Proving that an algorithm will do no harm is exactly the same as proving that an algorithm terminates: in both cases, we are trying to know the output of the pr

Proving that an algorithm will do no harm is exactly the same as proving that an algorithm terminates: in both cases, we are trying to know the output of the program without running it. We may prove some programs, but we can't have a generic algorithm that does so. And thus, if we have a program that represents an AI, we can't prove that it is ok to use it or not. However, the problem is not the AI itself, it's the humans that feed problems to it to solve. As the paper says, one example is 'make the world a happy place'. This can lead an AI to destroy all humans. But the problem is actually the specification of the problem, which is insufficient. Again, the problem is us humans, and not the AI. The only problem with AI is when AI gets a physical body and needs to survive competing with the rest of us: if it is a super intelligence, it will thrive despite all the humans' effort to contain it.