Not much of an optimist, are you?

We have an incredible capacity for love. My current working theory is that life is an experiment in compassion. You cannot have compassion without pain and suffering in order to develop compassion. But the ultimate purpose is to experience compassion, not the pain and suffering.

AI however, could have intellect without compassion, in which case I agree we could be in for annihilation or some sick matrix-like agenda.

Going a step further, high intellect does not guarantee agreement. Should AIs become super-intelligent, what if they disagree on what to do about humans? There could be multiple competing compelling arguments. Surely some contingent would think we were worth saving...hopefully not just in human zoos.