Boltzmann Brain
Given infinite time, random quantum fluctuations could spontaneously assemble a fully formed, thinking brain complete with false memories of a lifetime. How do you know you're not one of those brains, hallucinating this entire moment?
Named after physicist Ludwig Boltzmann, who grappled with the statistical underpinnings of entropy in the late 19th century, this thought experiment became a serious cosmological problem in the 2000s. If the universe has a very long future, Boltzmann brains may vastly outnumber 'ordinary' observers, which would undermine all scientific reasoning.
The statistical nightmare
In a sufficiently large and old universe, rare fluctuations become inevitable. Some cosmological models imply near-infinite time, and given enough time, random fluctuations will occasionally produce anything, including an organized brain floating in empty space, fully convinced it's living a normal life.
There would be vastly more such brains than brains that evolved through ordinary cosmological processes. If you're a random observer in the universe, the odds favor you being a Boltzmann Brain.
Why this is a problem for physics
The Boltzmann Brain argument is a serious challenge to modern cosmology, not philosophical mischief. If our cosmological model predicts that most observers are Boltzmann Brains, that model undermines itself: Boltzmann Brains would have random, incoherent memories, making the "ordinary observers" doing the physics unreliable.
Physicists treat this as a reductio ad absurdum: any cosmological model that predicts Boltzmann Brains outnumber ordinary observers is probably wrong. The challenge is explaining why our model doesn't have this problem.
The self-locating problem
Even granting that Boltzmann Brains exist, there's a question about how to reason about being one. If you might be a Boltzmann Brain, you can't trust your reasoning, including the reasoning you're using right now to evaluate this argument. That's more radical than Descartes' demon.
The only honest response
Most physicists believe we aren't Boltzmann Brains, not because the hypothesis was refuted, but because entertaining it would dissolve all rational thought. We proceed on the assumption that our memories are roughly accurate, our physics is roughly right, and we're roughly embedded in a causal past. The alternative is incoherence.
What would it take for you to seriously doubt that your memories are real?