@derwinmcgeary I suppose in theory you could balance out the bias with careful manipulation of the AI, but it's going to take something a lot more sophisticated than adding a couple of words to the prompt.