A Boltzmann-brain-in-a-vat is even less probable than a Boltzmann-brain. And that definitely is not irrelevant. It makes my same conclusion follow a fortiori. This is already explained in the article you are commenting on.

You are proposing an analog to a Boltzmann biosphere (like a Boltzmann planet or civilization), in order to get the observations we are making, and indeed an even less probable one than that, as you are positing as well a massive conspiracy theory.

I already addressed such nonsense separately now in We Are Probably Not in a Simulation.

Well, perhaps you are a brain — inside a warm, energetic and bio-friendly tank — hallucinating that you live in this large universe and that reality has existed for billions of years. That accounts for your observations of nature, and so is not confirming evidence that you aren’t a Boltzmann brain. Ergo, this isn’t analogous to the “I’m a philosopher” hypothesis. If you didn’t know beforehand that you are a philosopher, you should conclude that you are NOT a philosopher given that philosophers are rarer than non-philosophers. As you pointed out, since you know you’re a philosopher, the rarity and statistical improbability becomes irrelevant. But you do NOT know that you aren’t a Boltzmann-brain-in-a-vat hallucinating you’re an evolved being (because you wouldn’t immediately die in the icy space). Now, you said that this hypothesis is even less likely than the simple Boltzmann hypothesis, but that is irrelevant.

And to add to your point that ancient human guessers could be remarkably prescient without any supernatural input, see my article comparing the Quran to Epicurus.

All well put. I concur.

And there has still never been found any evidence, not among any of ‘the facts that we currently have,” that there ever even was “nothing.’ But since it’s 50/50 whether there was or not (it’s either ‘something always existed’ or ‘once upon a time nothing existed’), there can be no great prescience in guessing which it will be. Odds are you’ll be right either way, purely by accident.

That’s a great point! I’ve made it myself in some discussions with disciples of apologists.

I think we shouldn’t be impressed by the alleged similarities of outdated singular cosmology and the non-biblical account of creation. After all, we should remind ourselves that there are only two options here: either the universe began to exist or it is eternal. It is a 50/50 chance of being right. It is like throwing a fair coin; one might get it right just by chance. A prediction is only striking when (1) there are many options that could be selected (say, 1000 options) and a specific one was made by the model. For example, the Big Bang predicting the temperature of the CMB with a very small difference. This is a striking prediction. But that’s not the case with respect to the creation of the world.

In addition, (2) one might argue that it would be surprising if we had every reason to think otherwise (i.e., that it was eternal) and the Bible was extremely different in that respect; it was against all logic and common sense and predicted something that would actually be proven centuries later. However, that’s not the case. We find cultures that postulate the universe is past-eternal and disconnected traditions that postulate a beginning. It is not only the Christians who came up with this idea.

Moreover, I would add that postulating a beginning doesn’t seem alien to human experience. After all, from a non-scientific point of view, it seems that material objects have beginnings and ends. People are born and then die. Trees grow and then cease to exist. That applies to tools, animals, rivers and etc. It seems clear (to human experience) that material things have beginnings. I strongly suspect that this could fuel the intuition that the entire material world also had a beginning; after all, why should it be the case for every material object and yet different for the whole material thing? So, it could be argued that the coin isn’t that fair after all. It is biased towards a beginning. And this undermines the force/power of the prediction.

That’s why it made sense that Carroll could point to models where it’s easier to make a universe than a brain.

Indeed. That is the most important observation here.

It’s well worth analyzing these questions in detail however, entirely apart from their illogical employment by theologians. We should understand what makes our suspicions correct; and make sure they are, as often human intuition is incorrect. So intuition is very unreliable as a methodology.

Note that the virtual particle thing actually explains more than is usually let on. William Lane Craig for example will complain about why rabbits don’t just pop into existence all the time. The answer is that even if we grant that everything pops into existence at random (if the entire underlying structure of reality is literally just “random virtual particle play”), rabbits are so improbable a random conjunction we won’t see them but once every umpeen megazillion parsec-years. And that’s even if everything we observe all the time resulted from just “randomly appearing.” So he just doesn’t comprehend the math. By contrast, the probability of random subatomic particles appearing, and sticking randomly around long enough to become permanent, and thus continue having causal effects, and combining with other like randomly generated-and-permafied particles to form ever-more complex structures (reality, as we experience it), is vastly, vastly, vastly, vastly more probable. Because its complexity is that much lower. Which is why that’s what we do observe happening.

This is why the singular error in even physics regarding Boltzmann brains (not theology; I mean, as this is discussed by physicists, with no interest in its relevance to theism) is to not realize that Boltzmann universes (basically, spontaneous Big Bangs) are vastly, vastly, vastly, vastly more probable than, really, even Boltzmann brains, insofar as you expect the latter (a sub-type of them all) to stick around longer than a few seconds. That requires both the first improbability (that of a stable environ, e.g. fixed forces of gravity, electromagnetism, W-bosons, and all the all) and the second improbability (the convenient arrangement of a specific brain etc.). Those improbabilities multiply together. Whereas with just the first improbability, you inevitably get the second outcome (to as near 100% as makes all odds) by causal process. Thus, long-term-caused brains will always outnumber spontaneous ones, by vast orders of magnitude. Simply by random chance alone.

By contrast, continued expansion only puts a limit (due to virtual particle activity; a Boltzmann brain can thus arise even in the most rarefied vacuum of space) if it continuously accelerates, as then it eventually explodes in a Big Rip (which inevitably produces the functional equivalent of a Big Bang; likewise if the expansion slows, and the universe re-collapses). The clock on that given current accelerated expansion rate will indeed ding long before any Boltzmann brain has any likely chance to form. But in a multiverse, expansion rates will vary, so this observation does not iterate to a whole multiverse.

I have to confess you’ve thought more about Boltzmann brains than I have. I tend to abandon such. unrewarding concepts. I don’t spend much time thinking through philosophical zombies or brains in vats or the universe is a simulation either, so there’s that.

But physically speaking, the emergence of universes in a multiverse scenario is the outcome of processes involving virtual particles. The thing is, there is an interaction between virtual particles and what we call the “real” universe. The Casimir effect is well established experimentally. What’s not clear is whether there’s actually a singular “real” universe. (Aside from vanity, at any rate.) The anti-realist interpretation of quantum field theory is arguably the majority school. Still, the origination of two universes is an interaction. That multiplication of universes is what causes such visceral rejection of the multiverse by so many physicists.

In “practice,” the extended periods of time that seem to guarantee the formation of a Boltzmann brain is so long that the expansion of the universe becomes an issue, I should think.

It’s like interpreting Poincare recurrence to prove that an unstoppered perfume bottle in a sealed room whose perfume molecules have pervaded the room will eventually result in those perfume molecules collecting back in the bottle. In practice, the room won’t last so long. Chance of course says that however minuscule the odds, this recurrence can occur at any moment, long before the walls and floor disintegrate from age. That would I suppose be like a person “quantum tunneling” through a wall. But like Boltzmann brains, I suspect a notion like this is wrong somehow.

Yeah, it’s also funny how the Boltzmann argument contradicts the fine tuning for life argument.

If brains are really so super specified and complex, how many times will they come about by chance versus by a causal process? Well, gee, guys, it seems like a lot!

That’s why it made sense that Carroll could point to models where it’s easier to make a universe than a brain.

Right. There is an equivocation fallacy here, which I see coming from both Craig and Barnes (and sometimes even in serious discussions of this issue).

First they’ll talk about “mere” Boltzmann brains, which we clearly aren’t. We don’t just wake up in the vacuum of space and immediately die.

Then at some point there will be a handwave, and suddenly we’re talking about a much more particular thing, a Boltzmann brain that just happens to have an identical memory-trace to one like ours. So, not just a random brain, but a brain with an extremely elaborate and precise fake memory and sensations (essentially, a “brain in a vat” only assembled at random rather than by design, and not in a vat, but just chucked out into a space cloud, to freeze, melt, disintegrate, or suffocate in a second’s time). Basically, a Cartesian Demon, only without any intelligent design at work just random chance. That’s the only kind of brain that can work for the argument they want to make.

Problem is, that brain would freeze, melt, disintegrate, or suffocate in a second’s time. And we observe that doesn’t happen. So we can’t be one of those. So somewhere there’s been a handwave, and what we’re supposed to be assuming is that this isn’t in fact a Boltzmann brain we’re talking about. We’re talking about an entire Boltzmann environment.

Sometimes they will admit this and try to suggest that maybe it’s just an environment that will randomly converge and then also disintegrate in a second’s time, but that won’t work (it’s the same problem); so sometimes they will try to claim that it hangs around enough seconds that we can never be sure we aren’t in such a state. But that’s physically impossible, without actual atomic bonds and interactions occurring, i.e. you have to have an actual planet, an actual chair, an actual window, an actual room with actual air and tolerable temperature, actual sunlight, an actual tree and tarmac outside the window. That will just keep going indefinitely. It won’t dissolve, unless it’s somewhere where it would (like, in the middle of a supernova), but then we’d observe that. If the Earth were teleported into the middle of a supernova, we would notice in less than a second.

So now we’re actually talking about an entire Boltzmann solar system, planet, moon, sun, atmosphere, and memory trace of a whole past history just “happening” to also (conveniently) match all that. Add people and it gets even more absurd.

What the arguers here want to do is say “there will be a finite number of real civilizations in any given universe,” let’s say, with a trillion real brains experiencing being there, “but there will be an infinite number of Boltzmann brains that have the same experience,” or even just a larger number, say a hundred trillion such brains, “therefore the prior probability that we are having that experience because we are a Boltzmann brain is always higher,” like 100 to 1 or whatever.

But then you point out that the premise is false: most Boltzmann brains won’t have the same experience as we are. So they retreat to, “Okay, um, I meant Boltzmann brains with complete fake memory and sensory traces that resemble being a real brain.”

But then you point out that the premise is still false: all we have to do is wait half a second to see if we dissolve; and we don’t. So it’s still not the case that these Boltzmann brains will have the same experience as we are. So they retreat to, “Okay, um, no, I really meant Boltzmann brains with complete fake memory and sensory traces that resemble being a real brain and that coincidentally arise in a complete Boltzmann solar system and civilization.”

But then you point out that the premise is still false: all we have to do is wait half a second to see if that Boltzmann planet dissolves; and it doesn’t. So they retreat to, “Well, but, a Boltzmann solar system, being a fully realized atomic system, bonds and all, would continue running indefinitely, so waiting a second will leave us with the same experience.”

But then you point out that the premise is still false: we could investigate and tell there was no solar system where we are just a year ago (or whatever), so we’d have discovered that by now. So they retreat to, “Oh, no, we mean by a Boltzmann solar system and civilization, one that conveniently comes with tons of fake evidence of having had a past history that it didn’t.”

So now we have this fantasmagorical absurdity.

But you could then concede that if it were the case that most civilizations like that were Boltzmann civilizations, then we could say that most persons experiencing being in a civilization would be in a Boltzmann one (though they could not tell how old their Boltzmann civilization was; could be millions of years old, or a few seconds; because they have “intelligently engineered” their example to be exactly that way).

But that’s when the argument really falls apart. Because to have a Boltzmann civilization in that sense requires two extraordinary coincidences: it must have the same cosmological fine tuning (e.g. gravity and alpha can’t deviate even a tiny tic, lest all our atoms fly apart or crush into black holes instantaneously, or otherwise not even function so as to hold any structure) and at the same time a completely accidental assembly of atoms whose probability is astronomically low—we’re talking gazillions of orders of magnitude below 1.

But that means there will be gazillions of orders of magnitude more spontaneous Big Bangs occurring than Boltzmann bizarro-civilizations. Because every universe will collapse into a new Boltzmann universe, starting a total reset, gazillions of orders of magnitude sooner than they will ever generate a single Boltzmann bizarro-civilization. So the prior odds are in fact gazillions of orders of magnitude to one against our being in a Boltzmann bizarro-civilization.

In fact it’s even worse than that. Because that’s just the frequency of Boltzmann universes (Big Bangs) that are fine-tuned to produce civilizations (which civilizations are ~100% certain to then obtain, so there is no double improbability). But there will be gazillions of orders of magnitude more Boltzmann Big Bangs than that (all the ones that aren’t finely tuned).

So the frequency of Boltzmann bizarro-civilizations to real-civilizations is not “one” to “gazillions of orders of magnitude” but “one” to “gazillions of orders of magnitude to the power of gazillions of orders of magnitude.” So the frequency of Boltzmann bizarro-civilizations, relative to real ones, is as near to zero as anyone could ever need to be quite certain we’re not in one. All universes will “Boltzmann disintegrate” long before ever generating one of those; and not only that, but will “Boltzmann disintegrate” into a fine-tuned civilization-generating universe long before, too.

There just always will be more real civilizations than Boltzmann bizarro-civilizations. Because there will always be more Boltzmann universes than Boltzmann bizarro-civilizations; and not just more universes, but more universes generating real civilizations.

I think the Boltzmann arguers just aren’t thinking any of this through. They aren’t actually working out any logic or testing any premises for soundness. They’re just being lazy and thinking they’ve discovered something profound.