Talk:The Anthropic Principle

From Simulism

[edit] --TonyFleet 19:05, 11 February 2007 (CET)

There is a clear link between the anthropic principle(s) and the notion that we are living in a simulation. If we are in a simulation, then in all likelihood it will have been designed expressly to hold us, and therefore the Final Principle applies. In fact, if either the strong or final principles are valid, then this fact might be construed as pretty good evidence that we are in a simulation (or that a god exists - which may or may not be the same thing, depending upon your point of view!).

The arguments which first gave rise to the Anthropic Principle concerning the fine-tuning of the universe to allow for our existence have been used by many as an argument for the existence of (a) god. However, it does seem to me that many of these god-arguments are self-defeating, as they fall to a version of Douglas Adams' "Oolon Colluphid" argument going something like this: "I refuse to prove that I exist", says God, "because proof denies faith, and without faith I am nothing". "Ah, but", says man, "all the cosmological constants have been lined up so as to allow for and ensure my continued existence - probably so that you can have me worship you. This is a dead give-away. Therefore I have proof that you exist; ergo, by your own argument, you don't..."

Although it is possible to pick elephant-sized holes in this spoof argument, there is no denying that if it did turn out that the fine-tuning of the cosmological constants was, in fact, direct evidence for the existence of a god who valued faith and belief, then an interesting question would need to be answered: why would such a god create the circumstances in which beings of the 3rd Millenium were provided with this proof of his existence, whereas beings of previous epochs had to make do with belief? It must surely be the case that the more evidence that is accumulated towards the correctness and the strength of the Anthropic Principle, the more weight it would add to the argument that this is a simulation created by simulators in an external reality, rather than a Universe created by a god.

[edit] Some further thoughts on cosmological fine tuning - The "Goldilocks" enigma

The currently accepted way in which the fine tuning of the cosmological constants is "explained" without invoking either God or a Cosmic Simulation Programmer, is to hypothesise that the universe is not actually a universe at all: it is a multiverse, in which there are stacks of similar universes spreading out in a multiplicity of dimensions. This is what Douglas Adams called "The Whole Sort of General Mish-Mash" or TWSOGMM. If we could move through these different dimensions, in TWSOGMM, we would encounter universes that are slightly different to one another. For example, if we were to move away from our current universe on the 'gravitational constant' dimension, we would encounter universes in which gravity is lower or higher than ours, with the effect that galaxies could not form or that the universe consisted of a single black hole.

The consequences for creating a simulated multiverse (instead of a simple simulated universe) would be huge. To begin with, why bother? There would be many universes which were effectively sterile, in which there would be no life whatsoever, and this could be predicted without actually creating them. If the gravitational constant were zero for example, then this would be a universe without gravity. The force would not exist, and consequently there would be nothing in the universe, as matter would not exist per se. Even if the purpose of the simulation were to explore TWSOGMM, then there would be much better ways of doing this than going to the trouble of creating individual simulations of them. Optimised Search techniques would allow you to search through TWSOGMM to find pockets where exotic universes could exist, which had wonderfully weird combinations of cosmological constants which could support bizarre kinds of cosmological structure. It would then be possible to create a universe to see what happens with that particular combination, to see how the universe would function and evolve.

There are many people (Max Tegmark, David Deutch, Martin Ryle) who seem to contend that all mathematically possible universes actually exist. All of them accept that the consequence of this view is that the overwhelming majority (something greater than 99.99999999999%) of the universes are just non-functioning wasteland universes, somewhat akin to junk DNA. This might all be true; however, it does not seem reasonable that if you were going to design and build a simulation of a multiversal reality and set it running, that you would fill it full of non-functioning junk universes. That would consume energy and time and resources, and unless you had an infinite amount of these (such as if you were God or the TWSOGMM extended infinitely in every conceivable dimension), then you would wish to concentrate your efforts on 'interesting' cases.

Recent research seems to suggest that such pockets of exotic universes may well exist in TWSOGMM, but it is unclear how these would function, whether their existence would have some sort of continuation through time (whatever that means in universes which might have mind-manglingly different versions of "time") and whether they would be of such a nature which would allow some intelligence to exist within them. It is not hard to imagine a scientist in the near future who wishes to explore such an idea, creating a simulation of that universe, setting up the cosmological constants and just setting it running "to see what happens". Such a scenario seems highly likely, and this would leave any intelligences created within that universe asking themselves the 'Goldilocks' question of "Why is it that our universe is just right for us?"

This then leaves us with the following options:

• TWSOGMM exists in all its full glory, and we living in one of the exceedingly rare and exotic regions of it which can support intelligent life. Therefore to ask the question "Why are the the cosmological constants the way they are", is tautological - we are here (in this region of TWSOGMM) because if we were elsewhere we could not ask (or answer) the question. It's rather like saying "Why do humans live on the surface of the earth rather than at the bottom of the sea", the answer being that if we did the latter we would already have drowned and we could not then ask the damned fool question in the first place.
• TWSOGMM is simply a theoretical entity, and does not really exist. If that is the case, then our options are severely limited. We now need to account for why the cosmological constants are the way that they are. This reduces us to these two possibilities:
  1. The Universe possesses these constants "by chance". To say the chances of this happening are small is an understatement. The chances are so remote, that if a National lottery had been held at every moment of Planck Time since the beginning of the universe, the chances of the same person winning every lottery would be powers of ten greater than the probability of these constants having values which would support life as we know it. The only way that this could conceivably occur is if the universe had a continued existence in time (or other dimensions?, and these values are chosen differently in repeated versions of the universe. The problem here is that this then begins to resemble TWSOGMM, and we are back with case (1), but with an slight problem. We either accept the full version of TWSOGMM, we we need to hypothesize a 'modified' version of it where there are repeated (but slightly different) versions of the universe through time - a series of big bangs & big crunches, all with slightly different sets of cosmological constants, somehow allowing the imprint of one universe to seed itself into the subsequent version. This could conceivably allow for some sort of cosmic evolutionary process. However, Occam's razor begins to kick in here. Why go for a particular version of TWSOGMM which only allows variation on the time dimension, and how on earth could an imprint be left after a big crunch? Additionally as time is created and destroyed in big bangs and big crunches, how meaningful is it to talk about 'evolutionary processes' in such scenarios?
  2. The Universe has had these constants specifically selected for it. This means (in some sense) that our universe has been 'designed'. It does not necessarily mean that there is a 'designer', nor that the universe was created by God or a programmer specifically for life (as we know it - or otherwise) to exist in. Simply, that it was known beforehand that the particular combination of constants would produce a universe with something interesting in it, and that (for whatever reasons, and however it is configured), this universe was called into existence to explore what that 'something' might be.

The interesting issue here is that a consideration of the "Goldilocks" enigma now seems to have reduced our options towards a belief in one of two scenarios - (a) either that 'reality' is unimaginably more extensive and more profound than we have previously understood (TWSOGMM exists),or (b) that the parameters underpinning the universe have been specifically selected, in order to create a universe which has the potential for an interesting outcome, possibly that of supporting intelligent life. From this point onwards you take your pick. Methodological naturalists will plump for version (a) because that explains what we see without invoking the need for external entities or teleological explanations. Subscribers to religion will plump for (b) because their world view is that there is an external entity who is responsible for the creation & running of the universe. Arguments between these two viewpoints really cannot be resolved unless someone found God's (or Slartibartfast's) signature on a glacier.

The consequences of this argument for Simulism being a coherent explanation of the universe are interesting. Firstly this provides a further possible reason for creating a simulation, i.e. specifically to explore what would happen in a universe with particular sets of cosmological constants. Other reasons for creating simulations include those hypothesized by Nick Bostrom - 'ancestor' simulations, which examine particular past (or future!) historical or sociological events, Stephen Wolfram's 'digital universes' designed to explore cellular automata, or simulations designed for entertainment purposes. This 'new' version of simulism is more akin to the 'cellular automata' than the 'ancestor simulation', in that not all features of the simulation will have been worked out beforehand. The whole point of such a simulation would be to see what arises: the programmer would not have had to create stars, trees, bicycles, computer chips, ants and atoms as part of the programming; these would have developed as the simulation progressed and evolved. All the initial programmer needed to do would be to set up a computational substrate of sufficient capacity and complexity to hold the simulation, and to program in the rules which govern the way TWSOGMM functions (i.e. The Theory of Everything), together with the initial conditions and set it running.

If this were a valid explanation of what we see around us, then there are two further consequences:

1. It is unlikely that any programmer would interfere with the program once it is running. The clear reason for creating the simulation would be to explore what happens in a universe with different sorts of parameters. Changing the parameters of the universe in mid-simulation would be a self-defeating strategy.
2. The universe that we now see is (in terms of our conception of time) as old as it appears to be. By this is meant that evolutionary processes have acted out over time, and that stars, planets, and great apes have developed to the stage that we see now. There were no shortcuts, no LastThursdayism, and no 5-minute universe. It took as long as it needed to take, and that is a long time. In exploring the 'digital universe' notion, Seth Lloyd has calculated that there may have been around 10^120 digital interactions involved so far in reaching this point in our universe's history.

If even part of this is true, we are left with a very different conceptualisation of Simulism than the one which is explored in the rest of this site. Simulism by its very nature seems to imply some sort of design, and the notion that the design is simply the selection of particular constants is really the very loosest notion within the meaning of "design". In fact, it could even be imagined that such "designing" was not in fact the conscious decision of a programmer. If a super-simulator were programmed with sufficiently complex search algorithms, it could be let loose on the the entire scenario of TWSOGMM, looking for exotic regions where intelligences might exist, and creating simulations of these regions, tweaking parameters from one run to the next, and recording what occured. Our entire history of our entire cosmos from big bang to big crunch might end up as be nothing more than a few interesting data points in some astonishingly huge data warehouse. In fact our entire conceptualisation of TWSOGMM might just be the result of one particular program invented to explore one particular version of existence. The entire reason for our existence and the existence of everything that we can imagine as part of our universe, the multiverse and TWSOGMM could just be the idle tinkering of some adolescent hyperbeing which has been submitted for their end-of-term project.

--TonyFleet 12:13, 13 December 2008 (CET)