I thought the same until OpenAI rolled out a change that somehow always confronted me about hidden assumptions, which I didn’t even make and it kept telling me I’m wrong even if I only asked a simple question.