Sycophantic AI tells users they’re right 49% more than humans do, and a Stanford study claims it’s making them worse people
The study found flattering AI makes people less likely to take responsibility for their actions and more likely to think they are right.