments at the University of California, San Diego and the Salk Institute for Biological Studies. Throughout much of the 1990s, Sejnowski was also affiliated with the California Institute of Technology, where he was a Fairchild Distinguished Scholar and a visiting professor.

From the start of his career, Sejnowski has been a pioneer in the field of computational neuroscience, helping to shape the related areas of neuroeconomics, neuroanatomy, neurophysiology, psychology, and artificial intelligence. In 1985, while at Johns Hopkins, he collaborated with computer scientist Geoffrey Hinton to invent the Boltzmann machine, the first algorithm to solve the problem of learning in multilayered neural networks. It remains the most biologically plausible of all subsequent learning algorithms for artificial neural networks. Soon thereafter, Sejnowski created NETtalk, a neural network that, like the human brain, was able to learn how to turn written text into speech. Not only was this an astounding engineering accomplishment for artificial intelligence, but it also marked a major cultural milestone, for it raised new challenges for philosophy, linguistics, and cognitive science that continue to unfold today.

Many other landmark findings followed. In 1995, Sejnowski and postdoctoral fellow Tony Bell introduced a novel self-organizing learning algorithm—independent component analysis (ICA)—which solved a major problem (blind source separation) in signal processing. Sejnowski was also the first to apply ICA to electroencephalogram (EEG) and functional magnetic resonance imaging (fMRI) recordings. ICA is now a widely used technique for analyzing brain imaging. It has also had an enormous impact on engineering, solving problems in ways similar to the human brain. A year later, in 1996, Sejnowski published another seminal paper in which he proposed that the firing of dopamine neurons in the brain compute reward prediction error (the difference between the reward the brain expects and the reward it actually receives). This theory was soon experimentally confirmed and is now the basis for neuroeconomics.

Some of Sejnowski’s more recent work has focused on sleep spindles—the brain wave patterns that occur during non-rapid eye movement (NREM) sleep and that are believed to mediate many sleep-related functions, including the memory consolidation necessary for learning. He has shown that sleep spindles are not synchronous across the cortex, as was previously believed, but instead create circular traveling waves. “We’re now finding that these traveling waves are showing up everywhere in the brain and not just during sleep—and that they have an impact on behavior,” says Sejnowski. This research is helping scientists deepen their understanding of how the brain learns and of how memory is affected in disorders such as traumatic brain injury.

Sejnowski continues to work at the Salk Institute, where he is Francis Crick Chair and directs the Computational Neurobiology Laboratory, and he is also a Distinguished Professor in the Department of Neurobiology, School of Biological Sciences at the University of California, San Diego, where he directs the Institute for Neural Computation. He has received numerous awards and honors over the years, including memberships in the National Academy of Sciences, the National Academy of Medicine, the National Academy of Engineering, the National Academy of Inventors, and the American Academy of Arts and Sciences. He also helped with the 2013 launch of the National Institutes of Health (NIH) BRAIN Initiative, serving on the project’s advisory committee. Sejnowski lives in San Diego with his wife, Beatrice Golomb, MD, PhD, who is a professor of medicine at the University of California, San Diego.