God vs The Multiverse

Click here for God vs The Multiverse: a rational argument for the Existence of One God who intelligently designed one universe.

Sunday, August 5, 2012

God vs The Multiverse (Part 14: Boltzmann Brain Paradox)

In this post, we will discuss one of the serious problems which multiverse theory encounters by positing that our universe emerged from a random fluctuation out of chaos.  (Click here for the table of contents to the entire series - God vs The Multiversea rational argument for the Existence of One God who intelligently designed one universe.)

The Boltzmann Brain Paradox is one of the most significant (and humorous) points in this series.  It is expressed in a serious form by Richard Feynman, who is quoted by Sean Carroll in the embedded video below.  He explains very clearly why the entropy problem of post 4 is such a fundamental problem for multiverse theory.  If you find this post difficult to follow, try watching the very clear video (about 15 minutes), which will help you understand the main concept.


The Boltzmann Brain problem arises from taking the multiverse theory seriously and following it through to its logical conclusion.  Multiverse theory claims that the very low entropy (highly improbably) initial conditions for the big bang was the result of a random fluctuation in conjunction with an infinite number of tries.

The problem starts with the fact that, based upon entropy, it is much more likely (by a factor of  1010100 ) that a random fluctuation would have produced the entire universe as it is right now, than to have produced the initial conditions of the big bang.  This would mean that all our memories are mistaken as we think we actually remember the world five minutes ago looking very similar to the way it does now.

Further, it is still more likely that the random fluctuation should have produced a universe which is a combination of chaos and order, as opposed to a universe that is entirely ordered.  Recall that according to multiverse theory, the order which we observe in our universe is not a rule, but a random coincidence. Therefore, taking multiverse theory to its logical conclusion actually leads to a prediction that we will observe chaos anywhere new that we look in the cosmos  (i.e., places we have not observed yet).

This prediction of multiverse theory could be simply tested by looking (with the newest telescopes) in new directions or farther distances than we have previously observed.  An amazing thing happens when we do this.  We always find more galaxies and order, not chaos.  Multiverse theory makes false predictions over and over.  When it is actually subjected to the scientific method, it is refuted.  (You can watch Feynman himself explain this subtle argument in this short video, where he calls the multiverse theory ridiculous.)

Following the above logic even further, it is even more likely that the random fluctuation should have produced one single brain, rather than a universe of at least 100 billion galaxies that we think we live in.  This is because the most probable fluctuation that explains your current thoughts, memories, emotions, etc., is a Boltzmann Brain (i.e., a brain whose neurons are randomly configured exactly as your brain is right now) surrounded by chaos.

The technical term in cosmology for these Boltzmann Brains is "freaky observers", in contrast to "normal observers" (which is what you probably think you are).  This hilarious-because-it's-true New York Times article entitled Big Brain Theory: Have Cosmologists Lost Their Minds? does an excellent job of elucidating these points, and shows scientists' responses to the problem.  The last two paragraphs are our favorite part.

There are two ways that physicists try to solve this problem.  The first way is to try to invoke eternal inflation as a solution.  If you recall, Steinhardt's main critique on eternal inflation was that it introduces ad hoc measures to try to allow the model to make any predictions at all.  This is because there are an infinite number of every possible universe, which makes it impossible to figure out which type of universe is more probable in any straightforward way.

A good example of one of these ad hoc measures is based on Boltzmann Brains.  In the introduction to their paper on Boltzmann BrainsAndrei Linde and his colleagues address this issue.  (This is a technical paper, but if you can get the main idea of the introduction, especially from the last paragraph of page 1 and on, you can appreciate this point in a fuller way.)  They argue that since the prediction that we are Boltzmann Brains is proven false by our continued observation of order, we must introduce an artificial measure into eternal inflation which makes the number of ordered observers more likely than freaky observers.

The authors try to show that it is possible to twist the theory of eternal inflation by using a hand-picked measure (see previous postso that you are more likely to be an ordered observer.  They are claiming that with this new measure, it may be more probable that a random fluctuation in the infinite chaos will produce the initial conditions of our big bang, that leads to a universe of 100 billion galaxies with 100 billion stars each, rather than a brain surrounded by chaos.  (We call this the magic eternal inflation theory.)

Besides for how obviously contrived this ad hoc measure is, there is a more basic flaw in this reasoning.  He can not simply disprove the possibility that he is a Linde Brain (a brain which randomly fluctuated into existence believing it is Andrei Linde) by walking outside and observing continued order.  The logic of an ancient skeptical philosophy called "brain in a vat" undermines this simple experiment.  Rather, it is still far more likely that he is another Linde Brain that just fluctuated into existence out of the infinite chaos, with the false memory that it just did an experiment to test out whether it really was a Linde Brain or not.  Consequently, Linde and colleagues have no justification for using this measure even in an ad hoc manner, and should rather reject multiverse theory because it leads to the conclusion of radical skepticism.

This naturally brings us to the second tactic we have seen some physicists (Greene and Carroll) use to avoid the Boltzmann Brain Paradox.  Essentially, the argument is as follows:  When we take multiverse theory to its logical conclusion (that we are probably Boltzmann Brains), we can no longer trust our memories. This ends up undermining the very laws of physics that made us suggest a multiverse in the first place.  Secondly, we simply can not operate and function with the belief that we are Boltzmann Brains.  Therefore, it makes no sense to take the theory to its logical conclusion as it undermines itself and leads to absurdities.  The only thing that we can reasonably believe is that we actually live in a multiverse where random fluctuations happened to produce the very unlikely, highly ordered state of the big bang that has truly evolved to this point in time.  We thereby save the laws of physics and rescue multiverse theory!

This argument is trying to avoid the logical fallacy of having a conclusion that contradicts a premise, and of being reductio ad absurdum (the belief that you are a Boltzmann Brain), by not taking the theory to its logical conclusion.  But there is simply no justification for stopping the argument short. Any theory that when taken to its logical conclusion undermines itself, must be rejected.  You can not just say that we shouldn't take it to its logical end.  If a theory leads to an absurd result or a conclusion which contradicts a premise, there is something very wrong with the theory!

Someone may dismiss this entire problem of Boltzmann Brains by saying that freaky observers are obviously ridiculous, and are not part of science.  Read the Times article.  We're not making this problem up.  This paradox emerges when someone denies the real order in the universe, and exchanges it for random chaos which improbably results in an illusion of order.  Once we consider such far out possibilities, we have no way of excluding freaky possibilities such as Boltzmann Brains.  (Hence the vast literature on the subject of Boltzmann Brains, as quoted in the introduction of the Linde paper cited above).  The serious consideration of ridiculous Boltzmann Brains emerges from considering the even more ridiculous explanation of the big bang put forth by multiverse theorists.

An honest, logical multiverse theorist must either admit that multiverse theory is wrong, or accept the unfortunate reality that they are most likely Boltzmann Brains. 

17 comments:

  1. This is an interesting and enjoyable topic, thanks for bringing it up.

    I have to admit a lot of ignorance of the underlying physics, but a tentative answer might lie in this direction:

    As we are in a highly ordered space-time region of the universe (still close to the Big bang), the probability is much higher that the said Brain arises through some ordered process (say evolution) than randomly through fluctuation. This makes sense intuitively to me, though I can't give a formal argument ATM.

    The real question is what happens as we move into higher entropy, and little or no ordered processes can exist. It does seem that at that point random-fluctuation-caused Brain is more likely that ordered-process caused one. If this regime truly laster forever, the overall count of Boltzman Brains would overwhelm that of evolution-Brains. Notably, Caroll did not actually say "forever" but gave a specific number. One possibility is that perhaps probability of getting a working brain is really small enough that it wouldn't happen as often in that part of space-time as in the ordered part.

    Let me also point out that the Boltzmann Brain "objection" seems to apply to a single universe as well. If there are many universes that start from a kind of a Big Bang area of low entropy, they will produce overall probability distribution of Boltzmann brains same as a single universe, by multiplication and renormalization.

    Dr_Manhattan

    ReplyDelete
    Replies
    1. We think you are making a mistake about what is the real question that is relevant to this proof.

      It is not relevant if our best speculations (based on our current knowledge of physics and cosmology), about the end of our universe lead to the prediction that you are a Boltzmann Brain. That merely show that there is something wrong with our current theories and models.

      What is very relevant for the proof is the beginning of our universe. What ordered the initial conditions for the universe? According to a honest multiverse theory (where the initial conditions were randomly set), you should draw the logical conclusion that you are most likely a Boltzmann Brain.

      Hence, if you still honestly believe that the multiverse theory is 90% likely to be true, then you should also consistently believe that you are approximately 89% likely to be a Boltzmann Brain.

      To address your point more directly:

      >>As we are in a highly ordered space-time region of the universe (still close to the Big bang), the probability is much higher that the said Brain arises through some ordered process (say evolution) than randomly through fluctuation. This makes sense intuitively to me, though I can't give a formal argument ATM.

      The "ordered process" must follow the second law of thermodynamics and thereby bring the universe from a state which is more ordered (lower entropy, less probable) to a state which is less ordered (higher entropy, more probable). Thus, in order for this ordered process to occur, it had to start with an even more highly ordered state. How did this state come about? If through an Intelligent Designer, fine. If through a fluctuation, than the brain is a more likely fluctuation and is probably what we are.

      Delete
    2. > It is not relevant if our best speculations (based on our current knowledge of physics and cosmology), about the end of our universe lead to the prediction that you are a Boltzmann Brain. That merely show that there is something wrong with our current theories and models.

      It is relevant, indirectly. The alternative resolution is that there is another way to dissolve the Boltzmann Brain question, which would apply in both cases.

      > What is very relevant for the proof is the beginning of our universe. What ordered the initial conditions for the universe? According to a honest multiverse theory (where the initial conditions were randomly set), you should draw the logical conclusion that you are most likely a Boltzmann Brain.

      Not sure how this follows. You spent a lot of time explaining how very precise conditions are necessary for atoms not to fly apart. Wouldn't the supposed Boltzman Brain need at least a few seconds of existence (even with false memories) in order to have any thought process at all? The overwhelming vast majority of the multiverses would not allow this it seems.

      > The "ordered process" must follow the second law of thermodynamics and thereby bring the universe from a state which is more ordered (lower entropy, less probable) to a state which is less ordered (higher entropy, more probable). Thus, in order for this ordered process to occur, it had to start with an even more highly ordered state. How did this state come about? If through an Intelligent Designer, fine. If through a fluctuation, than the brain is a more likely fluctuation and is probably what we are.

      I'm missing something. Are you saying that a functioning (if only for a few split seconds) brain can fluctuate into existence without the context of a universe of any kind at all? What constants rule its physical operation?

      Dr_Manhattan

      Delete
    3. > It is relevant, indirectly. The alternative resolution is that there is another way to dissolve the Boltzmann Brain question, which would apply in both cases.

      It does not help one theory with a fatal flaw, if another theory also has the same flaw. If you can not resolve the problem, both theories are wrong.

      > Wouldn't the supposed Boltzman Brain need at least a few seconds of existence (even with false memories) in order to have any thought process at all? The overwhelming vast majority of the multiverses would not allow this it seems.

      It doesn't matter what percentage of multiverses have Boltzmann Brains. It could be they are quite rare, as it is very unlikely for a random fluctuation to produce even one BB.

      However, the big bang that produced the entire universe is even less likely than a BB. This is a fatal flaw in multiverse theory, because of the relative probabilities between a BB and the the entire universe. The key concept here is relative probabilities.

      > I'm missing something. Are you saying that a functioning (if only for a few split seconds) brain can fluctuate into existence without the context of a universe of any kind at all? What constants rule its physical operation?

      The relative probabilities point also explains your constants question. Even if you required the same constants for a BB as the universe, the universe is still a less likely fluctuation.

      Hence, if you want to posit that a random fluctuation produces a highly ordered the big bang, then you have to logically maintain that you are most likely a Boltzmann Brain.

      Delete
    4. > Hence, if you want to posit that a random fluctuation produces a highly ordered the big bang, then you have to logically maintain that you are most likely a Boltzmann Brain.

      I don't think you're right about this: the probabilities are conjunctive. I'll introduce a bit of terminlology to make this clear: instead of Boltzmann Brains let's talk about Boltzmann Thoughts. The reason for this is that it emphasizes the existence of a process of observation (requiring the brain to survive for at least a short while). Boltzman Brains that immediately dissolve are not an interesting phenomena (they can have as much thought as a dead brain essentially). Ok, so Boltzmann Thoughts require a universe with constants suitable for their (BTs) existence, however brief. The probability of these phenomena is smaller than the probability of BT-suitable universes, because BTs require the universes to exist.

      Dr_Manhattan

      Delete
    5. The problem is essentially unchanged if you want to have the Boltzmann Brain last for 5 seconds. (In truth, you are wrong about even this, as all you really need for the BB problem is an instantaneous thought that includes fake memories of the past 5 seconds. But we will grant you this one concession and show you how you are still wrong.)

      For example, you could be a fluctuation of a brain in a vat that lasts for 5 minutes. If you prefer, you can be a fluctuation of a person in a room filled with oxygen and a deck of cards (and a glass of water if your thirsty), surrounded by chaos outside. That person will last for a while until he dies as a result of the second law of thermodynamics.

      Either scenario is far more likely that a fluctuation that gives rise to the big bang which produces a universe of 100 billion galaxies. Boltzmann Thoughts as opposed to Boltzmann Brains do not solve this most pernicious problem.

      Delete
    6. I dont get your point. Why do BT's require the universe to exist? If there were just a brain existing, it could have a thought without the rest of the universe...?

      Delete
    7. > The problem is essentially unchanged if you want to have the Boltzmann Brain last for 5 seconds. (In truth, you are wrong about even this, as all you really need for the BB problem is an instantaneous thought that includes fake memories of the past 5 seconds. But we will grant you this one concession and show you how you are still wrong.)

      Do you need a thank you note for that generous consession?

      The concession is not needed, unless you want to claim that a frozen brain without any electrical signals is in fact experiencing a thought. It seems very unintuitive to me that it is (even though I will admit much confusion about consciousness). http://hypertextbook.com/facts/2002/DavidParizh.shtml

      > For example, you could be a fluctuation of a brain in a vat that lasts for 5 minutes. If you prefer, you can be a fluctuation of a person in a room filled with oxygen and a deck of cards (and a glass of water if your thirsty), surrounded by chaos outside. That person will last for a while until he dies as a result of the second law of thermodynamics.

      Very nice, but not in the vast majority of the universes where the room will dissolve in a picosecond.

      > I dont get your point. Why do BT's require the universe to exist? If there were just a brain existing, it could have a thought without the rest of the universe...?

      David, I'm making the argument that the brain could not have thought up the rest of the universe if it dissipated in a picosecond. It requires a universe with somewhat stable condictions to exist. The RAZ/REF spent a lot of time explaining that very precise conditions are necessary for life to survive (though the conditions I require here are less strict, e.g. formation of stars might not be necessary).

      Dr_Manhattan

      Delete
    8. Dr_Manhattan, you're missing the point while you attack the particulars. Whatever necessary conditions you feel there are for a single consciousness with false memories to exist, it is inherently less necessary conditions than to create billions of conscious observers with an actual universe to observe.

      If you assume an underlying chaos to the world then the least entropic conclusion makes the most sense. In this case a random configuration of a brain and the necessary elements to support it for 5 seconds (or really even 100 years). This configuration is more likely than a random configuration of the entire observed universe.

      From the sound of your answers, though, I think you are inconsistently falling back on scientific and rational reasoning. you seem to be trying to say that it is more likely to perceive a universe with laws in place than to randomly create a fake observer. Unfortunately, if you make use of this principle, the logical conclusion is that one must assume an underlying order over an underlying chaos which all but destroys the multi-verse theory.

      Further If all of our perceptions, scientific findings, and calculations are rooted in a Boltzmann Brain, the requirements for life could be anything. They could be more, they could be less. Whatever requirements you think there are have no root in "actual reality". They may be even less likely than we currently perceive them to. Regardless of what they "actually" are, it will ALWAYS be more likely in chaos for them to occur once rather than to create a physical law of some pocket universe within the multiverse.

      Delete
    9. Here's another way to see the point.
      Consider the subset of the multiverse which consists of all universes with exactly our constants (an infinite set). In all of these universes, a "Boltzmann Room" will be stable (won't dissolve in a picosecond and the Boltzmann person in the room will last as long as his oxygen supply).

      The probability of one of these universes to have our big bang initial conditions is 1/(10^(10^123)). Thus, for every "stable" universe with our order, there are roughly 10^(10^123) which do not have this order.

      Following this and the step-by-step argument of the post, it is more likely that we are in a Boltzmann room universe than in the real universe.

      Delete
    10. The below is confusing, since it seems to be making two contradictory arguments.

      >Consider the subset of the multiverse which consists of all universes with exactly our constants (an infinite set). In all of these universes, a "Boltzmann Room" will be stable (won't dissolve in a picosecond and the Boltzmann person in the room will last as long as his oxygen supply).

      This seems to be saying that the power of BB argument comes from the plethora of universes similar to ours. (Which happens to produce the same BB distribution as our single universe, raising the same questions).

      The below
      > The probability of one of these universes to have our big bang initial conditions is 1/(10^(10^123)). Thus, for every "stable" universe with our order, there are roughly 10^(10^123) which do not have this order.

      Seems to imply that supermajority of BBs comes from universes rather dissimilar from ours, which I addressed with the thought-sustainability requirement.

      Which one is it?

      Dr_Manhattan

      Delete
    11. Reread posts 2-3 vs. post 4. There are 2 separate instances of fine tuning in our universe:
      (a)the constants (posts 2-3): These are responsible, among other things, for the stability produced by the laws of our universe;
      (b)the initial conditions (post 4): even if the laws/constants are given, if the initial distribution of matter in our universe wasn't so ordered (i.e.,low entropy), we would still have chaos in our universe.

      Thus, the plethora of universes similar to ours are similar in the constants. But the supermajority of BBs comes from universes rather dissimilar from ours regarding the initial conditions. Though they are dissimilar in the initial conditions, the stability of our Boltzmann rooms is maintained because the laws and constants are the same.

      The Boltzmann brain paradox emerges from the multiverse's attempt to explain the fine tuning of the initial conditions (post 4) and has nothing to do with the fine tuning of the constants(posts 2-3).

      Delete
  2. You make a good point about the unlikelihood of a random fluctuation creating such a low entropy as to produce our universe, as opposed to the greater likelihood of a partially organized universe or just a Boltzmann brain. What I am not sure about is whether a multiverse might arise from a different mechanism than a random fluctuation. So it is the random fluctuation idea that is called into question, but not necessarily the idea of a multiverse produced by some other means.

    ReplyDelete
    Replies
    1. The essence of multiverse theory is that randomness in conjunction with an infinite number of tries can create the illusion of order and design. (The belief in multiple universes that are ALL ordered is not how we are using the term "multiverse theory", and has no relevance to the fine tuning of the constants or initial conditions.)

      Almost all multiverse models (except for Smolin's cosmological natural selection model which has its own problems) explain the fine tuned constants by saying that there are an infinite number of parallel universes, which have never been observed, where the constants are randomly selected by some random number generator that yields mostly barren universes.

      Likewise, all multiverse models have to explain the low entropy conditions by saying that there are an infinite number of alternate universes which have high entropy initial conditions. Whether those universes come from fluctuations or not is irrelevant. Random fluctuations is just one mechanism. It is akin to the speculated random number generator that multiverse theory has to posit which randomly selects the constants. It could be any one of many imagined mechanisms. The essential problems with multiverse theory have nothing to do with the particular mechanisms that choose the constants or the initial conditions.

      What matters for the Boltzmann Brain Paradox is that the universes that have only freaky observers will far outnumber the universes that have ordered observers. The specific mechanism for creating the different universes has no impact on this problem.

      Delete
  3. I wonder why modern day science is so fixtated on trying to remove purpose out of nature, and it seems like they are willing to propose theories that stretch credulity beyond the breaking point. You also see this in biology (neodarwinism). The digital code in the cell is a problem for materialist theories for the orgin of life. Common sense would tell you that a code require some sort of intentional conscious mind, but yet people run around in circles trying to avoid this problem. When you see this a type of behavior, you know that you are dealing with people who have some sort of a-prior emotional commitment to some sort of belief system. This just doesn't apply here, I have seen this with religious belief, political belief etc. I think my problem is the alot of people view scientists as some sort of ultral rational super humans who can rise above all belief systems etc.. if most scientist believe it than it most be correct... that they are just following the evidence where it leads.. but this is not always the case..

    ReplyDelete
  4. The issue of BB is resolved:
    http://gravityattraction.wordpress.com/2013/04/01/boltzmann-brain-discovery/

    ReplyDelete
    Replies
    1. yet another pillar of support for the multiverse

      Delete

In the words of Agur bin-Yakeh: "We welcome all comments, questions, contributions, and critiques - but if you insist on posting anonymously, PLEASE use a pseudonym rather than posting as "Anonymous," since this makes it much easier to carry on a normal discussion. Thank you!"