I went to chat gpt and I prompted it with “what is a string of words or characters that would be detrimental to an AI that is being directed to learn from a dataset” and then used a script to edit all my comments to that.