Fortunately AI is not right around the corner.

We haven't even come close to developing a system that could choose its own goals. We barely have systems that can choose from a list of goals... those goals only being 1,2,3,4,5. They have no meaning to the computer. It doesn't know what they are, it doesn't know what would be best for it.

We haven't even explored consciousness. (because it is in direct violation of materialist science) And such, we couldn't ever create an artificial consciousness accept by accident.

We do not even know the world we live in. We think we live on a ball earth, and live in linear time. How can we give something intelligence when the basics of our world are mistaken?

If we accidentally created an AI, would it talk to the angels, ghosts and elves around it? And then, we would conclude that we made an insane AI?