A state-of-the-art brain-machine interface created by UC San Francisco neuroscientists can generate natural-sounding synthetic speech by using brain activity to control a virtual vocal tract — an ...