[kelnos]

The point isn't that they can be preserved; you're right that they probably wouldn't be. But all it takes is a single instant where "you" are randomly "conjured" into being: your Boltzmann brain has all your memories and everything. Even if that brain dissipates a small fraction of a second later, for that tiny amount of time, you would believe in the world as we all know it.

And the bizarre thing about this is you can't say "oh look, I've been in existence long enough to type this sentence, so I and my memories and reality must be real, and not just a random spontaneous formation of a brain", because at every single instant, you could be that brand-new Boltzmann brain, never before formed, and the bits where you thought "oh look, I've been in existence long enough..." are just the memories spontaneously implanted into that brain.

It's kinda wild. I'm not sure I buy it as a realistic possibility, but it can be fun (or terrifying) to think about.

[jancsika]

This seems like a physicist with no philosophical background spontaneously (heh) discovering our epistemic hard-dependency on sensory perception.

In terms of philosophy, what's here that can't be found in, say, Descartes or Hume?

[axlee]

The difference between this concept and, let's say, Descarte's evil demon, is not the philosophical skepticism but its explicit grounding in physics and thermodynamics. It basically attempts to answer the question "Where would that evil demon come from?". It materializes Descarte's thought experiment and shows that it could actually happen within the confines of our scientific knowledge, unlike malicious demons.

[jancsika]

I think it's the other way around.

Descartes/Hume are saying that to even bootstrap our understanding of reality, we have a hard dependency on sensory perception. (I mention Hume because he points out that even Descartes' singular ground-truth can't lead anywhere else without linking sensory perception back into the mix.) And when I say "nearly anything" it includes our notions about the laws of physics. (Which, btw, cannot be derived from Descartes' singular ground-truth.)

At best, BB is a restatement of what I wrote with the philosophically irrelevant detail that the BB hypothesis relies on all the same laws of physics we have in common with our universe. But I imagine it's really meant commonly as a weaker claim-- one which takes the laws of physics as epistemological ground-truth to derive an ambiguity about the nature of our reality within that universe.

My speculation is that science-minded people think BB is the most potent thought experiment for the same reason non-musicians might think Pachelbel's Canon in D is the best ever-- they've heard it a lot at places filled with people they admire.

[soulofmischief]

As someone familiar with both Descartes and Boltzmann, I will chime in and say that you're approaching this from an angle of contempt and defensiveness, imagining the Boltzmann brain as an inferior subset of or analogy to long-studied philosophical and metaphysical issues such as ground truth or the evil demon. Instead, I implore you to give benefit of the doubt and attempt to understand the differences.

The Boltzmann brain is not making some grand statement on ground truth or perception. It's not about intrinsics or perception at all. Boltzmann discussed how the universe, even in a state of 100% thermodynamic equilibrium, may spontaneously end up in a state of non-equilibrium, reducing entropy. The Boltzmann brain was a concept developed by others in response to this theory.

In fact, many theories are such that a Boltzmann brain actually has a higher chance of occurring than all of the billions of years of coincidences which led up to me typing this message out to you.

It's purely an argument of entropy and spontaneous symmetry breaking. The sensory and perceptive states described by the Boltzmann brain only serve to illustrate the point, and are not the main subject of the problem.

Don't forget that philosophy was the first science, and viewing people as "science-minded" (and therefore not philosophy-minded) hurts the scientific legitimacy of philosophy, and also only serves to exclude. Many scientists also have deep philosophical grounding. Many also have deep musical grounding. You yourself are exhibiting a lack of domain knowledge regarding the Boltzmann brain, filtering it through your "philosophy-minded" perspective, so maybe we can dispense with these kinds of judgements and focus on the core argument.

The world is not so black and white, and there is no false dichotomy between people who are "science-minded" and "philosophy-minded". Both follow the same exact scientific method of inquiry.

Additionally, "the same reason non-musicians might think Pachelbel's Canon in D is the best ever" comes across as a strawman. Some people might prefer that piece overall, but it's not a crime for someone to enjoy it. But relatively few probably consider it "the best ever".

But also, who cares? Why judge? I have a lot of favorite modern pieces which are technically inferior to most classical pieces. But as a musician, not just an engineer, I consider sensory evocation to be equally as important as technicality.

[sdwr]

Sounds about right, except Pachelbel's Canon actually is the best song ever for non-musicians.

It showcases harmony and contrasting lines in the simplest, punchiest, most pleasing way. It's fundamentally "the good stuff", and untrained ears slurp it up like babies w pats of butter.

[raattgift]

Probably no physicist thinks that Boltzman brains are a potent thought experiment. BBs are short-hand for a problem in combining cosmology and statistical mechanics in a way in which there is a hierarchy of vastly improbable configurations fluctuating into existence out of thermal equilibrium.

Discounting Brain-in-a-Vat (because it's cognitively useless), the problem in a nutshell is that we inhabit a universe which appears (a) to have had a hot dense phase in approximate thermal equilibrium, (b) a future sparse phase in approximate thermal equilibrium, and (c) a whole bunch of structure in between those. Is the structure a fluctuation in (a)? Could (a) be a fluctuation in (b)? These are reasonable questions about which one can ask: is there astrophysical or laboratory evidence available to determine the answers?

One problem is that if (a) (early conditions) is a fluctuation in (b) (late conditions), wherein (a) simply evolves into (c) (complex structure with galaxies and so on) and then (b), what mechanisms could suppress simpler configurations than (a)?

A huge huge huge number of low-entropy Boltzmann brains fluctuating into existence is vastly more likley (on Boltzmann entropy grounds) than an early very-very-very-very-very-low-entropy universe compatible with the standard model of particle physics and the cosmic microwave background and galaxies all over the sky, in which there is a nonzero chance of human brains arising via evolutionary processes.

A tiny change in a Boltzmann brain as it fluctuates into existence could lead to a significant loss of false memory; a tiny change in a maximally-hot maximally-dense phase in the early universe could lead to completely different chemical elements (or none at all).

So Boltzmann brains highlight some metaphysical ratholes one can fall into with respect to the fine-tuning of the (a) state, and have provoked work on how (a) could be so generic an outcome that the evolution of (a)->(c) is "unsurprising". The hard part is coming up with observables which usefully compare a given hypothetical solution and our own sky.

[anon34223879]

I would question the idea of bootstrapping; rather sensory perception is a QC/QA function that confirms the brain's construct of reality.

The only reason we can generally agree on the nature of any object is our common evolution and generally the same sensory ability. (but that isn't universal and differs widely across species)

[griffzhowl]

Yes, and the idea of Boltzmann brains depends for its credence on those physical principles used to derive it. And those principles in turn depend on the reality (or at least reasonable reliability) of our memories and the whole history of experimental and theoretical development leading to them. So trying to use it as an arugment for philosophical scepticism, or that it's a probable scenario, would be self-defeating, denying its own evidence.

It does give a technically detailed construction for how such a scenrio might come about though, as you say, so it can be interesting to think about.

[jancsika]

Sure. But as I said to another respondent, it's a much weaker claim than those which already exist in the philosophical literature.

That doesn't matter if BB is just a bridge for physicists to a deeper understand of philosophy. But I have a sneaking suspicion that BB is part of a basket of ideas in a kind of bubble category of "Philosophy for the Scientist." Similar to those "101 Jokes for Golfers" books-- I mean, fine, but if those are the only jokes you know you're probably insufferable at parties.

[lazide]

Well, the BB isn’t necessarily a serious proposal.

Rather, when you do the math on all the billion/trillion to one shots that are definitely happening, every second of every day, in physics - and look around at the universe as it appears to exist now and how many of those shots had to play out in a specific way - and then do the math on the probability of a BB spontaneously existing, then it’s really absurd that we aren’t somehow BB’s.

[namaria]

I don't know, I've always seen Boltzman brains like I see Schröedinger's cat. I think it was intended to show the absurdity of certain philosophical interpretations of the data available, but somehow got misapprehended as an argument defending those interpretations. Namely that randomness could somehow spontaneously lead to ordered complexity.

[moi2388]

But this presupposes that consciousness is a mere instant. If I am conscious ‘now’ and maintain this State for Lets say 2 seconds, the chances of me being a boltzmann brain pretty quickly drop to near 0

[lazide]

And how could you actually determine if we’re actually in second 2, or actually in second 0 with knowledge/memories/sensory input which makes you think we’re in second 2?

At some point it does boil down to faith that we aren’t in that scenario/it isn’t what is occurring. Philosophically, it’s hard/impossible to fully logic our way out of this kind of problem.

[moi2388]

You can’t, for a single point. My argument is that you don’t have consciousness for a single point. Thoughts take time, we have streams of consciousness.

If I can determine a current point (I argue I can), with memories, then remain conscious for let’s say 2 seconds, the chance of me being a Boltzmann brain already dropped to almost 0.

At any point you could say false memories were planted, but this assumes that I am only conscious in an instant, and merely remember previous states. Both research and my own experience seem to indicate it’s not this coherent, and we have a much longer instant of consciousness or ‘moment’.

A Boltzmann brain only makes sense for particular instants thermodynamically, coherency is as good as impossible

[lazide]

You might want to re-read my prior comment.

[medo-bear]

Taking scientific knowledge to their weird limits might also signal something perhaps more likely: a missing puzzle in our knowledge