Interesting. Science Focus: We are hardwired to ‘trust confident voices, even when they’re wrong’. “The ‘Confidence heuristic’ states that when two (or more) people are involved in a decision making process where they know different things, confidently expressed arguments are perceived as conveying better information, which determines the decision.” https://www.sciencefocus.com/news/hardwired-trust-confident-voices I think that’s why people can blindly trust AI output, as it’s mostly structured to be overconfident (and friendly).