I think we need to start from a more controllable, basic premise: Can we really correctly evaluate our own species' intelligence? If yes, under which conditions?