The "Confidence Trap" occurs when an LLM sounds authoritative while...

https://sierra-wiki.win/index.php/The_Mechanics_of_Correction:_How_Suprmind_Audits_Multi-Model_Workflows

The "Confidence Trap" occurs when an LLM sounds authoritative while hallucinating, a dangerous scenario for high-stakes workflows. You cannot trust a single model blindly