it's interesting that you start from the assumption that a piece of software should be judged by the same measure as a human.