Alvin Plantinga and Sean Carroll are on the same page

I recently read an article by Sean Carroll called "Why Boltzmann Brains Are Bad." What jumped out at me when I read this article was how similar it was to Alvin Plantinga's Evolutionary Argument Against Naturalism (EAAN).

Boltzmann Brains are not in a position to know true from false because all the information that comes their way just fluctuated into being without having any connection with reality. This could happen because the information fluctuated inside their brains, or it could happen because the world in their immediate vicinity fluctuated into existence. Either way, they cannot use their perceptions or any of their tools of reasoning to reliably come to true beliefs about the world.

If you have a model of the universe that predicts you are a Boltzmann Brain, then that model undermines any justification you would have for believing that model. The model is self-stutilfying because as soon as you believe it, for whatever reason, you lose your justification for believing it.

Carroll thinks this is a good reason to reject models that generate Boltzmann Brains. Since Boltzmann Brains are "cognitively unstable," we shouldn't even consider models that generate them. They could still be true, of course. It's just that we could never be justified in believing them since they undermine the reliability of the very process we used to come up with them.

This argument is just like Alvin Plantinga's EAAN. According to Plantinga, if both evolution and naturalism are true, then it's unlikely our brains would be able to reliably distinguish between true and false. Evolution combined with naturalism generates unreliable belief-producing cognitive faculties. So if we believe in both evolution and naturalism, then we have an undercutting defeater for all of our beliefs, including our belief in evolution and naturalism.

In both cases, they are considering models of the world that generate unreliable belief-producing brains, and they are both saying that even though it's possible for such models to be true, we can never be justified in believing them. We shouldn't even consider a model of the world that makes it likely that we can't tell true from false because if we can't tell true from false, then we can't know whether the model is true or false.

Neither of them claim to have proved these models to be false. They only claim to have shown the models are not reasonable to believe or even consider.