Neuralink tests Telepathy as Blindsight targets the visual cortex

By Elena Vasquez

Neuralink is testing Telepathy in people with quadriplegia through three tasks—cursor control, typing and video-game play—with an implanted N1 device under the PRIME Study.

The announcement followed the recent launch of Musk’s so-called “flying car.” Neuralink presented Telepathy and Blindsight as two medical brain-computer interfaces designed to link the brain directly with external devices: one aimed at restoring motor control, the other at producing a rudimentary visual signal. The car supplied the news hook. The interfaces must now earn their claims in a laboratory and a clinic.

In an October 30, 2026 social media post, Elon Musk said: “Neuralink’s first product is called Telepathy. It enables high-speed communication with computers or humans simply by thinking. Output bandwidth will vastly exceed the fastest human speech or typing in a few years. Read that sentence again.”

The claim establishes the target, not the result. The post gives no output rate, latency, participant count or comparison data. “In a few years” also places the forecast beyond the current trial. In a reading room, that sentence is a hypothesis awaiting measurement, not a bandwidth specification.

What does this change for patients and clinicians?

For people with severe spinal cord injury or amyotrophic lateral sclerosis who have lost limb control, Telepathy offers a potential route to computers, smartphones and other compatible devices through decoded neural activity rather than physical movement. For routine care, the change remains small: Neuralink must show that this route works reliably enough to matter.

Telepathy captures the brain’s electrical signals through the implanted N1 device and converts them into digital commands in real time. The recording method is implanted neural-signal capture, not scalp EEG; the description provides no scalp montage or noninvasive recording setup.

The PRIME trial, formally the Precise Robotically Implanted Brain-Computer Interface study, evaluates four practical domains: implant safety, wireless connectivity, accuracy of real-time neural-signal decoding, and long-term stability of the thread-like electrodes in brain tissue. The goal is reliable, sustained device control over extended periods.

These are monitoring-unit questions with clinical consequences. Loose hardware can interrupt control. An unstable interface can erode performance. The trial must show that useful control persists.

For participants with quadriplegia, investigators are assessing the ability to control a cursor, type and play video games. The announcement reports no trial results, participant count or follow-up duration. Telepathy therefore remains an investigational BCI, not a clinically validated device-control method. The announcement does not describe it as an FDA-cleared device.

How does Blindsight reach the visual cortex?

Blindsight addresses a different deficit. Neuralink developed it for people with severe or complete vision loss, including potential candidates whose optic nerves have been damaged or severed. Rather than send image data through the usual visual pathway, the system converts external visual information into stimulation of the primary visual cortex, or V1.

Its proposed pathway has four stages:

• Digital capture: External cameras or optical sensors record images and visual information from the user’s surroundings.
• Signal translation: Processing hardware converts visual patterns, edges and spatial details into electrical stimulation patterns.
• Cortical delivery: Electrode arrays implanted in V1 deliver controlled electrical pulses to targeted cortical areas.
• Phosphene mapping: Stimulation activates selected groups of neurons, producing tiny perceived flashes or points of light called phosphenes.

By stimulating multiple neural locations, the system aims to build a representation of the environment that could eventually support perception of shapes, patterns and movement. The first building block is not scene vision. It is a field of dots.

That distinction matters. A phosphene shows that stimulation can produce a perceptual event; it does not establish natural, detailed vision. Neuralink still needs evidence on safety, effectiveness and practical use before Blindsight can support real-world decisions. The program remains experimental, and the announcement identifies no regulatory clearance for either technology.

Where does the evidence stand?

Telepathy has entered human clinical testing through PRIME; Blindsight remains an experimental neural-stimulation concept. Neither part of the announcement supplies completed efficacy results. For clinicians and laboratory teams, the next useful evidence will come from sustained follow-up: decoder accuracy, wireless performance, electrode stability, safety and whether users can perform meaningful tasks without ordinary motor control.

The next decisive evidence will come from sustained PRIME follow-up and further Blindsight safety and efficacy work, not from another speed claim.