In 1865, William Stanley Jevons observed that James Watt's more efficient steam engine, which consumed less coal per unit of work, led to increased overall coal use as cheaper engines were installed everywhere. This is relevant to AI, whose value consistently appears somewhere other than where it was installed, in a different P&L, business line, sometimes as a currency other than money. For example, UPS's ORION routing system, a decade in development with over a billion dollars invested, was built to solve a logistics problem. According to UPS's own figures, cited by INFORMS, the fully deployed system cuts roughly 100 million miles driven annually, saves 10 million gallons of fuel, and delivers $300–400 million in yearly cost avoidance, none of which is revenue. Rolls-Royce's "Power by the Hour" model, which has sold guaranteed flying hours instead of engines since 1962, was scaled by AI starting with the IntelligentEngine program in 2018. Predictive maintenance algorithms let Rolls-Royce underwrite uptime guarantees across a larger, more complex fleet with less financial risk. AI generated the confidence to sell the business model at industrial scale.