Forgive me if I'm reading this incorrectly but doesn't your premise that "for every Boltzmann brain B that has the memory M, there is an evolved brain E that actually has done the things that make up M" undermine the argument you make in the post? A BB could observe things that would not be true in the real world because it is most likely just hallucinating a world it thinks is real. For example, it could imagine strong evidence of God such as meeting God or even being God. If this post is arguing in favor of atheism, then the premise that any BB's memory corresponds to an evolved brain undermines atheism. There is a huge range of false memories that a BB could have of fake laws of physics that don't correlate to anything in reality. This is why physicists such as Sean Carroll claim we should reject any theory that leads us to conclude that we are BBs. If we were BBs, we would not be able to reliably trust science (since our "science" would be hallucinated memories) and thus it's ultimately a self-undermining hypothesis. (https://arxiv.org/pdf/1702.00850)
This is an interesting point. I think the argument can maybe be rescued by the idea that even non-Boltzmann brains sometimes have faulty memories. E.g., if I take lots of Ambien, the next morning I'll remember having met a talking walrus (https://www.toothpastefordinner.com/index.php?search=ambien+walrus), even though that isn't true and is impossible.
Of course, that answer seems a little disastrous for me, and probably does lead to total skepticism...
Ok then, let's try again, and maybe consider this in the broader context of the multiverse. First off, in the Ambien case, I'll probably also remember the Ambien-taking and realize that the walrus wasn't real. My argument is that if I don't remember the Ambien-taking and really think I did talk to a walrus, my brain state has now become indifferentiable from an identical version of me that really did talk to a walrus—and in the infinite multiverse, that version of me must exist. So my consciousness and his are identical, meaning they're the same, meaning that, in a sense, that memory is true.
We can reason the same way about physical laws. A Boltzmann brain that remembers physical laws different from ours is identical to a brain in a universe with those physical laws. And since our brains remember our physical laws, our selves can be said to reside in our universe. Hooray, skepticism averted!
What to do about a Boltzmann brain that remembers meeting or being God? I don't know exactly! I think there's reason to doubt that this is possible—it seems like memories should be restricted to the set "things that exist in the multiverse" and could probably include a plethora of God-ish beings and experiences... but not God Himself unless we take for granted that He exists. In other words, to propose a Boltzmann brain hallucinating God is to beg the question.
(Or am I begging the question in order to refute this challenge by claiming my hypothesis is true? I think not, since the challenge already assumes my hypothesis is true. We can still fidget with God's existence, though, and it seems like doing that materially changes the outcome.)
Edit: To clarify, the parenthetical is responding to the idea that I might be begging the question when I say "memories should be restricted to things that can actually exist." But this claim follows directly from my hypothesis that all memories are somewhere instantiated, so I think it remains ok to assume it's true and shift the onus to the theist's claim.
You say that false memories could be correlated with a consciousness in the multiverse where those memories are true, but it seems like an unjustified assumption that those other universes really exist. The laws of physics might differ in other universes or they might not but that's just speculation. For example, maybe the speed of light is constant in all universes. We've measured it to be 3.0*10^8 m/s. But in a large enough universe there will be BBs or deceived observers who think it is 3.1*10^8 or 5.5*10^15 or a very large (but still finite) number of different values. And only one of these values reflect the real speed of light. The reason we even speculate we might be BBs in the first place is evidence from thermodynamics and cosmology. But if we are BBs, that evidence is most likely incorrect.
Another thing, if talking walruses and meeting God is not possible in any universe, a large enough universe could have computer simulations created somewhere where the inhabitants live in these impossible worlds. But these would be skeptical scenarios. Any consciousness making an observation that is not true in "base reality" is in a skeptical scenario.
The original post seems to be written in the context of the MUH. I have a superficial understanding of the MUH so I'm not sure how it deals with situations like this. Something I have heard is that BB's actually have a higher algorithmic complexity (and thus a lower probability) than a universe and that it's simpler to describe a universe with simple rules that will evolve a brain than it is to describe the brain itself which is actually pretty complex (here is one reference I have for this but it is just a reddit comment https://www.reddit.com/r/askphilosophy/comments/71bq2t/how_accepted_is_tegmarks_mathematical_universe/ Also this article, while not strictly about the MUH, seems to be about a tangentially related theory and dismisses the BB problem because of their high algorithmic complexity https://arxiv.org/pdf/2412.02826).
This comment got quite long. The bottom line is that we either reject BBs because then we can't reliably do science (and thus have no evidence to believe we are BBs) or we develop a metaphysical theory like the MUH or Muller's algorithmic idealism which gives low probability to us being BB's.
“On MUH, this is happening constantly. Your consciousness expands to include any identical Boltzmann brains as they fizz into existence. And then when they fizz out, it snaps back to (what you consider) you. The evolved brain on Earth that’s reading this blog and loving it.”
Won’t this violate the standard No Duplicates condition on psychological continuity theories of personal identity? According to Parfit and his minions, a brain with my quasi-memories appearing *while I’m still alive* won’t cause me to temporarily relocate. If I died a moment before the brain appeared, I would go where the brain goes (assuming there doesn’t need to be the appropriate sort of causal connection) — but in light of the No Duplicates condition, I can’t be taken into some other region of outer space if the brain appears while I’m still alive.
Moreover, the No Duplicates condition seems really hard to dispense with. It’s important that — if a psychological clone of me is made during my lifetime — that it’s not the case that both are identical to me. If both are identical me, we’re going to run into the logical problems that come up in virtue of our different persistence conditions.
Eegh, I'm expressing it poorly I think. In the bit you've quoted, I'm using "consciousness" in the folk, "self"-imbued sense. Let me try to rewrite the claim more precisely:
- You have an evolved brain E.
- There exists some Boltzmann brain B with the same memories as E.
- By my selflessness understanding, we can say that the consciousness arising from E, C_E, is identical to the consciousness arising from B, C_B, so really they're the same consciousness.
- In the next instant, B will disappear.
- So we expect that C_B will disappear.
- That implies that C_E will also disappear in the next instant.
- Which has a seeming contradiction with the fact that E is not about to disappear.
But, on the selflessness understanding of consciousness, that contradiction doesn't exist! Whether a matching B was about to disappear or not, C_E was going to disappear anyway. Because in the next instant, E will have some slightly different set of memories, and so will have a slightly different consciousness—call it C_δE. And our folk understanding of "self"-ness will associate C_δE with C_E because they have so nearly identical memories, and it'll *feel* like the consciousness keeps existing.
I think that oughta dodge most of the duplication issues you're bringing up—in a sense, we *are* all Boltzmann brains, just much more predictably changing ones.
Maybe I'm not getting half of it, but why talk of self, consciousness, and brains as if they're separate things?.. Chemical memory exists in simpler organisms, and it is capable of being trained / retrained. Why not regard those notions (S, C, B) as being facets of physical *interactions* –– albeit complex ones –– occurring on / supported by a particular physical substrate (what we call the physical organ of brain, but also with assistance of the rest of our physiology and surroundings)? Doesn't that more or less resolve the whole thing without needing actual multiple universes? (Modulo the mathematical construct possibly being useful for some sort of prediction or therapeutic purposes, why bother with it?)
It kind of seems like you're asking something equivalent to "why talk of walking, steps, and legs as if they're separate things?" They're certainly all facets of physical interactions, and very closely causally connected, but we experience them in extremely different ways. That's the puzzle of it—why it feels like something to have a brain, a consciousness, a self.
You can say "consciousness arises physically from the brain," but there's still the questions of how, exactly, and why. And then there are more issues with identity—would a physically identical brain, a clone, have the same consciousness? What would that *feel* like? Who would you be? Who are you now? And so on.
It's good to talk about S, C, and B separately because they all seem to relate to each other in different ways. You can have two identical yet distinct brains—can you have two identical yet distinct selves? It sure doesn't seem like it.
I do agree that it’s useful and interesting to perform the analysis of the mechanism and composition of the thing we call consciousness (like the the dynamics and physiological substrate for walking).
What I’d emphasize viz a vis my earlier comment, however, is that all of that reasoning and examination of consciousness can be done adequately (and consequentially / actionably) using purely physical substrates and their interactions, without the need to invoke multiverses, be they physical or mathematical (abstract concepts for the purposes of “calculation”).
That said, I also see some potential value in making these mathematical constructs (MUHs), albeit their utility in this application is lost on me at the moment. For example, in solid state physics, there is the concept of a “hole”, which is kind of like the opposite of an electron, except it makes the math way simpler and allows for accurate predictions of measurable properties (behavior) of a material. Do these “holes” truly exist?.. No. They are mathematical constructs with some physical manifestations we choose to name in a way that becomes a shorthand and refers to familiar objects that share in some of the characteristics with these mental / mathematical constructs to describe the less intuitive physical phenomena. A simpler, more accessible analogy from the world of physics might be how we describe light. Thinking of it as being made up of particles allows us to make great predictions in some instances. They break down in other instances, where thinking of light as a wave makes more accurate predictions. Or the concept of parallel lines — easy to define on a plane, but needing additional rules and/or looking different when doing so on a curved surface…
Forgive me if I'm reading this incorrectly but doesn't your premise that "for every Boltzmann brain B that has the memory M, there is an evolved brain E that actually has done the things that make up M" undermine the argument you make in the post? A BB could observe things that would not be true in the real world because it is most likely just hallucinating a world it thinks is real. For example, it could imagine strong evidence of God such as meeting God or even being God. If this post is arguing in favor of atheism, then the premise that any BB's memory corresponds to an evolved brain undermines atheism. There is a huge range of false memories that a BB could have of fake laws of physics that don't correlate to anything in reality. This is why physicists such as Sean Carroll claim we should reject any theory that leads us to conclude that we are BBs. If we were BBs, we would not be able to reliably trust science (since our "science" would be hallucinated memories) and thus it's ultimately a self-undermining hypothesis. (https://arxiv.org/pdf/1702.00850)
This is an interesting point. I think the argument can maybe be rescued by the idea that even non-Boltzmann brains sometimes have faulty memories. E.g., if I take lots of Ambien, the next morning I'll remember having met a talking walrus (https://www.toothpastefordinner.com/index.php?search=ambien+walrus), even though that isn't true and is impossible.
Of course, that answer seems a little disastrous for me, and probably does lead to total skepticism...
Ok then, let's try again, and maybe consider this in the broader context of the multiverse. First off, in the Ambien case, I'll probably also remember the Ambien-taking and realize that the walrus wasn't real. My argument is that if I don't remember the Ambien-taking and really think I did talk to a walrus, my brain state has now become indifferentiable from an identical version of me that really did talk to a walrus—and in the infinite multiverse, that version of me must exist. So my consciousness and his are identical, meaning they're the same, meaning that, in a sense, that memory is true.
We can reason the same way about physical laws. A Boltzmann brain that remembers physical laws different from ours is identical to a brain in a universe with those physical laws. And since our brains remember our physical laws, our selves can be said to reside in our universe. Hooray, skepticism averted!
What to do about a Boltzmann brain that remembers meeting or being God? I don't know exactly! I think there's reason to doubt that this is possible—it seems like memories should be restricted to the set "things that exist in the multiverse" and could probably include a plethora of God-ish beings and experiences... but not God Himself unless we take for granted that He exists. In other words, to propose a Boltzmann brain hallucinating God is to beg the question.
(Or am I begging the question in order to refute this challenge by claiming my hypothesis is true? I think not, since the challenge already assumes my hypothesis is true. We can still fidget with God's existence, though, and it seems like doing that materially changes the outcome.)
Edit: To clarify, the parenthetical is responding to the idea that I might be begging the question when I say "memories should be restricted to things that can actually exist." But this claim follows directly from my hypothesis that all memories are somewhere instantiated, so I think it remains ok to assume it's true and shift the onus to the theist's claim.
You say that false memories could be correlated with a consciousness in the multiverse where those memories are true, but it seems like an unjustified assumption that those other universes really exist. The laws of physics might differ in other universes or they might not but that's just speculation. For example, maybe the speed of light is constant in all universes. We've measured it to be 3.0*10^8 m/s. But in a large enough universe there will be BBs or deceived observers who think it is 3.1*10^8 or 5.5*10^15 or a very large (but still finite) number of different values. And only one of these values reflect the real speed of light. The reason we even speculate we might be BBs in the first place is evidence from thermodynamics and cosmology. But if we are BBs, that evidence is most likely incorrect.
Another thing, if talking walruses and meeting God is not possible in any universe, a large enough universe could have computer simulations created somewhere where the inhabitants live in these impossible worlds. But these would be skeptical scenarios. Any consciousness making an observation that is not true in "base reality" is in a skeptical scenario.
The original post seems to be written in the context of the MUH. I have a superficial understanding of the MUH so I'm not sure how it deals with situations like this. Something I have heard is that BB's actually have a higher algorithmic complexity (and thus a lower probability) than a universe and that it's simpler to describe a universe with simple rules that will evolve a brain than it is to describe the brain itself which is actually pretty complex (here is one reference I have for this but it is just a reddit comment https://www.reddit.com/r/askphilosophy/comments/71bq2t/how_accepted_is_tegmarks_mathematical_universe/ Also this article, while not strictly about the MUH, seems to be about a tangentially related theory and dismisses the BB problem because of their high algorithmic complexity https://arxiv.org/pdf/2412.02826).
This comment got quite long. The bottom line is that we either reject BBs because then we can't reliably do science (and thus have no evidence to believe we are BBs) or we develop a metaphysical theory like the MUH or Muller's algorithmic idealism which gives low probability to us being BB's.
“On MUH, this is happening constantly. Your consciousness expands to include any identical Boltzmann brains as they fizz into existence. And then when they fizz out, it snaps back to (what you consider) you. The evolved brain on Earth that’s reading this blog and loving it.”
Won’t this violate the standard No Duplicates condition on psychological continuity theories of personal identity? According to Parfit and his minions, a brain with my quasi-memories appearing *while I’m still alive* won’t cause me to temporarily relocate. If I died a moment before the brain appeared, I would go where the brain goes (assuming there doesn’t need to be the appropriate sort of causal connection) — but in light of the No Duplicates condition, I can’t be taken into some other region of outer space if the brain appears while I’m still alive.
Moreover, the No Duplicates condition seems really hard to dispense with. It’s important that — if a psychological clone of me is made during my lifetime — that it’s not the case that both are identical to me. If both are identical me, we’re going to run into the logical problems that come up in virtue of our different persistence conditions.
Eegh, I'm expressing it poorly I think. In the bit you've quoted, I'm using "consciousness" in the folk, "self"-imbued sense. Let me try to rewrite the claim more precisely:
- You have an evolved brain E.
- There exists some Boltzmann brain B with the same memories as E.
- By my selflessness understanding, we can say that the consciousness arising from E, C_E, is identical to the consciousness arising from B, C_B, so really they're the same consciousness.
- In the next instant, B will disappear.
- So we expect that C_B will disappear.
- That implies that C_E will also disappear in the next instant.
- Which has a seeming contradiction with the fact that E is not about to disappear.
But, on the selflessness understanding of consciousness, that contradiction doesn't exist! Whether a matching B was about to disappear or not, C_E was going to disappear anyway. Because in the next instant, E will have some slightly different set of memories, and so will have a slightly different consciousness—call it C_δE. And our folk understanding of "self"-ness will associate C_δE with C_E because they have so nearly identical memories, and it'll *feel* like the consciousness keeps existing.
I think that oughta dodge most of the duplication issues you're bringing up—in a sense, we *are* all Boltzmann brains, just much more predictably changing ones.
Maybe I'm not getting half of it, but why talk of self, consciousness, and brains as if they're separate things?.. Chemical memory exists in simpler organisms, and it is capable of being trained / retrained. Why not regard those notions (S, C, B) as being facets of physical *interactions* –– albeit complex ones –– occurring on / supported by a particular physical substrate (what we call the physical organ of brain, but also with assistance of the rest of our physiology and surroundings)? Doesn't that more or less resolve the whole thing without needing actual multiple universes? (Modulo the mathematical construct possibly being useful for some sort of prediction or therapeutic purposes, why bother with it?)
It kind of seems like you're asking something equivalent to "why talk of walking, steps, and legs as if they're separate things?" They're certainly all facets of physical interactions, and very closely causally connected, but we experience them in extremely different ways. That's the puzzle of it—why it feels like something to have a brain, a consciousness, a self.
You can say "consciousness arises physically from the brain," but there's still the questions of how, exactly, and why. And then there are more issues with identity—would a physically identical brain, a clone, have the same consciousness? What would that *feel* like? Who would you be? Who are you now? And so on.
It's good to talk about S, C, and B separately because they all seem to relate to each other in different ways. You can have two identical yet distinct brains—can you have two identical yet distinct selves? It sure doesn't seem like it.
I do agree that it’s useful and interesting to perform the analysis of the mechanism and composition of the thing we call consciousness (like the the dynamics and physiological substrate for walking).
What I’d emphasize viz a vis my earlier comment, however, is that all of that reasoning and examination of consciousness can be done adequately (and consequentially / actionably) using purely physical substrates and their interactions, without the need to invoke multiverses, be they physical or mathematical (abstract concepts for the purposes of “calculation”).
That said, I also see some potential value in making these mathematical constructs (MUHs), albeit their utility in this application is lost on me at the moment. For example, in solid state physics, there is the concept of a “hole”, which is kind of like the opposite of an electron, except it makes the math way simpler and allows for accurate predictions of measurable properties (behavior) of a material. Do these “holes” truly exist?.. No. They are mathematical constructs with some physical manifestations we choose to name in a way that becomes a shorthand and refers to familiar objects that share in some of the characteristics with these mental / mathematical constructs to describe the less intuitive physical phenomena. A simpler, more accessible analogy from the world of physics might be how we describe light. Thinking of it as being made up of particles allows us to make great predictions in some instances. They break down in other instances, where thinking of light as a wave makes more accurate predictions. Or the concept of parallel lines — easy to define on a plane, but needing additional rules and/or looking different when doing so on a curved surface…