Eshan Chattopadhyay joined the faculty of Cornell’s Computer Science Department in July 2018. Chattopadhyay received his undergraduate Bachelor of Technology degree in Computer Science from the Indian Institute of Technology, Kanpur. He earned his Ph.D. in Computer Science from the University of Texas at Austin. Before coming to Cornell, Chattopadhyay spent two years as a postdoctoral researcher at the Institute for Advanced Study (IAS) in Princeton, New Jersey.
“My time at IAS was amazing,” says Chattopadhyay during a recent conversation in his Gates Hall office. “I was able to explore my own ideas in the company of some very high profile people. I was surrounded by winners of the Fields Medal…it was super humbling to be the stupidest guy on campus.” Chattopadhyay had the opportunity to work with Avi Wigderson at IAS and found him to be very generous with his time and his ideas. “Avi would talk with me for hours. And he wasn’t the only one—I really didn’t feel the hierarchy in my field at all.”
Chattopadhyay’s field is theoretical computer science. More specifically, complexity theory, pseudorandomness, and cryptography. In his years at UT-Austin Chattopadhyay, working with his advisor David Zuckerman, solved a problem that had been vexing the field for more than 25 years. Together, Chattopadhyay and Zuckerman found an efficient way to produce purely random bits from two independent weak sources of randomness.
The need for truly random bits is great in many applications of computer science. Most encryption relies on random numbers. Traditional sources of randomness have been shown to be less random than originally thought. For instance, a computer might use the timing of the intervals between a user’s keystrokes to create a set of random numbers. Over time, patterns emerge that disqualify these sets of numbers from being truly random.
Chattopadhyay and Zuckerman found a way to combine two low-quality sources of randomness (as long as they are unrelated sets) to create one set of high-quality randomness that can be used for safer encryption, better simulations of complex phenomena like weather and the economy, and more truly random sample groups for pollsters.
At Cornell, Chattopadhyay plans to continue and deepen his work on pseudorandomness, complexity theory, and cryptography. “This is such a fertile area to be working in,” says Chattopadhyay. “Many people believe that if we can solve a problem using randomness then we should eventually be able to solve it without randomness. But this is still a huge open question. Complexity theory tries to understand which problems we can solve efficiently and which we can not solve efficiently.”
Chattopadhyay joined the faculty at Cornell in part because when he came to visit he found people who were friendly, humble, and eager to answer questions and give helpful advice. “There are so many gifted people here doing such strong research,” says Chattopadhyay. “It wasn’t a very hard decision.” He will be teaching a graduate level class in Pseudorandomness and Combinatorial Constructions in the fall of 2018 as he recruits graduate students to join him in his work. “Cornell should be a great place for me to find students,” says Chattopadhyay. “The work has appeal to people from several fields: applied mathematics, optimization, and electrical engineering to name a few.”
--by Chris Dawson