> I've seen many instances of users needing to yell at, abuse, or manipulate ChatGPT to get the desired answers.
I would like some examples of the filters getting in the way of 'desired answers'.
Three would be great.