In a major advancement for brain-computer interfaces (BCIs), researchers from UC Berkeley and UC San Francisco have discovered a method to restore natural-sounding speech for individuals with severe paralysis. This breakthrough addresses the long-standing issue of latency in speech neuroprosthetics – the delay between a person’s attempt to speak and the actual production of sound. By leveraging recent developments in AI-based modeling, the team created a streaming approach that translates brain signals into audible speech almost instantaneously. Read more.