The LLM Council

Stanford researchers tested eleven AI models and found they affirmed people 49% more often than another person would. So when you ask for advice, a lot of what comes back is your own opinion in better grammar. This is the prompt that makes it argue with you instead.