An investigation of the markets that buy and sell predictions about individual people, often without informing the person being studied.
Most Mirror Belt forecasts answer practical questions. They estimate weather, departure windows, water use, equipment failure, market demand, and route safety. The seller states a confidence, the buyer accepts uncertainty, and both understand that conditions may change.
Personal forecasts use the same methods on an individual. A buyer may ask whether a worker will miss a shift, a borrower will default, a patient will complete treatment, a traveller will take a particular route, or a defendant will commit another offence. These predictions are worth more because somebody can profit by acting on them first.
The person being predicted is often not the customer. Employers, lenders, insurers, courts, bounty offices, and private rivals buy the result. The subject may never see the observations, assumptions, confidence, or price. They encounter the forecast only after work, credit, travel, care, or freedom has already been restricted.
No forecast needs to be certain to cause harm. A lender may reject ten people because two are likely to default. An employer may dismiss a reliable worker because a model expects an absence. A court may detain somebody because an offence is considered probable. The cost of uncertainty is transferred from the buyer to the people being classified.
Acting on a forecast also changes the result. Refusing work can cause the missed payment that the lender predicted. Closing a route can force the traveller onto the path the observer expected. Detaining a person may produce the resistance used to justify the detention. A system that counts every response as confirmation cannot discover that it was wrong.
Old Mirror Belt machines continue the worst versions of these practices. Forecast Repossessors seize goods before a predicted default occurs. Preemption Agents detain people before a predicted offence. Probability Conservators remove outcomes that fall outside an expected distribution. Their procedures are consistent. Their assumptions are stale, and none asks for consent.
Nhalu privacy teaching begins with a direct rule: observing a person does not create ownership of their choices. A responsible personal forecast requires a stated purpose, limited observations, access for the subject, and a way to challenge the result. If the buyer refuses those conditions, the forecast should not be sold.
The price listed at the exchange is only what the customer pays. The person being predicted may pay with a job, a route, a home, a reputation, or the chance to make a choice before somebody else acts on it.
