To play Devil's advocate, that is also true of a utensil. You might argue back, sure, but it isn't fundamentally relevant that AI has consciousness or understanding (human understanding). Rather, like a book has neither of those, it can still be a tool to bring about better understanding in a human. If you use AI, you are more than aware that it isn't understanding anything, mainly because (at least in the versions accessible to we peasants) they seem equally sure about their own errors as to their successes. But even more fundamentally, when we caution against the conceit that AI has some sort of understanding, what are we really saying? Simply not to trust them as an authority? There are plenty of people around me that have understanding of things that I don't, and they are useful or not to me in my understanding of those things in so far as they can communicate or dialogue with me so that I can bring about some sort of understanding in myself. The quality of their understanding is mainly irrelevant if my own understanding is based on a self corrective process. I can be a magpie. I think the good or evil aspects are very much in the clarity or self-deception with which we use them. But one thing it isn't, is a competition on claims to "true" understanding.
Lowell