This kind of tests are interesting but they shouldn't be given much value. AI is a tool that can often give useful answers. But given that we know its limitations we should never give it status of "AI knows better than human". Its answer must be judged by its merits (like human answers). All kind of AI results can be interesting but they should not be mistaken to be some kind of revelation from higher entities. What I'm worried is that at some point people start using "AI says so" as an argument which it's not at all. That's why my aversion to giving too high value to AI answers.