> I don't know what we want to call this, but in my opinion, having to convince your tools is not computer science.

My vote is "machine psychology".