Re: Boltzmann Brains, consciousness and the arrow of time

From: Bruno Marchal <>
Date: Thu, 1 Jan 2009 21:44:27 +0100

On 01 Jan 2009, at 21:10, Brent Meeker wrote:

> Bruno Marchal wrote:
>> It seems to me that your reasoning illustrates well the problems with
>> physical supervenience and physicalism, and perhaps ASSA.
>> In any case the Universal Dovetailer generates all such gaz universes
>> generating the Boltzmann brains. Now the probability that you are
>> implemented by a particular Boltzmann brain is null, as it is null
>> for
>> any particular. With the comp supervenience you have to "attach"
>> consciousness on ALL the histories going through your computational
>> state. It is a sort of double cone of histories.
> Are you assuming time as fundamental here? If time is merely
> inferred then it
> seems that states of Bbs could fit into the inferred time sequence
> as well as
> states that arose in some other way.

I assume only the sequence 0, 1, 2, 3, 4, 5, ... or the axioms of
Robinson arithmetic, or Peano.
This is enough to recognize the working of a universal dovetailer, and
the execution of all programs. It is not infered but postulate. You
can call it a digital time, or you can unravel such a dynamical
deploiment into a statical n n+1 dimensional cone (with n the
dimension of the space used by your starting universal machine (but
some have no concept of dimension, and the statical picture is more a
logical than a geometrical one). It is not "physical time", nor even
the subjective time builded by internal entities.

>> We cannot belong to the aleph_zero Boltzmann brains state, because,
>> from our first person (plural) point of views we already belongs to
>> the 2^aleph_zero "winning" (infinite) histories. (or comp is wrong).
> I don't understand the counting measure. Why are histories order
> 2^apleph_0?

Well, in part this results from the unbounded dumbness of the
universal doevtailing procedure which dovetails on all programs but
also on all non interacting collection of programs (as all interacting
one). In particular each computation is "entangled" to dovetailing on
the reals, and infinite computations are multiplied into 2^aleph_zero
by this entanglement with the reals. Now this is a good thing because
it means that the stable histories will be those who manage that
background noise, who exploits it probably.

Our mind states are enumerable, but our histories are not.


> Brent
>> This is a case for RSSA indeed. I think. Bb are reduced to the
>> "usual"
>> white rabbits histories, with RSSA, it seems to me.
>> Bruno
>> On 31 Dec 2008, at 22:58, Hal Finney wrote:
>>> Sometimes we consider here the nature of consciousness, whether
>>> observer
>>> moments need to be linked to one another, the role of causality in
>>> consciousness, etc. I thought of an interesting puzzle about
>>> Boltzmann
>>> Brains which offers a new twist to these questions.
>>> As most readers are aware, Boltzmann Brains relate to an idea of
>>> Boltzmann
>>> on how to explain the arrow of time. The laws of physics seem to be
>>> time
>>> symmetric, yet the universe is grossly asymmetric in time.
>>> Boltzmann
>>> proposed that if you had a universe in a maximum entropy state,
>>> say a
>>> uniform gas, then given enough time, the gas would undergo
>>> fluctuations
>>> to regions of lower entropy. Sometimes, purely at random, clumps of
>>> molecules would happen to form. Even more rarely, these clumps might
>>> be
>>> large and ordered. Given infinite time, one could even have an
>>> entire
>>> visible-universe worth of matter clump together in an ordered
>>> fashion,
>>> from which state it would then decay into higher entropy conditions.
>>> Life
>>> could evolve during this decay, observe the universe around it, and
>>> find
>>> itself in conditions much like our own.
>>> The Boltzmann Brain is a counter-argument, suggesting that the
>>> universe
>>> and everything else is redundant; all you need is a brain to form
>>> via
>>> a spontaneous random fluctuation, and to hold together long enough
>>> to
>>> engage in a few moments of conscious thought. Such a Boltzmann Brain
>>> is
>>> far more likely to form than an entire universe, hence the vast
>>> majority
>>> of conscious thoughts in such a model will be in Boltzmann Brains
>>> and not
>>> in brains in large universes. If we were tempted to explain the
>>> arrow of
>>> time in this way, we must accept that the universe is an illusion
>>> and
>>> that we are actually Boltzmann Brains, a conclusion which most
>>> people
>>> don't like.
>>> Now this scenario can be criticized in many ways, but I want to
>>> emphasize
>>> a couple of points which aren't always appreciated. The first is
>>> that the
>>> Boltzmann scenario, whether a whole universe or just a Brain is
>>> forming,
>>> is basically time symmetric. That means that if you saw a movie of a
>>> Boltzmann universe forming and then decaying back to random entropy,
>>> you would not be able to tell which way the movie was running, if it
>>> were to be reversed. (This is an unavoidable consequence of the time
>>> symmetry of the underlying physics.) It follows that while the
>>> universe
>>> is moving into the low-entropy state, it must be evolving backwards.
>>> That
>>> is, an observer from outside would see time appearing to run
>>> backwards.
>>> Eggs would un-scramble themselves, objects would fall upwards from
>>> the
>>> ground, ripples would converge on spots in lakes from which rocks
>>> would
>>> then leap from the water, and so on.
>>> At some point this time reversal effect would stop, and the universe
>>> would then proceed to evolve back into a high entropy state, now
>>> with time
>>> going "forwards". Now, the forward phase will not in general be an
>>> exact
>>> mirror image of the reverse, because of slight random fluctuations
>>> and
>>> the like, but it will be an alternate path that essentially starts
>>> with
>>> the same initial conditions. So we will see one path backwards into
>>> the
>>> minimum-entropy state, and another path forwards from that state.
>>> Both
>>> paths are fully plausible histories and neither is distinguishable
>>> from
>>> the other as far as which was reversed and which was forward, if you
>>> ran a recording of the whole process backwards.
>>> One might ask, what causes time to run backwards during the first
>>> half of
>>> the Boltzmann scenario? The answer is, nothing but very, very odd
>>> luck.
>>> Time is no more likely to continue to run backwards, or to run
>>> backwards
>>> the same everywhere in the local fluctuation-area, than it is to
>>> start
>>> running backwards right now in the universe around you. Nothing
>>> stops
>>> eggs from unscrambling themselves except the unlikelihood, and the
>>> same
>>> principle is at work during the Boltzmann time-reversal phase. It is
>>> merely that we select, out of the infinity of time, those rare
>>> occasions
>>> where time does in fact "happen to happen" like this, that allows us
>>> to
>>> discuss it.
>>> I want to emphasize that this picture of how Boltzmann fluctuations
>>> would
>>> work is a consquence of the laws of thermodynamics, and time
>>> symmetry.
>>> Sometimes people imagine that the fluctuation into the Boltzmann
>>> low-entropy state is fundamentally different from the fluctuation
>>> out
>>> of it. They accept that the fluctuation out will be similar to our
>>> own
>>> existence, with complex events happening. But they imagine that the
>>> fluctuation into low entropy might be much simpler, molecules simply
>>> aggregating together into some convenient state from which the
>>> complex
>>> fluctuation out and back to chaos can begin. While this is not
>>> impossible
>>> and hence will happen occasionally among the infinity of
>>> fluctuations in
>>> the Boltzmann universe, it will be rare. It will be no more common
>>> for a
>>> "simple" fluctation-in process to occur than for a simple
>>> fluctuation-out
>>> process. In our universe, knowing it will evolve to a chaotic heat
>>> death, we might imagine that molecules would just fly apart into
>>> chaos,
>>> but we know that is highly unlikely. Instead, by far the most likely
>>> path is a complex one, full of turbulence and reactions and similar
>>> activity. By time symmetry, exactly the same arguments apply during
>>> the fluctation-in phase. The vast majority of Boltzmann fluctuations
>>> that achieve a particular degree of low entropy will do so via
>>> complex,
>>> turbulent paths which if viewed in reverse will appear to be
>>> perfectly
>>> plausible sequences of events for a universe which is decaying from
>>> order to disorder, like our own.
>>> Following on to this, let us consider the nature of consciousness
>>> during
>>> these Boltzmann excursions. Again let us focus on larger scale ones
>>> than
>>> just Boltzmann Brains, although the same principles apply there.
>>> During
>>> the time reversal phase, if conscious entities are present, their
>>> brains
>>> are running backwards. They are talking backwards, walking
>>> backwards,
>>> doing everything in reverse. They remember things that are coming in
>>> the future, and forget everything as soon as it has happened.
>>> The question is, is there any difference in consciousness during the
>>> reverse and forward phases? Consider that during the forward
>>> phase, we
>>> started with a low entropy state, and now the laws of physics are
>>> playing
>>> out just as they do in our own universe. Everything is happening
>>> for a
>>> reason, depending on what has happened before. Events cause memories
>>> to
>>> appear in brains by virtue of the same causal effects which give
>>> rise
>>> to our own memories. Hence I imagine that most would agree that
>>> brains
>>> during the forward phase are conscious.
>>> However, during the reverse phase, things are quite different.
>>> Brains
>>> have memories of things that haven't happened yet. Again, one might
>>> ask how this can be. The reason is because we stop paying attention
>>> to fluctuations where this doesn't happen. We only focus on
>>> Boltzmann
>>> fluctuations which take the universe into a plausible and consistent
>>> low-entropy state, one from which things can evolve in a way that is
>>> similar to what we see. When a brain remembers something, if that
>>> doesn't
>>> happen, the fluctuation is inconsistent. We skip over that one and
>>> look
>>> for one that is consistent.
>>> In the consistent fluctuations, brain memories turn out to be
>>> correct,
>>> purely by luck. Similarly, every internal function of the brain
>>> which
>>> we might attribute to macroscopic-type causality, like neuron A
>>> firing
>>> because neuron B fired, will happen instead by luck, with neuron A
>>> firing
>>> as though neuron B is going to fire, and then neuron B just
>>> happening
>>> to fire in precisely the anticipated way.
>>> The point is that during the time-reversal phase, causality as we
>>> normally think of it is absent. Subjectively-past events do not
>>> cause
>>> subjectively-future ones; rather, subjectively-future events take
>>> place
>>> before subjectively-past events, and it is merely through luck that
>>> things
>>> happen in a consistent pattern. Again, if we hadn't gotten lucky so
>>> that
>>> things work out, we wouldn't have called this a Boltzmann
>>> fluctuation of
>>> the kind we are interested in (Boltzmann Brain or Boltzmann
>>> Universe).
>>> By paying selective attention to only those fluctuations where
>>> things
>>> work, we will only observe cases where luck, rather than causality,
>>> makes things happen.
>>> But things do happen, in the same pattern they would if causality
>>> were
>>> active. So the question is, are brains conscious during this
>>> time? Do
>>> the thoughts that occur during the time reversal (which recall is
>>> not
>>> exactly the same as what happens during the forward-time phase) have
>>> the same level of subjective reality as thoughts which occur when
>>> time
>>> runs forward?
>>> We can argue it either way. In favor of consciousness, the main
>>> argument is that time is fundamentally symmetric (we assume). Hence
>>> there is no fundamental or inherent difference between the forward
>>> and
>>> reverse phases. The only differences are relative, with the arrow
>>> of
>>> time pointing in opposite directions in the two phases. But within
>>> each
>>> phase, we see events which can both be equally well described as
>>> leading
>>> to consciousness, and therefore conscious experiences will occur in
>>> both phases.
>>> On the other side, many people see a role for causality in the
>>> creation
>>> or manifestation of consciousness. And arguably, causality is
>>> different
>>> in the two phases. In the forward phase (the part where we are
>>> returning
>>> from a low-entropy excursion to the high-entropy static state),
>>> events
>>> follow one another for the usual reasons, and it is correct to
>>> attribute
>>> a role for causality just as we do in our own experience. But in the
>>> reverse phase, it is purely by luck that things happen in a
>>> consistent
>>> way, and only because we have an infinity of time to work with
>>> that we
>>> are able to find sequences of events that look consistent even they
>>> arose
>>> by simple happenstance. There is no true causality in this phase,
>>> just a
>>> random sequence of events where we have selected a sequence that
>>> mimics
>>> causality. And to the extent that consciousness depends on
>>> causality,
>>> we should not say that brains during this reverse phase are
>>> conscious.
>>> I lean towards the first interpretation, for the following reason.
>>> If
>>> consciousness really was able to somehow distinguish the forward
>>> from
>>> reverse phases in a Boltzmann fluctuation, it would be quite
>>> remarkable.
>>> Given that the fundamental laws of physics are time symmetric,
>>> nothing
>>> should be able to do that, to deduce a true "implicit" arrow of time
>>> that
>>> goes beyond the superficial arrow of time caused by entropy
>>> differences.
>>> The whole point of time symmetry, the very definition, is that there
>>> should be no such implicit arrow of time. This suggestion would
>>> seem
>>> to give consciousness a power that it should not have, allow it to
>>> do
>>> something that is impossible.
>>> And if the first interpretation is correct, it seems to call into
>>> question
>>> the very nature of causality, and its posible role in consciousness.
>>> If
>>> we are forced to attribute consciousness to sequences of events
>>> which
>>> occur purely by luck, then causality can't play a significant role.
>>> This
>>> is the rather surprising conclusion which I reached from these
>>> musings
>>> on Boltzmann Brains.
>>> Hal Finney
> >

You received this message because you are subscribed to the Google Groups "Everything List" group.
To post to this group, send email to
To unsubscribe from this group, send email to
For more options, visit this group at
Received on Thu Jan 01 2009 - 15:44:42 PST

This archive was generated by hypermail 2.3.0 : Fri Feb 16 2018 - 13:20:15 PST