[InternetPerson]

That's why Boltzmann brains are a problem. If a theory predicts the appearance of Boltzmann brains, then that theory is (arguably) self-defeating.

For example, let's say you tell me your theory of the universe. And then I say, "Wait a minute, doesn't your theory lead to an infinite stretch of time where random brains can randomly spring into existence?"

If you say, "yes", then I'd say, "If your theory is true, then I'm probably just a Boltzmann brain, and this whole conversation is just a figment of my imagination."

I would assume that I'm probably a Boltzmann brain because the number of Boltzmann brains that ever exist will be far larger than the number of human brains that ever exist. Even if it takes a zillion years for a Boltzmann brain to appear, it will happen zillion times, over an infinite stretch of time.

Sean Carroll discusses this in much more depth in "Why Boltzmann Brains Are Bad" [1]. In the paper, he argues that "the theories that predict [Boltzmann Brains] are cognitively unstable: they cannot simultaneously be true and justifiably believed."

[1] https://arxiv.org/abs/1702.00850