Neuralink's N1 Brain-Computer Interface Enables ALS Patient to Speak Through Thought Alone

March 25 News: The brain-computer interface chip from Neuralink, a company under Elon Musk, has already helped people with disabilities remotely control mouse cursors or robotic arms. Now, the technology has achieved a new breakthrough—successfully converting brain signals into audible speech.

On Tuesday local time, the company announced the results of the relevant trial, featuring Kenneth Shock, a user who underwent brain-computer interface chip implantation surgery in January this year. Shock suffers from amyotrophic lateral sclerosis (ALS), a neurodegenerative disease that gradually deprives patients of their ability to walk and speak, causing severe speech impairments.

However, in a video, Neuralink demonstrated that its N1 brain-computer interface chip could help Choke communicate more easily, even without moving his mouth.The implantable device can decode his brain signals into text, which is then read aloud by a computer program.

Neuralink Briefly Explains How the Technology Works. The chip does not directly translate inner monologue as in telepathy; instead, it reads specific brain signals and matches them with the words the user intends to express. In introducing this clinical trial, the company stated, “Specific areas of the brain are activated and generate signals that would normally be transmitted to the muscles of the mouth, tongue, and throat.”

The video also shows that the implanted device, paired with Neuralink software, can recognize Noland’s brain signals and match them to “phonemes”—the smallest units of speech sounds.

When building the speech decoding system, Skylar Grana tier, a machine learning engineer at Neuralink, explained, “We guide him to try speaking certain sentences and use this data to map neural intentions to actual words.” In the first phase, Shock reads sample sentences aloud; in the second phase, he silently mouths the words without producing any sound.

Granatir added, “Our goal is for him to simply intend to move his mouth, and our brain-computer interface (BCI) will decode what he wants to say.”

In the Phase III trials, video footage showed that the Neuralink software successfully recognized Choke’s speech despite his complete lack of mouth movements. The implanted device mapped neural signals to individual phonemes, combined them into words, and ultimately formed complete sentences, which were then synthesized and spoken aloud in Choke’s original voice by the program.

With the aid of the chip, he said, “I am speaking to you with my thoughts.”

According to IT Home, this test is part of Neuralink’s ongoing “VOICE clinical trial,” so it may still take several years for the technology to become widely available. The video also shows that the current decoding process still needs optimization; for instance, there is still a certain latency in the software’s reading of brain signals and their conversion into speech. Nevertheless, Neuralink plans to significantly upgrade this technology over the long term.

Granatir stated, “We will continue to enhance the quality and quantity of our sensors. We aim to develop a system that can directly convert brain activity into speech in real time.”