Whatever AI optimizes, the world will reorganize around. The question is who writes the objective function.
Every previous wave of technology amplified human behavior. AI amplifies whatever objective we hand it — including the perverse ones already running in the systems we have. A platform paid for engagement will deploy AI to maximize engagement. A hospital paid per procedure will deploy AI to schedule more procedures. The model is not the problem. The incentive is.
Future-crafting is the discipline of choosing the objective on purpose. It pairs the Flip from Chapter 06 with the foresight tools the Lab uses — scenario simulation, equation runs, and pre-mortems — so the objective handed to the model is the one we actually want amplified.
Get this layer right and AI compounds the redesign. Get it wrong and AI compounds the perverse incentive faster than any institution can respond. There is no neutral deployment.
Treat the objective function as policy. Write it the way you would write a constitution.