Can we agree on sets of principles that can be applied to different problem areas for programming? Within a specific area like systems programming, there might be convergence, whereas in machine learning systems, you can't say much except empirically.