A genuine brain-computer interface does not teach the brain to operate a machine; it teaches the machine to enter the nervous system’s native causal chain. Most of today’s astonishing brain-computer interface (BCI) demonstrations—whether moving a cursor by intention, controlling a robotic arm, or translating attempted speech into text—represent real and important progress. Yet by rigorous neuroscientific standards, they still remain some essential distance from a “true” brain-computer interface. In most cases, we have not truly understood how the brain forms intention, organizes action, or integrates sensation; nor have we genuinely inserted a machine into the closed neural pathway running from intention to planning to execution to feedback to correction. More often, we identify reproducible statistical correlations on the surface of brain activity and, through engineering, domesticate those correlations into usable control signals. It is a bridge, certainly—but more like a temporary pontoon bridge than a high-speed railway. For people living with paralysis, that floating bridge is already immensely valuable; yet to make it a system that truly resolves paralysis, we are still far from the goal.
Dongsheng Xiao (Thu,) studied this question.