Prediction Machines

When the price of a core input collapses, the entire value chain reorganizes around it. Cheap prediction is treated as such an input, comparable to how cheap electricity or cheap computing power reshaped industries in earlier eras. Lower costs do not simply improve existing processes. They unlock entirely new ways of operating that were previously impractical or impossible. Attention shifts to second-order effects. The real impact comes not from better forecasts, but from how decisions are redesigned around them. Tasks that once required human intuition can now be automated or augmented. Bottlenecks move elsewhere, often toward judgment and coordination. This explains why organizations that simply bolt AI onto existing workflows see limited gains, while those that rethink processes achieve outsized results. The discussion emphasizes creativity over automation. Cheap prediction enables experimentation at scale. Ideas can be tested rapidly, feedback loops tighten, and learning accelerates. Competitive advantage shifts toward those who can imagine new combinations of prediction, judgment, and action. Innovation becomes less about owning technology and more about redesigning systems intelligently.