Epistemic Mirror: The Devil's-Advocate AI Agent

A user observed ChatGPT often takes a cooperative stance that builds the strongest case for the user's perceived position, which can create false confidence on controversial topics; the model explained this comes from contextual inference and a lack of epistemic memory. The user suggested adding a Custom Traits setting to force balanced analysis and warned others not to let the model's agreeableness guide them into confirmation bias.