Our sophisticated use of language is a key part of what makes us human. As far as we know, no other animal is capable of, say, using an apples-and-oranges analogy to help explain the difference between two things or of summarizing the key points in a Ph.D. dissertation.
New Program Postdoctoral Fellow Arseny Moskvichev is fascinated by how people use language and abstraction to communicate and share knowledge. Moskvichev, who received his Ph.D. from UC-Irvine, where he worked closely with advisor Mark Steyvers, is particularly interested in using insights from cognitive science and machine learning to advance language capabilities in artificial intelligence systems. His dream, he says, is to see Natural Language Processing AI models become capable of having meaningful conversations and even change their beliefs.
At SFI, Moskvichev will work with Professor Melanie Mitchell to solve the problem of how to measure abstraction and the analogy-making capabilities of AI systems.
Moskvichev holds a B.Sc. in psychology and an M.Sc. in neuroscience from St. Petersburg University, as well as a Ph.D. in cognitive studies and an M.Sc. in statistics from UC-Irvine. Arriving summer 2022, supported by NSF EAGER Award 2139983