The Neuroscience of Skill: Lessons from a Decade of Musculoskeletal AI
Alexander Mathis, Associate Professor at the Brain Mind Institute, EPFL
Motor control, the art of coordinating muscles to produce intricate movements, is a marvel of biological intelligence: skills like playing soccer take years of practice to master. Over the past decade, AI competitions have produced musculoskeletal agents that solve dexterous manipulation tasks, play table tennis and kick soccer penalties, controlling up to 290 muscles. Alexander Mathis will link these results to classic theories of skill learning from psychology, asking what they reveal about how we learn and how we control our bodies. He will argue that we are ready to tackle skill learning computationally, at the scale of the whole body.
This talk is part of the Kempner Seminar Series, a research-level seminar series, open to the public, that covers topics related to the basis of intelligence in natural and artificial systems.
