In January 2024, surgeons implanted a chip into the brain of a man named Noland Arbaugh, paralyzed from the shoulders down after a diving accident. Weeks later he was moving a cursor and playing online chess by thought alone, for the first time in years not depending on someone else's hands. This is not a demo reel — it happened, and it genuinely helped one specific person. What follows is the harder question this piece takes on: the chip implanted to give a paralyzed man back a cursor and the chip implanted to "achieve symbiosis with AI" so a healthy person doesn't fall behind the machines are, physically, the same product from the same company. The border between medicine and market doesn't run through the electrodes. It runs through intent — and intent shifts faster than you can blink.
What already works — and it genuinely works
Neuralink is neither the only player nor the most senior one. The Utah Array, from Blackrock Neurotech, has sat in research labs since the nineties, and researchers have spent decades on it learning to read movement intent straight out of the motor cortex. Synchron took a different route entirely: its Stentrode is threaded in through a blood vessel, no skull opened, and it already lets people with ALS type by thought. Against that backdrop, Neuralink's contribution isn't a discovery of principle — it's engineering of execution. Its R1 surgical robot stitches in ultra-thin threads carrying on the order of a thousand-plus electrode channels, by the company's own account, with a precision no human hand can hold, threading around the cortex's blood vessels.
The PRIME study, in which Arbaugh and several other patients are enrolled, runs under the oversight of the FDA, which granted the device "breakthrough" status — not full approval, but an accelerated track for a promising, not-yet-proven technology. The stated goal is narrow and genuinely worthy: let people with quadriplegia and ALS control a computer, a cursor, a robotic arm directly — no keyboard, no voice, no dependence on another person's hands. Inside that frame, there's nothing to argue with. This is medicine, and it's needed.
Where the frame ends
The problem isn't what Neuralink does today. It's what the company itself says about tomorrow. Its founder has publicly framed the goal not as "treatment" but as "symbiosis with artificial intelligence" — a way to avoid becoming economically and cognitively irrelevant next to machines that will soon think faster and wider than a human can. That's no longer a medical indication. It's a bet on a future where a neural interface isn't a crutch for the sick but an upgrade for the healthy — a competitive edge in a world where the person at the next desk types by thought while you're still hitting keys.
A company that has raised, by various estimates, several hundred million dollars at a valuation in the billions cannot live forever on a "quadriplegia" indication — that patient population is real, but it will never justify the capitalization investors are talking about. The economic logic pushes the product the same direction every venture-funded medical startup gets pushed: from a narrow therapeutic use toward the widest possible market. Today: give a paralyzed person back the ability to write a letter. Tomorrow, with the exact same hardware in the skull: sell a healthy person "enhanced memory" or "direct hands-free interface access." The line between "treat" and "sell an upgrade" isn't technical. It's commercial, and it gets drawn not by a surgeon but by a board of directors.
Who owns the signal from inside your head
A neural interface doesn't just send commands outward — it reads raw electrical patterns from the brain in order to decode them. That's denser data than has ever been collected about a person: not what you clicked, but what was happening in your cortex the moment you thought about clicking it. Who stores those recordings, for how long, who they can be handed to by court order, by a service subscription, by a merger with another company — in most jurisdictions these questions still have no clear answer, because the law was written for data about behavior, not for a raw stream from under the skull.
Our record: in the vocabulary of Maat, the Ba is mobile consciousness — the part of you that moves, acts, chooses. A neural interface plugs literally into the Ba's signal and turns it into a stream that can be recorded, sold, run through someone else's algorithm. And the Ren — your name, your essence, the thing that naming grants power over — stops being a metaphor here: a decoded intent pattern is about as precise a "name" for your inner state as anything could be. A company holding the decryption key holds not a metaphorical but a literal claim on your Ren. That's not a reason to reject a technology that genuinely gives people back the ability to move and speak. It's a reason to ask out loud, before signing the consent form, exactly who you're handing that key to.
What happens when the company doesn't survive
This isn't a hypothetical worry. It has a precedent. Second Sight made the Argus II retinal implant, which partially restored sight to blind patients — a real, working, regulator-approved technology. When the company ran out of money, support for the device stopped: no more firmware updates, nowhere to replace a failed external unit, and the implants themselves stayed wired into the eyes of hundreds of patients for decades to come — working or not, a question now decided not by medicine but by the balance sheet of a company that no longer existed in any functional sense. A brain implant isn't an app you can delete with one tap. It's hardware stitched into living tissue, whose future maintenance depends on whether a private company stays solvent for the next forty or fifty years — a horizon most venture-backed startups survive only on the pitch deck.
That's the bottleneck that usually gets lost behind the footage of a man playing chess by thought. The medical benefit is real. The dependency it creates is just as real, and it isn't symmetric: the patient implants the company's hardware once; the company gets to decide the hardware's fate — expand, go bankrupt, pivot the business model — fresh every quarter.
A technology that gives a paralyzed person back the ability to write to a friend without anyone's help deserves to exist and keep improving — that isn't up for debate. But precisely because the stakes are this personal and this irreversible — a signal from under your own skull, hardware wired in for decades — the question of who holds the key, and on what terms, doesn't get to wait until the line between treatment and product has already been erased by default. The mind is the last territory that still, unconditionally, belongs only to you. It's worth reading the contract before that territory acquires a landlord.