How Neuralink Patients Are Actually Living Right Now — Beyond the Headlines
Forget the viral clips and Elon tweets: the real story of life with a brain chip is messier, more mundane, and far more moving than any press release will tell you.
When Noland Arbaugh revealed himself as the first human to receive a Neuralink implant, the clip that went viral was him playing chess with his thoughts. That one image — a paralyzed man moving pieces on a board purely through brain signals — pretty much consumed the media cycle for a week. What it didn’t show was what happened three months later, when a significant portion of the device’s ultrafine electrode threads quietly retracted from his brain tissue, degrading his control and leaving the engineers scrambling. Arbaugh chose not to say anything publicly at the time. He thought it would be “extremely rash” to go public before Neuralink had a chance to fix it.
He was right to wait. They fixed it, mostly — by rewriting the signal detection algorithm to become more sensitive to the remaining electrodes. But the episode tells you something the highlight reel doesn’t: this technology is genuinely experimental, living in the gap between medical device and work-in-progress, and the people using it are navigating that gap in real time, every single day.
By early 2026, Neuralink has implanted roughly 21 participants, referred to internally as “Neuralnauts,” across its PRIME Study and newer VOICE trial, covering people with cervical spinal cord injuries and ALS. The trials now run in the United States, Canada, the United Kingdom, and the UAE. We know the names of a handful of those participants, and their stories are worth telling properly — not as miracles, not as cautionary tales, but as lives being rebuilt with one of the most unusual tools in medical history. 💡
What Noland Arbaugh’s days actually look like
Arbaugh is 31, quadriplegic since 2016 when a swimming accident dislocated two of his vertebrae, and he now uses his Neuralink implant for roughly 10 hours a day. That number is worth sitting with. This isn’t a lab device he activates for a two-hour research session. He uses it to browse the internet, read books, write emails, play video games, and, as of fall 2025, take a full semester load of chemistry, biology, and pre-calculus courses. He’s earning top grades, which he credits directly to the chip.
Before the implant, by his own account, he couldn’t be left alone for 30 minutes. The contrast since then has been so stark that he’s begun building a professional speaking career. In September 2025, he was a paid speaker at Fortune’s Brainstorm Tech conference in Park City, Utah. He has described that sentence — “I was a paid speaker at a conference” — as something that seemed literally impossible two years ago. 🎤
But the daily reality has friction in it that the conference appearances don’t capture. A few specifics that Fortune reported after spending time with him 18 months post-surgery:
The N1 implant runs on a battery that needs charging approximately every five hours. Neuralink heat-treats the charging coil into some of his hats, so he charges by literally wearing a hat.
Recalibrating the “model” — the software that maps his imagined physical movements to cursor movements — degrades over hours and days. During a November 2024 livestream, he mentioned spending up to 45 minutes redoing a calibration task before getting usable control.
His cursor speed in grid-based tasks has measured around 8 to 9 bits per second, compared to the approximately 10 bits per second that the median able-bodied mouse user achieves. That gap sounds small, but it matters for usability.
He has named his implant “Eve,” as Wikipedia documents in his public updates, which says something about how personal this relationship has become.
The thread retraction setback was real, and it shaped how Neuralink approached subsequent implants. The company modified algorithms, but also learned from the engineering failure in ways that directly affected patient two. 🔧
Alex Conley is flying drones. With his brain.
Alex Conley, who received the second implant in July 2024 after a 2021 car rollover left him paralyzed from the waist down, has become one of the more remarkable data points in BCI history. In late 2025, he confirmed publicly that he is using the Neuralink device to control a robotic arm and to pilot a drone. With his thoughts. I’ll admit I had to re-read that sentence the first time I encountered it.
Conley’s background is in welding and fabrication — he worked in mechanics before the accident. The BCI has pulled him back toward that world in a way that a purely digital device couldn’t have. Using Neuralink-enabled CAD software, he’s returned to designing tools and parts on a computer. He and the Neuralink engineering team collaborated on a custom mount for his charging system; he designed the part, the team sent it to a printer in Phoenix, and he had it the next day. That workflow — paralyzed man designs physical hardware he cannot touch, engineers manufacture it overnight — is not something you can describe without feeling the weight of it. 🦾
Conley’s experience also avoided the thread retraction issue that plagued Arbaugh’s first months. That suggests Neuralink incorporated learnings between the first and second implants quickly, which matters for what the PRIME Study is actually producing: real iterative data on how to make the next implant better.
What Conley can do now versus a typical BCI recipient from even five years ago:
Drone piloting via neural signals, translating imagined movement into real-time directional control
Robotic arm operation for daily physical tasks, allowing interaction with a world that otherwise requires hands
CAD design on a computer using cursor control via thought, enabling professional work he thought was permanently behind him
Posting to social media independently, which may sound trivial until you consider he couldn’t operate a phone without assistance before
Bradford Smith found his voice, literally
The third Neuralink patient — Bradford Smith, who has non-verbal ALS — represents the application that may matter most in the long run. ALS is a disease that leaves the mind intact while systematically dismantling the body’s ability to communicate it to the world. Smith can no longer speak. His motor neurons are dying. But he can still intend to speak, and the Neuralink device reads those intentions.
What makes his case different from Arbaugh’s or Conley’s is the AI layer on top of it. Smith’s speech output uses an AI model trained on recordings of his voice from before he lost the ability to speak. When he types using brain-controlled cursor movement and a virtual keyboard, the system synthesizes his words in his own voice, not a generic text-to-speech voice. As Neuralink co-founder DJ Seo described it when discussing the moment Smith played Mario Kart with his children again: “That moment was incredible.” 🧬
Smith’s daily use is different in character from Arbaugh’s. He’s not playing video games for fun — though he does that too. He’s communicating with his family, attending his kids’ soccer games, speaking at his local church, and planning travel for the first time in five years. The Deseret News profiled him in May 2025, and the portrait is of a man who says “ALS still really sucks” while simultaneously describing the implant as a form of answered prayer. That combination of clear-eyed acknowledgment of the disease and genuine gratitude for what the device gives back is probably the most honest account of what these trials are producing for the people in them.
The VOICE trial, Neuralink’s follow-up program explicitly designed for speech restoration, enrolled its second participant — Kenneth Shock, another ALS patient — in January 2026. Musk confirmed in late March 2026 that the VOICE trial is producing results. 🗣️
The real-world friction nobody talks about
For all the remarkable capabilities, the honest accounting of Neuralink’s current state requires naming the things that don’t work yet or haven’t been solved. The MIT Technology Review analysis from January 2025 is still largely accurate in its framing: these are true experiments, not products.
Some unresolved realities as of early 2026:
Thread retraction has happened and will probably happen again. The company has partial software mitigations, and the surgical robot is being refined to improve thread placement, but it’s not solved. Neuralink is targeting a “streamlined, almost entirely automated surgical procedure in 2026” that threads through the intact dura without removing it, which should reduce complications.
Battery life at roughly five hours creates a rhythm of the day that able-bodied people never think about. You charge a phone when convenient. Charging a brain implant is more negotiated.
Model drift means cursor control isn’t set-and-forget. The neural patterns the software reads shift over time, and recalibration is part of the regular routine. That’s not a dealbreaker, but it’s real work.
Data security for a device that transmits neural signals wirelessly is a topic that gets almost no public discussion. Arbaugh’s X account was hacked and a false tip to local law enforcement sent a SWAT team to his home. These are separate events, but the collision of celebrity, vulnerability, and neural data creates risks that nobody has fully mapped yet.
What Neuralink is building in parallel is a surgical robot — already in use for implantation — that the company wants to eventually run with minimal human surgeon involvement. The goal is a procedure that looks less like brain surgery and more like an outpatient visit. Whether that’s achievable within the timelines Musk has publicly described is a genuinely open question. 💊
What the PRIME Study is actually producing
Beyond the individual stories, the aggregate data from PRIME is legitimately interesting. By February 2025, trial participants had collectively logged over 4,900 hours of device usage across more than 670 implant-days. That’s not cherry-picked demo footage — it’s sustained, daily, practical use by people for whom the alternative is profound isolation from the digital world.
The participants are scattered across four countries now. The most recent trial enrollment numbers put the PRIME Study at 21 participants, with the company aiming to expand toward larger enrollment as it pursues FDA Premarket Approval. The Blindsight implant — Neuralink’s visual cortex device — received an FDA Breakthrough Device Designation in June 2025, which doesn’t mean approval but does mean accelerated review.
What I find genuinely hard to resolve is the gap between those 21 participants and the scale of need. According to the Christopher and Dana Reeve Foundation, approximately 5.4 million Americans live with paralysis from various causes. ALS affects roughly 30,000 Americans at any given time. The PRIME Study is not a treatment rollout — it’s a controlled experiment to determine whether this can become one. The distance between 21 participants and clinical availability is measured in years and regulatory requirements that won’t bend for hype. 📈
The people in those trials know this. Arbaugh knew when he signed the consent forms that something could go terribly wrong. He said so publicly. “Even if it didn’t work — even if something went terribly wrong — I knew that it would help someone down the road.” That’s not the language of a true believer; it’s the language of someone who thought carefully about risk and decided the bet was worth making. The question worth sitting with isn’t whether the technology is impressive — it clearly is. It’s whether the systems being built around it, from surgical robots to signal security to regulatory frameworks, are advancing at the pace the technology demands.
So what should actually change in how this technology gets covered and discussed — and are there parts of the real patient experience that you think the industry is still not being honest about?