Those prompts are likely asked by right leaning people who believe in the pseudoscience. And it is textbook pseudoscience.

AI as it exists right now also can’t predict anything, because it doesn’t really “know” anything. They’re language models trained on examples of responses. It knows what a response to your prompt looks like and can draw on similar prompts.

I don’t think AI can determine that more accurately than a human can because any accuracy, even in humans, is based on luck and maybe intuition, not science.

“You Libs may think I’m a troll, but have you considered pseudoscience and quackery?”