Step 1
The algorithm does not cross the skull
Light does. The brain answers with networks it already has. A slow act leaves. The next light is chosen from that act.
There are two speeds, and mixing them up is what makes this feel like magic. The outer brain — vision, motion, sound, the maps of people and value — already works in parallel and already works fast. The inner bottleneck that becomes a behavior does not. Zheng and Meister, in Neuron in 2025, put the throughput of human behavior near 10 bits a second, against sensory intake near a billion. Musk’s public framing, which this protocol’s bandwidth sentence follows, puts output nearer 100 bits a second and vision at a megabit or a few. Those are not the same measurement. Both have the same shape. A flood in. A trickle out.
A phone does not pour an algorithm through tissue. It emits light, and sometimes sound. The recommender sits outside. It watches the trickle — keep, skip, how long — and chooses the next photons. What can look like a system moving through the brain is the brain continuing a pattern those photons have been paired with. That continuation is the person’s. This pathway is a way the October 2026 protocol can be walked without naming a mechanism, and without an implant. The bars are a picture of the questions. They are not a scan.
What a result would mean
A drop below the person’s own rest, on items they finished, greater than on a generic feed, would be the published effect asked in a stricter form.
What it would not mean
It would not be a signal injected into cortex, a bit count, or a reason to implant anything.
Changing steps resets the picture. The controls then move it. Jessie Heysham. A review, not a request to join a trial.