I described a couple of AIs that do a couple of very specific things well. They probably won't even be able to understand each other. An AI that "is never wrong

I described a couple of AIs that do a couple of very specific things well. They probably won't even be able to understand each other. An AI that "is never wrong" is complete science fiction. General AI is a fascinating thing to imagine, and great fodder for fiction, but it has no basis in science.