A Power User Emerges
Casey Harrell, a man living with amyotrophic lateral sclerosis (ALS), has become the first 'power user' of a speech brain-computer interface (BCI), according to a study published today in Nature Medicine. Over 22.6 months following implantation, Harrell accumulated more than 3,800 hours of independent home use, with the device achieving 99% accuracy in decoding his intended words.
The milestone represents a significant leap for BCI technology, shifting from controlled lab demonstrations to real-world, daily utility. Harrell, now 48, uses the implant to speak, send messages, browse the web, and continue working as an environmental activist — all without researchers present.
“Living with a disease like ALS, you are supposed to have diminished dreams. I do not,” Harrell told MIT Technology Review. “Any one of these things would be an absolute godsend of improvement. To have all of them, and many, many more, is truly revolutionary.”
How the Implant Works
In July 2023, during a five-hour surgery, surgeons at the University of California, Davis, implanted four arrays of 64 electrodes each into Harrell’s brain. The electrodes were placed in his speech motor cortex — the region controlling the muscles required for speaking. Two pairs of arrays were connected to pedestals on the exterior of his skull, allowing a computer to be plugged in.
The team, led by neuroengineers Sergey Stavisky and Nicholas Card, developed algorithms that decode neural activity into phonemes — the 39 sound units of American English. “We first go from brain data to phonemes, and then from phonemes to words,” Card explained. Within a month of surgery, the decoder was operational, achieving 99.6% accuracy with a 50-word vocabulary on the first day. That vocabulary was later expanded to 125,000 words, maintaining 97.5% accuracy before subsequent updates pushed it to 99%.
Harrell can now control a cursor, enabling him to use personal computers independently. He also has a “privacy mode” that automatically deletes decoded text, and a “profanity filter” for conversations with his young daughter.
From Lab to Living Room
A critical advancement highlighted in the study is the transition to independent home use. In 2023, researchers had to visit Harrell to connect and disconnect the system. Now, a care partner — trained by the team — can handle the setup. Harrell wakes up, gets plugged in, and proceeds with his day. This automation is crucial for real-world adoption, according to Mariska Vansteensel, a BCI researcher at Utrecht Medical Center not involved in the trial.
“For these technologies to be relevant for patients, we really need to test them in settings in which they will eventually be used,” Vansteensel said. “This study demonstrates that it has value, that it’s usable, and that it functions well without the constant involvement of a research team.”
Harrell’s usage includes sending text messages, emails, and performing his job. He also reads to his seven-year-old daughter — something impossible before the implant. “This has allowed me to keep working and earn money and insurance for my family. This is reconnecting me with friends and family,” he said.
Limitations and Open Questions
Despite the success, the study does not guarantee the approach works equally well for all ALS patients. Vansteensel noted that in her previous work, a woman with ALS used a fully implanted BCI for seven years before it stopped working, likely due to brain degeneration. The long-term stability of electrode recordings remains a challenge: scar tissue can form and degrade signal quality.
Moreover, not all patients are willing to undergo invasive brain surgery. Jane Huggins, a BCI researcher at the University of Michigan, pointed to a “consistent aversion to hospital stays among people with progressive conditions like ALS.” Noninvasive alternatives remain under development, but none have matched the performance of implantable systems.
The UC Davis team acknowledges these hurdles and is already working on further improvements. “We’re never satisfied,” Stavisky said. Future efforts will focus on making the system fully implantable — eliminating external connectors — and testing it in additional patients to establish generalizability.
Implications for the BCI Field
Harrell’s case provides the strongest evidence yet that speech BCIs can deliver sustained, clinically meaningful benefits outside research settings. The 3,800-hour usage record far exceeds previous demonstrations, and the 99% accuracy approaches the reliability needed for natural conversation.
For the AI and neurotechnology community, the study underscores the importance of closed-loop, user-driven design. Many of the features — privacy mode, profanity filter, cursor control — were added in response to Harrell’s specific requests. That iterative, patient-centered approach may prove as important as the hardware itself.
The findings also reinforce that brain-computer interfaces are not just about restoring function but enabling new forms of independence and participation. As Harrell put it, “We are making the road as we walk it, or roll it, so to speak.” The next step will be to see whether that road can be paved for others facing similar diagnoses.
评论