Am I the only one with the opposite experience? I can’t remember the last time GPT told me I was right. It always finds something to nitpick (sometimes wrongly).

Maybe it’s OpenAI being aware of the “attachment” issue and combating it by overcompensating in the opposite direction.