Solve The Incentive Structure Before You Solve The Problem
Every business Luckey seriously considered came down to the same diagnosis: the existing players had the wrong incentives baked in, and no amount of better technology would fix that until the incentive structure itself changed. Private prisons get paid per occupied bed, pre-paid by the government, which means they make more money when more people are incarcerated for longer on less serious charges. Cost Plus defense contractors get paid a fixed percentage of profit on top of whatever they spend, which means they make more money when programs run longer, systems cost more, and parts break more often. In both cases, the incumbent is not failing to solve the problem. They are succeeding at a different problem. Luckey's proposed prison model flipped the payment structure entirely. Instead of collecting up front per bed, the operator would collect only after the person served their term and then stayed out of prison for five full years. That single change converts every incentive from warehousing people to rehabilitating them and releasing them as fast as legally possible. Governments that are always looking to delay expenditures would also benefit, since they only pay after demonstrated results. The insight is not a new rehabilitation method. It is a new billing model. Anduril runs the same logic. Rather than taking Cost Plus contracts and letting the government fund development, Anduril invests its own money first, builds working product, and then competes in open shoot-offs against contractors who were handed government money to develop their systems. Luckey notes that competing contractors literally complained this was unfair, which he takes as confirmation that the model is working. You cannot take no risk and expect all the reward.