(This page is still a draft, gathering 2 separately written versions of the same topic, which will be cleaned up later)

The entropy creation process can be roughly described as a process of evolution from a system with a number N of possible states, into a system with a much larger number N' of possible states, where the N states evolve into the space of N' states with near certainty, and are "dissolved" in it (distinctions with the other states there are no more accessible to observation in practice).

The very fact of assuming the initial state as given, while the final state is not determined yet, is an assumption that treats time asymmetrically. This assumption can be understood as based on the metaphysics of time, and more precisely the time of consciousness.

This would mean that consciousness was already there and observed the universe since its beginning (the big bang).

However, this reasoning is not very rigorous (as it is quite metaphysical...), so its status as a "proof" remains subject to personal opinion.

For more details, see: The Mind makes collapse interpretation - A mind/mathematics dualistic foundation of physical reality

An alternative candidate explanation, is to consider the state of the big bang. Let us discuss this in more details.

- The flow of entropy creation from a human body (or any animal), is very much bigger than the flow of sensorial perceptions in the same body, making it much too unlikely for entropy to be reduced, or even to not be created "as it should", at the macroscopic scale.
- With the expansion of the universe, we can consider that the increase of available space for every quantity of matter, provides it manier possible states for its evolution, thus more opportunities for entropy creation. Thus we have 3 growing quantities, from the biggest to the smallest (in each big volume of the universe):

- the "possible entropy" (log of the total number of states into which matter might potentially evolve);
- the entropy as it is (the "measure of our ignorance" on the state of the universe);
- the total information that has been received from concious perceptions of matter (the "measure of our knowledge" on the universe, even if we forgot most of it).

But what we take the final observable system of N' states and compute the inverse evolution ? Then, for a reason of numbers, the probability to get back among the initial N states is extremely small. A still larger set of more than N' possibilities is needed, including the N first ones.

But then, how can we know from the present state in the N' set, that the past state of the system was more probably among the first N than among the others that the theory "retro-predicts" from the present observed state ?

We could say: because if it was others, then it would probably not have evolved into what it is now.

But this argument requires to assume that, inside this larger set, the previous state of the system was only possibly oriented towards "previously observable" directions, and did not have a "hidden purity" precisely designed to make it much more probably evolve towards what it is now.

These considerations seem to confirm that the reasoning which explains the thermodynamic time orientation (irreversible entropy creation), requires the assumption that the probability laws of the present state of the universe is precisely what is given by the formulas of the theory as determined by past observations.

- The universe there is quite more irregular, with diverse parts collapsing (to the past) in separate black holes rather than converging to a single origin; as opposed to the remarkable homogeneity of the real big bang where distances between future galaxies are arbitrarily small near the birth of the universe.
- At many places, there is more energy for every material particle at every time defined by a specified density, which explains the higher number of available states (we said that the number of possible states of a system depends both on the available space and available energy, so more states with the same space, takes its room by involving higher energies)

Let us explain something well-known by cosmologists that naively looks very paradoxical : in usual life, states of uniform perfect mixtures are usually those of maximal entropy, while the structured, non-uniform states (with low density at some places and high density at other places) have lower entropy. But with the Big Bang, things go the other way round. Why ? Because in usual life, states of equilibrium that maximize the entropy come in a context of a limited amount of energy that is conserved, so the maximal entropy is achieved by uniformly distributing the energy.

But this law only appears as a "constant energy" in the case of an isolated systems with a fixed volume, which is not the case of the expanding universe: in the big bang it does not work like this because the energy is not conserved during the universal expansion. Here in the present universe, matter emits radiation (above the temperature of the cosmic microwave background) from nuclear energy coming from the fact that the energy of the Big Bang had largely separated protons and neutrons, putting them in the form of hydrogen and helium that can make nuclear reactions in stars. A reversed evolution from this would emit light that would increase its energy by blueshift (inverse of the cosmological redshift), giving the involved supplementary energy.

The result is that, when taking other "possible states of the universe" slightly different than what the big bang predicts (and thus with higher entropy), and computing their backwards evolution into "alternative versions of the big bang", we no more get a clean uniform big bang but a chaotic one with higher entropy, as a big crunch would look like with time reversed : a big crunch is made of

- Some cold regions but also some hot regions with quite higher
temperature for a given density of matter (i.e. density of
nucleons), which explains the higher entropy ;

- Separate collapses of diverse regions into black holes that
only merge later.

Between the very first time of the big bang supposedly with "very low entropy" which we cannot yet describe by our theories, a lot of things could happen to create the relative lot of entropy there is later in the rather uniform soup that we can describe (such as the time of big bang nucleosynthesis).

See also:

Notes on decoherence and interpretations of quantum physics