Surya Ganguli joined Microsoft CTO, Kevin Scott, for a technical conversation about artificial intelligence on Kevin's podcast, "Behind the Tech."
Surya is a Stanford University Assistant Professor in Applied Physics and, by courtesy: Neurobiology; Electrical Engineering and Computer Science. He’s also a faculty Associate Director at Stanford’s Institute for Human-Centered Artificial Intelligence.
Surya talked about his path to artificial intelligence (AI), which began with a desire to reverse engineer the human brain. But when he was a freshman, AI was done very differently, and mostly ignored the brain. So in college, he veered off to study both physics and mathematics, in addition to computer science. He discovered a love of quantum mechanics, which used the power of mathematics to yield penetrating insights into the microscopic world. This led him to a PhD in string theory (he describes this as “the one theory that can unify general relativity and quantum mechanics”).
After his PhD, Surya found his path back to AI and the brain, following his intense desire to connect with nature, especially to understand and recreate fascinating aspects of our inner mental worlds. He now works across the fields of neuroscience, psychology, AI and physics.
Kevin and Surya discussed a number of areas ripe for research, including energy efficient AI, catastrophic forgetting, unsupervised learning and the importance of building AI systems that learn with much less data — like humans.