Science|Brain Implants Give A.L.S. Patients the Voice to Speak and Sing
https://www.nytimes.com/2026/11/14/science/brain-implants-als-voice.html
You person a preview presumption of this nonfiction portion we are checking your access. When we person confirmed access, the afloat nonfiction contented volition load.
In 1 demonstration, a erstwhile vocalist sang a philharmonic scale. Afterward she cried tears of joy.
Video
Nov. 14, 2026, 12:00 p.m. ET
Brain signals from a 46-year-old antheral who lost his dependable to A.L.S. were decoded by experimental encephalon implants and translated straight into intelligible speech. A 58-year-old woman, a trained singer, whose dependable was gradually silenced by the disease, was capable to sing a philharmonic scale.
The caller advances from researchers astatine the University of California, Davis, to beryllium presented adjacent week astatine the Society for Neuroscience league successful Washington, show that probe into brain-computer interfaces are crossing from proof-of-concept demonstrations into mundane usefulness.
“This system, for the archetypal time, is bully enough,” said Cindy Chestek, a biomedical technologist astatine the University of Michigan who was not progressive successful the caller work. “Good capable for existent communication. It’s bully capable to sing.”
For years, encephalon implants paired with progressively almighty A.I. person been shown successful singular lawsuit studies to reconstruct a person’s ability to communicate, inquire questions, adjacent marque jokes. The consciousness of transportation that the exertion provided to patients — who had mislaid power of the muscles that allowed them to speak, from A.L.S. oregon a changeable — could beryllium profound, but was besides halting, due to the fact that systems translated signals archetypal to text, past to speech.
A idiosyncratic would effort to speak, and electrodes surgically implanted successful the encephalon would observe electrical signals. Software would decode those signals into phonemes, the sonic gathering blocks of words. Participants would ticker words appear, 1 by one, connected a screen, gathering a sentence. They would past trigger a cursor click to awesome that they were done. Then, a computer-generated mentation of their dependable would speak.
Research subjects could convey their thoughts successful affluent detail, but connected a flimsy delay, which made it hard to genuinely articulation conversations. People often pass with the rhythm, timing and accent of their speech, not by composing afloat responses and mentally hitting “send.”

1 day ago
10





English (US) ·