UI automation testing can be a nightmare: tests break all the time when developers are changing behaviors for new features or bug fixes. Here I'd like to propose why UI automation tests constantly break and a hypothetical solution to the nightmare.