How could a brain implant let a blind person see?

Elon Musk says Neuralink’s Blindsight implant is “ready from an engineering standpoint” and “just waiting on regulatory approval”, as he posted at the end of October 2026. The idea is to skip the eyes altogether: a camera sees for the person, and an implant in the back of the brain draws what it sees. So how could a brain implant let a blind person see?

▶ Try it yourself in Phosphene Lab: add electrodes from 1 to 10,000 and watch dots of light turn from a scatter of stars into a doorway, a letter or a face.

One electrode, one dot

Seeing happens in the brain, not the eye. The eye turns light into signals, and the visual cortex at the back of the brain turns those signals into a picture. In many kinds of blindness the eye or the optic nerve is damaged but that cortex still works. Stimulate one spot of it with a tiny electrical pulse and a person sees a dot of light, called a phosphene, always at the same place in view. In 2021 Eduardo Fernández and colleagues placed 96 electrodes in the visual cortex of a blind woman for six months; patterns of these dots let her identify some letters and the edges of objects.

A map with the center stretched big

The visual cortex is a map of what you see: neighboring spots of cortex see neighbouring points of view. But the map is badly out of proportion. The center of view, where you read and recognize faces, gets far more cortex than the edges. In the model behind Phosphene Lab, which follows a classic measurement by Horton and Hoyt, the central 2° of a 20° view is 1% of its area but takes about a quarter of the cortex, and the central 5° takes about half. So electrodes spread evenly over the brain land densely in the middle of the view and sparsely at the edges.

Dots for pixels

A camera turns each frame into a pattern of pulses: bright where the scene is bright, one dot per electrode. That makes the electrode count the resolution. In this model, 60 dots, about as many as the Orion implant had, read as a scatter of stars. About 1,000 start to draw a doorway, sharp in the middle and coarse at the edges. That is the same count as Neuralink’s N1 chip, which has 1,024 electrodes on 64 threads; Neuralink has not published how many Blindsight will have. In monkeys, Xing Chen and colleagues stimulated a 1,024-electrode implant in visual cortex in 2020, and the animals recognized shapes, motion and letters drawn in dots.

What Neuralink says

Neuralink announced in September 2024 that the FDA had given Blindsight Breakthrough Device designation, which speeds up review; it is not an approval. Musk wrote then that the vision will “at first be low resolution, like Atari graphics” and that it could one day be better than natural vision. Those are his claims. As of early November 2026 no Blindsight trial was listed on ClinicalTrials.gov and no human results had been published.

Phosphene Lab is a simplified, best-case model: real phosphene maps are irregular, not every electrode makes a dot, the dots move with the eyes, and real arrays cover only part of the map. The film uses short excerpts of Neuralink’s own videos on YouTube, credited on screen: its “Calibrated Vision” simulation, an animation of its visual-cortex map and its R1 surgical robot. The dots, the brain map and the numbers are drawn from the model.

Sources: E. Fernández et al., “Visual percepts evoked with an intracortical 96-channel microelectrode array inserted in human occipital cortex”, Journal of Clinical Investigation 131, e151331 (2021), doi; X. Chen et al., “Shape perception via a high-channel-count neuroprosthesis in monkey visual cortex”, Science 370, 1191–1196 (2020), doi; J. C. Horton and W. F. Hoyt, “The representation of the visual field in human striate cortex”, Archives of Ophthalmology 109, 816–824 (1991), doi; The Debrief, Neuralink receives FDA Breakthrough designation for Blindsight (2024); Tesla North, Elon says Blindsight is ready (2026).