Note: Eric Lacosse has transitioned from the institute (alumni).
The relationship between the Central Nervous System (CNS) and environment is extremely versatile. This versatility is what underlies all human learning. It is also what allows the CNS to gain so much from so little–so little energy, time, and data (sensory information and experience). This stands in stark contrast with today’s most advanced approaches of synthetic intelligence; enormous amounts of compute time, energy, and data are needed for the most rudimentary of effortless, commonsense tasks we take for granted.
One inspiring, general category of tasks–universally celebrated across cultures–is learning how to generate or adapt skilled movement, i.e., motor learning. How the biological feat of learning a new motor skill takes place–even in the simplest form conceivable--remains full of exciting, open questions. Helping answer them may also inspire us where we struggle to realize motor control synthetically–robots performing the simplest tasks still remain far behind human dexterity, reliability, and generalizability.
My research is at the interface between computational method development and the design of human motor learning experiments that can be brought into magnetic resonance (MR) environments for functional imaging.