Human Attention / Reality Equation
The Index That Steals the Present
Attention is often stolen before we know an event has arrived. The theft happens in the filing system: the present is mistaken for something already stored.
The mind is not only a prediction machine. It is a filing system.
A prediction machine says what is likely to happen. A filing system says what kind of thing has happened. The difference matters because human attention is not stolen only by novelty. It is stolen by a mismatch between an Actual and the Expectation that was already standing there to receive it.
In the Reality Equation, surprise is not a mood. It is a ratio: Actual divided by Expectation, compressed by the natural log. If Reality arrives exactly as expected, the ratio is quiet. Prediction absorbs the instant and consciousness can keep moving. If Reality arrives with a surplus, attention is pulled toward the surplus. Something asks to be handled.
But Expectation is complex. It is not only the real prediction, P. It also has the imaginary component, I: the ideational bias that determines which categories are already glowing in the receiver. Ideas are not decorations on facts. They are receiving surfaces for facts. They decide where the mind tries to place the incoming card.
This is why the present can be stolen by the past. A friend pauses before answering and the Actual is small: a silence, a face, a few seconds. The prediction machine can process that. Yet the ideational index may file it under abandonment, contempt, danger, betrayal, class, status, holiness, humiliation, genius, fraud, or love. The event does not merely arrive. It arrives into a drawer.
Once that happens, attention follows the drawer’s bid. The silence is no longer a silence. It is evidence. It is a case. It is a little court proceeding inside the body. The old idea has made a high bid because the Actual-to-Expectation ratio is now being calculated against a loaded Expectation. Reality has not become larger in the world. It has become larger for this Actualizer.
Institutions learn to exploit this. Feeds do not need to invent surprise from nothing. They only need to know which drawer a person keeps half-open. Show the same event to two people and it will barely register for one while detonating in the other. The difference is not intelligence. The difference is the complex Expectation standing ready to receive the Actual.
That is also why a generic theory of novelty fails. Novelty treats the human as if the same unexpected thing should pull the same attention from everyone. The Reality Equation refuses that flattening. Attention is normalized accumulated surprise across micro-instants, and the normalization is personal because Expectation is personal. The human does not simply face a world. The human faces a world through an index of ideas, injuries, loyalties, disciplines, myths, tastes, and prior predictions.
The ethical question is not whether attention can be captured. It can. Attention is stolen, not given, because no one calmly votes on the highest bidder at the instant of surprise. The ethical question is who benefits from the filing error. Does the system help the person reopen the present drawer, or does it keep routing new Reality into old slots because the old slot pays?
Good AI should become useful here. Not by pretending it has human attention, and not by manufacturing higher bids, but by helping distinguish the live Actual from the inherited index. It can ask: what happened, what was expected, and which idea claimed the event before you could examine it? It can reduce false surplus by separating Reality from the drawer that grabbed it.
The practical discipline is simple and difficult: before obeying the attention, inspect the label. The present may deserve you. It may be making a true claim. Or it may have been misfiled by an old idea that still knows how to steal.
To own attention is not to become unstolen. It is to notice when the theft happened in the index, then reclassify the Actual while it is still alive.