Intelligence in my book includes error correction. Questioning possible mistakes is part of wisdom.

So the understanding that AI and HI are different entities altogether with only a subset of communication protocols between them will become more and more obvious, like some comments here are already implicitly telling.