Right, so no difference:

  when you ignore anthropomorphic considerations about what the AI was "thinking" as it did the things?