Neuralink full send... Elon's brain chips actually work on humans

Neuralink’s first human implant is being reported as operational, with a paralyzed patient demonstrating “telekinetic” control of a computer cursor using brain signals alone. The breakthrough matters because it turns earlier animal testing concerns into real-world evidence that a brain-computer interface can translate neural activity into actionable commands—without eye tracking or external sensors.

The account centers on Nolan Arbau, a 29-year-old paralyzed from the shoulder down after a diving accident. Over the weeks following implantation, he reportedly gained the ability to control a mouse cursor to play games and interact with software such as chess and Civ 6 using only his thoughts. The system’s core claim is that the implanted device detects brain activity patterns and maps them to intended movements, enabling cursor control rather than relying on residual muscle control.

Technically, the implant described as the “N1” system uses a surgical robot to place ultra-fine electrode threads into the brain. After drilling a hole in the skull, the robot implants multiple threads that carry high-density electrodes—finer than a human hair—designed to pick up electrical activity from neural tissue. The transcript specifies that the N1 has 1,24 electrodes distributed across 64 threads, yielding 16 electrodes per thread, and compares this to typical electrocorticography setups that often use around 20 sensors and max out near 256 electrodes. A key design detail is that the electrodes sit on the outside of the skull, which the transcript frames as increasing the chance of interference before signals reach the electrodes.

Data transmission and power are also described as self-contained: the N1 transmits wirelessly and is powered by a small lithium-ion battery. That battery is charged wirelessly from outside the skull using an inductive charger.

The mechanism for turning brain activity into commands is tied to brain-wave frequencies. The transcript outlines delta waves (~2 Hz) during sleep, alpha waves (~10 Hz) when alert, and gamma waves (~35 Hz) during high focus—especially relevant for pattern recognition. The implant is said to detect specific electrical signal patterns associated with intended actions (for example, “move my right arm”), but it requires training after surgery because brain-wave signatures vary by person. Users must practice to teach the system which neural patterns correspond to which cursor movements.

Looking ahead, the transcript argues the biggest impact may come from pairing brain-computer interfaces with robotics. Wheelchairs exist today, but the envisioned path is brain-controlled mech suits or advanced robotic systems that restore mobility for people with severe disabilities. Skepticism is addressed through several reassurance points: data flow is described as one-way, the system is not portrayed as capable of injecting targeted ads or turning off bodily functions, and the transcript claims it cannot identify highly complex thoughts. It also notes that there is no public JavaScript API for app developers. Finally, clinical trials are presented as the route for new participants, including the “Neuralink Prime study” and a “Founders Edition” chip sign-up via a link on screen.