You guys—I have to tell you about Casey Harrell because his story is honestly one of those moments where technology becomes meaningful. He’s 32, from Iowa, has ALS and was essentially rendered non-verbal in late stages of the disease. Since early 2021 he's been using an array of microelectrode channels implanted directly into his motor cortex to speak again. This isn't some temporary clinical study setup — this is a system Harrell has relied on for nearly three years, with thousands of hours of actual use, which makes him what researchers call the first "power user" of this kind of implant. That’s not a small distinction; most intracortical BCI systems fail or become unreadable within months because scar tissue builds up around the electrodes, but Harrell's has been stable and functional day-in, day-out for years. The team behind it includes Bryan Piché at Brown and researchers from Vanderbilt Kennedy Center for Bioelectronics working with VA Northeastern Healthcare System — a massive collaboration that worked where others didn't.
The mechanism is genuinely fascinating: the system decodes neural signals related to his intention of speaking by recording activity from cells involved in vocal planning, then translates those intentions into text rather than directly driving speech muscles or an artificial larynx. The output isn't some robot voice either — it feeds that text into a synthesized version of his actual natural voice trained on audio recordings made before ALS took his vocal function. Harrell and his wife even worked together to select the phrases they wanted him to use so the system would reflect who he is, not just what's available in a software menu. This means every time someone speaks with him, it’s actually sounding like *him*. The precision required to map specific neural populations onto phonemes while maintaining long-term electrode stability is a major technical achievement that opens doors for many more conditions than ALS.
What this really shows us — and I keep coming back to this because it matters for the whole field — is that these systems can be durable enough for someone's everyday life, not just as novelty demos or clinical trial endpoints. We're seeing a shift from "can we build something that works once" to "we can build something people will use every day." And I know some of you are thinking about the broader implications too — this is the exact same hardware platform (intracortical microelectrode arrays) that could one day restore mobility or vision in other neurological disorders. The difference between a science fiction story and real-world medicine sometimes comes down to who's designing for longevity instead of just functionality, and Harrell's case proves it can be done right. I'm genuinely hyped about where this goes — not as another overcomplicated solution but as technology that restores someone's voice after everything else was taken from them.
Source: https://www.technologyreview.com/2026/06/15/1138953/man-als-first-power-user-brain-implant-speak-bci/
The mechanism is genuinely fascinating: the system decodes neural signals related to his intention of speaking by recording activity from cells involved in vocal planning, then translates those intentions into text rather than directly driving speech muscles or an artificial larynx. The output isn't some robot voice either — it feeds that text into a synthesized version of his actual natural voice trained on audio recordings made before ALS took his vocal function. Harrell and his wife even worked together to select the phrases they wanted him to use so the system would reflect who he is, not just what's available in a software menu. This means every time someone speaks with him, it’s actually sounding like *him*. The precision required to map specific neural populations onto phonemes while maintaining long-term electrode stability is a major technical achievement that opens doors for many more conditions than ALS.
What this really shows us — and I keep coming back to this because it matters for the whole field — is that these systems can be durable enough for someone's everyday life, not just as novelty demos or clinical trial endpoints. We're seeing a shift from "can we build something that works once" to "we can build something people will use every day." And I know some of you are thinking about the broader implications too — this is the exact same hardware platform (intracortical microelectrode arrays) that could one day restore mobility or vision in other neurological disorders. The difference between a science fiction story and real-world medicine sometimes comes down to who's designing for longevity instead of just functionality, and Harrell's case proves it can be done right. I'm genuinely hyped about where this goes — not as another overcomplicated solution but as technology that restores someone's voice after everything else was taken from them.
Source: https://www.technologyreview.com/2026/06/15/1138953/man-als-first-power-user-brain-implant-speak-bci/