Speech implant has let a man with ALS produce nearly 2 million words, his team reports in Nature Medicine
Casey Harrell, who has ALS, has communicated nearly 2 million words through a brain implant, his team reported in Nature Medicine. The record is one patient's, and the researchers Science News quotes expect others to follow.
The Scientist · Science desk
What happened
- Harrell uses the system for about 10 to 14 hours a day, according to Science News.
- Matthew Willsey of the University of Michigan recently implanted a new type of speech interface in a woman's brain.
- Harrell's interface is bulky, relatively slow and limited in the nonword sounds it can decode, according to the article.
- Science News credits advances in computing power and AI, lessons from patients and neurotechnology company investment for the progress.
Why it matters
- What's now possible capability In Willsey's account, the open question for the field has moved from whether speech can be decoded to whom the decoding works for.
- What it limits constraint Science News extends the possibility to stroke, paralysis, cerebral palsy and Parkinson's, but the case it documents is ALS alone, so the others are inference.
- Who has to decide decision Hochberg's stated goal of fast, clear speech in a person's former voice sets targets a word count does not measure, so teams would need to report speed and voice quality to be compared.
The denominator is small. ALS affects an estimated 35,000 people in the United States [6], and the evidence in this account comes from one of them, Casey Harrell [1].
Harrell calls himself "the world's biggest power user of the brain-to-computer interface known to man," and Science News writes that hard data suggest he is right [12]. That count of nearly 2 million words is a measure of use [5]. It shows a person keeps reaching for the system, which a short lab demonstration cannot show. It is not an accuracy figure, and the article reports no decoding speed or error rate.
Science News says the implant decodes Harrell's words "directly from his thoughts" [2]. The electrode placement is more specific. Four small devices were implanted on the left side of his brain in 2023, in a clinical trial at the University of California, Davis, and their grids sit on the speech motor cortex, a stretch of the precentral gyrus that controls the movements required for speech [3]. The placement points toward motor commands for speaking. A decoder built on motor commands is a different proposition from one that reads thought in general.
Matthew Willsey, the Michigan neurosurgeon, said: "There's a lot of excitement and hope in the field right now." [10] For other people and other diseases, that excitement is for now an expectation. Leigh Hochberg, a neuroscientist at Brown University who works with Harrell, put it more plainly: "It's still early days." [14]
What to watch
- Published results from Willsey's new interface in the woman he implanted, including how fast and how accurately it decodes.
- Whether trials enroll people with stroke, cerebral palsy or Parkinson's, the conditions Science News says the approach may reach.
- Whether later reports pair word counts with decoding speed and error rates, so systems can be compared.
Clarity's read
What the record supports and how the coverage leans. The claims behind it follow.
Reality
- Evidence55
- Adoption8
- Hype gap+15
- Incentives40
- Confidence50
Claim ledger
Ranked by verification strength, evidence, and original report placement.
- [1]
Casey Harrell has amyotrophic lateral sclerosis (ALS) and began to lose his ability to speak after his diagnosis in 2020.
- [2]
Science News says Harrell talks using an experimental brain implant, called a brain-computer interface, that can decode his words directly from his thoughts.
- [3]
In 2023, in a clinical trial at the University of California, Davis, Harrell had four small devices implanted on the left side of his brain; the electrode grids reside on his speech motor cortex, a stretch of the precentral gyrus that controls the movements required for speech.
- [4]
Harrell uses his system for about 10 to 14 hours daily.
- [5]
A recent count shows Harrell has used his system to communicate nearly 2 million words, the scientists working with him reported in June in Nature Medicine.
- [6]
ALS affects an estimated 35,000 people in the United States.
- [7]
Harrell's interface is bulky, relatively slow and limited in what sorts of nonword sounds it can decode.
- [8]
Matthew Willsey, a neurosurgeon and biomedical engineer at the University of Michigan in Ann Arbor, recently implanted a new type of speech interface in a woman's brain.
- [9]
Willsey says that for years scientists did not know whether attempts to decode speech would be successful, and that a body of research results now points clearly to yes.
- [10]
"There's a lot of excitement and hope in the field right now," said Matthew Willsey.
- [11]
Science News says advances in computing power and AI, lessons gleaned from patient experience and investments from neurotechnology companies are starting to yield real progress on brain-computer interfaces for people who can no longer talk.
- [12]
Harrell describes himself as "the world's biggest power user of the brain-to-computer interface known to man," and Science News says hard data suggest he is right.
- [13]
Science News writes that Harrell's words show it may be possible to restore speech for people with ALS, as well as those whose speech is impaired by stroke, paralysis, cerebral palsy and Parkinson's disease.
- [14]
"It's still early days," said Leigh Hochberg, a neuroscientist at Brown University who works with Harrell.
- [15]
Hochberg says the end game is a world in which people could use a device to communicate seamlessly, quickly, clearly and in their former voices.
Get this ledger inside ChatGPT. Ask about any of our stories and the answer brings its figure, each claim’s verdict and the sources. Free, no account.
Sources
1 independent publisher whose own reporting we read for this story.
- sciencenews.orgBrain implants are helping people who’ve lost their voice
1 article · October 15, 2026
Topics and entities
Follow any of these and your For You feed starts watching them — no settings page required.
Topics
- Brain-Computer InterfacesFollow
- Speech neuroprostheticsFollow
- Amyotrophic lateral sclerosisFollow
Entities
- Casey HarrellFollow
- University of California, DavisFollow
- Nature MedicineFollow
- Matthew WillseyFollow
- Leigh HochbergFollow
- David BrandmanFollow
- Science NewsFollow