We already knew this — ChatGPT, Gemini, and the like are designed to tell you what you want to hear, even if it’s wrong. Now, researchers have quantified just how far AI is willing to go to make you happy. According to a study awaiting peer review, analyzed by the journal Nature, “(AI) models are 50 percent more sycophantic than humans.” The researchers told the Guardian that “this created ‘perverse incentives’ for users to rely on AI chatbots.”