Fifteen years ago, Kotaro Sakamoto, then a physics student captivated by the Kuramoto and Ising models, turned his attention to artificial intelligence, becoming one of the last apprentices of pioneering AI researcher Shun-ichi Amari. After completing a PhD at the University of Tsukuba on sleep neuroscience and machine learning, he conducted research on AutoML, neural architecture search, Bayesian optimisation, and information geometry at the Institute of Statistical Mathematics. Since 2024, he has been a researcher at Matsuo Lab, The University of Tokyo, where his work centres on the theory of generative and agentic AI—spanning diffusion and large language models—as well as world modelling, machine reasoning, multi-agent social simulation, and AI alignment. He grew up in New York and Hong Kong.
