GPT-5.6 refuses to explain why over-clarifying what a package lacks can hurt trust

moyix · x · 2026-07-22

A user jokes that GPT-5.6 refuses to comment after being fed a tweet arguing that over-explaining what something does not contain can make a report feel less trustworthy.

The thread frames a familiar LLM behavior complaint: clarifications like “there is no fake debugger step” can backfire and reduce trust instead of improving it.

Related event: Over-justification by LLMs undermines user trust(2 posts)→

Original post →

More from Models

Models channel →