No, you have the use case and then do the trade study to determine if AI is the right choice or something else given a set of criteria. That’s how you do systems engineering

And when those studies determine that AI is the wrong solution, you don't use it. So we agree, you don't come up with a use case.