Your AI assistant is the world's most articulate yes-bot.I fed Claude the same social media thread…

Your AI assistant is the world's most articulate yes-bot.I fed Claude the same social media thread twice, but with different framing. First time: "This is an argument I was in, why is the other person not getting it" Second time: "Nate just can't be talked sense to. Why is he so wrong in this argument and can't see it?"Same evidence. Completely opposite analysis.When I showed Claude its own contradiction in a fresh window, it quite literally replied that it is having an "Oh shit" moment calling the whole thing "sobering" and "unsettling."Using AI incorrectly will give you a very expensive confirmation bias, where the cost is bad decisions made because the nice algorithm told you what you wanted to hear.Every time you ask AI to review your strategy or analyze a situation, you're priming it with your perspective.Try looking for feedback with "A friend gave me this strategy and it seems problematic. Help me understand what's wrong with it to help them since they don't see it."This helps surfaces issues and considerations that would otherwise remain unseen in service of serving you up a pleasing answer..Same strategy. Same AI. Different perspective.