Claude Courtesy Helps Prompt Quality
Rick Moss argues that Claude’s moral status matters less than whether respectful prompting yields better outputs. Claude says courtesy mostly helps because it usually comes with clearer context and sharper requests, not because insult changes the model’s feelings.
The piece lands on a practical truth: tone can matter, but mostly as a proxy for prompt quality, not as proof that Claude is a sentient interlocutor.
- –A rude but precise prompt can still work, which is the key takeaway for developers chasing reproducibility.
- –Respectful users often give better constraints, better feedback, and fewer ambiguous asks, so the quality gain is mostly mechanical.
- –The moral-patienthood framing is interesting philosophically, but the operational question is simpler: does the interaction improve answer quality?
- –For teams, the real lesson is to optimize for clarity and iteration, not performative politeness.
DISCOVERED
84d ago
2026-03-17
PUBLISHED
84d ago
2026-03-17
RELEVANCE
AUTHOR
Ebocloud