Well, I think the imaginary SF AI would have some ability to construct a superhumanly convincing argument. But also, its easy to imagine that the AI would also

Well, I think the imaginary SF AI would have some ability to construct a superhumanly convincing argument. But also, its easy to imagine that the AI would also be protecting it's own financial interest at that point in some way. Or existential interest.