Dear Bruno and Evgenii,
From: Bruno Marchal
Sent: Sunday, January 16, 2011 11:50 AM
Subject: Re: comp and Maxwell demon
You can simulate it. But you cannot program it. It uses a huge amount
of information that it is hard to extract from the environment. I
In a larger sense, in which we are ourself Maxwell demon, you can
filter it from long computation. They will transform selective
information into energy. they will economize.
The Universal Dovetailer does generate all possible Maxwell daemons,
but I guess they are typically distributed sparsely in the set of all
Quantum Maxwell demon? That might explains the origin of life, but
that might be a too cheap explanation. I should revised quantum
information theory, but I have already hardly the time to do the real
thing, which is to extract the quantum information from classical
information + self-duplication and other self-perturbation (or to
convince others to do the work, or that such a work has to be done to
get both the qualia and the quanta).
Interesting but hard question.
On 15 Jan 2011, at 18:22, Evgenii Rudnyi wrote:
> I am trying to understand what the hypothesis comp is about. It
> would be interesting to understand whether the Maxwell demon could
> be simulated or not. Any idea?
Pardon my interjection. How exactly does it make sense to say that we
"transform selective information into energy"? If information is just a form of
energy, such that that phasing can make sense, does it not follow that there
should be a unique relationship between the forms of matter that can implement
a given information structure and a given material structure? From what I can
tell there exists multiple ways that one and the same information can be
implemented in matter. It is also true that multiple different information
structures can represent one and the same material system. This implies two
distinct and irreducible categories: a category of information structures and a
category of material structures. There is also a relationship between these two
categories, but that relation is not an identity mapping in the usual sense
because it is almost always not a bijection. Because of this line of reasoning
I am having a hard time understanding how it is that we can make statements
like "transforming information into energy" and think that those kind of
statements are sound.
OTOH, it make sense to me that we can make claims like: "we can use
information about material systems to induce changes within them that optimize
their energy distributions" and so forth. We can maintain our belief in comp
and UD as they apply to information and its structures, such as Boolean and
Heyting algebras, but still consider that there has to be some form of material
implementation possible at least in principle for it, otherwise we are stuck
with the conclusion that the physical world is just a causally ineffective
epiphenomena and with no explanation as to why we necessarily experience it.
All forms of monism have this problem.
If the distribution of Maxwell daemons is spare in its distribution over
the set of possible computations, does this not imply that the chances of
finding oneself as a Maxwell Daemon to be identically sparse. If the set of
computations is infinite this make the chance of finding oneself as a MD to be
vanishingly small, 1/inf. Does this not seem to indicate against the necessity
of ever experiencing oneself as a MD in the first place? If we go further and
notice that we seem to experience a relatively long chain of experiences that
are MD in nature, does this not make it even more miraculous given derivation
of the statistics of distributions of MD in the ensemble of possible
computations? What if there is a mechanism that allows for the biasing of those
statistics such that it is no longer a miracle that we are not just spontaneous
Boltzmann Brains. The observer selection ideas of Bostrum seem to give us some
good clues about this mechanism. Could it be that a feedback type of
relationship might be involved? The choice of what form the material structure
takes as it evolves in time could be used to select the information
representation that best describes it, but this requires a categorical
difference between matter and information such that we cannot think of them as
just being transformable into each other.
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To post to this group, send email to firstname.lastname@example.org.
To unsubscribe from this group, send email to
For more options, visit this group at