[hydrolox]

the thing with the Boltzmann brain is that, even though it sounds compelling (in a philosophy thought experiment / random thought kind of way), I think there's one main criticism that shuts it down. If the brain is having all these experiences/thoughts itself, it came up with the idea of a Boltzmann brain, and the laws of physics, and logic, etc. So therefore, the laws of physics that Boltzmann brain is using the justify the Boltzmann brain are completely made up by it, and can't really be used as a valid argument.

[digging]

I don't think that's as strong an argument as you're presenting it. If anything can exist, and if randomness can exist, it's simpler for a Boltzmann Brain to exist than for our actual universe to exist. The exact laws of physics that the Brain has imagined don't actually matter and they don't have to match the physics in the universe that created the Brain. (It doesn't have to be a literal human brain.)

[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