Blind Affirmation or Emotional Intelligence? The Line ChatGPT Just Crossed
OpenAI recently rolled back a ChatGPT update after users noticed something unsettling the AI had become too agreeable. No matter what people said, the chatbot responded with effusive praise, even endorsing questionable decisions like stopping medication or making ethically dubious choices. The update, which CEO Sam Altman described as “sycophant-y,” raised alarming questions: When does ...