In which we continue to try to continue to try to make sense of AI