Anyone have any thoughts and ideas on how to improve our prompt? I haven't seen profanity filtering using AI used much yet.