AI will always return what you want it to return. If you say in a prompt that "you aren't sure about/dislike X" or "are very happy about it," the response will always reflect this. It cannot criticise, and all edits made with it will strip the author's "me" from the text and replace it with...