PGD Home

Simulated Reality

by on Dec.09, 2015, under Musings, On Game Development, P-0

I have been thinking about the concept of Simulated Reality for some time now. I came across this idea from reading the wikipedia article on it. It is closely related to the simulation hypothesis.

Basically, the idea is that is it possible to create a computer simulation that is indistinguishable from the real world, and that the liklihood is that we are currently living in such a simulation.

Firstly, is such a simulation possible? I believe it is. Look at how far we have come technologically in just a few decades. We have gone from Pong to Call of Duty, and from simple text-based MUD’s (Multi-User Dungeons) to World of Warcraft. 50 years is a completely insignificant amount of time in universal terms. What kind of computer simulations will we be able to run in 100 years? or 1000 years? Or 100,000, or 1,000,000? I think that given out current rate of advancement, developing the capability to create a simulation indistinguishable from the real one is only a matter of time.

The logic as to why we are probably living in such a simulation is quite simple. Humanity has only been around for a heartbeat, compared to the life of the universe. We have also only explored a fraction of our universe. It is reasonable to think that somewhere out there, in the vastness of space, there are a large number of intelligent civilisations. Given the age of the universe, it is quite probable that a great many of these civilisations are at a significantly more advanced state than ours.

If we assume that a civilisation with a sufficient level of technology will create these types of simulations, and we we assume that such a simulation is possible to create at any technological level, then it is far more likely that we are currently the ones being simulated rather than the ones who will eventually create the simulation.

But how would a simulation like this be created?  Is it even possible, for even an advanced civilisation? We have advanced computer games today, but even though they look quite realistic, they are certainly not at the point where one would mistake them for reality.

I think that fundamentally, our universe is just a very large number of particles. If we could create a computer that could store the position and velocity of every particles in our universe, then we could create a simulation of that universe which would be indistinguishable to it’s simulated inhabitants.
Think of it like a snooker player preparing to take a shot. If you know the position and velocity of every ball on the table, and you know the variables affecting the que, such as impact velocity, angle, etc, and you know the variables affecting the environment, such as surface friction, air resistance, etc, then you can predict the position of every one of those balls with exceptional accuracy, from the moment the que strikes the first ball, to the moment that the balls come to rest.
Obviously, simulating the entire universe would be vastly more complex, and would require a level of complexity far beyond anything we can even imagine, but given our rapid advanced in computing technology in the few years we have been on this Earth, I would not be at all hasty to rule out a computer like this.

This idea would encompass some fields of modern quantum mechanics, especially the Grand Unified Theory, and the Theory of Everything.

One problem with this theory is the idea of “nested” simulations. If we assume that the amount of processing power required to simulate a universe is given by the variable X, and we assume we construct a computer with just enough processing power, what happens if the blissfully unaware inhabitants of that system run their own simulation of their universe, not realising it is a simulation itself? That universe, remember, is a 100% accurate simulation of our universe, so it should require the same processing power to simulate. If the inhabitants of the simulated universe create their own computer to simulate their universe, then the original computer (in the “real” world) is now simulating two universes: the first simulated universe, and another nested simulation in that universe with the same level of complexity, simulating it’s own parent universe.

This means that the original computer would actually require X + X processing power, or 2X, but it only has X processing power available, just enough for one simulation. Does this break the theory?

I believe that it does not, for two reasons.

Firstly, it is possible that the original simulation is an “ancestor” simulation, IE, it only simulates a time period from the past, before the technology existed to create advanced simulations. Therefore, a nested simulation would not be possible.
Secondly, it is possible that an infinite number of nested simulations could be run without increasing the processing load on the original computer in the real world. How? Well, assuming that the original variable X is the processing power required to simulate the position and velocity of every particle in the universe. This means that when the inhabitants of this simulation build their computer, they are merely rearranging the particles which are already being simulated. When that nested simulation is run, the electrons running through it’s wires are merely particles whose variables, position and velocity, are being updated in the parent simulation. It really doesn’t matter what configuration they are in, the processing requirements from the parent computer will be the same. Again, think of the snooker table example. I can place the snooker balls on that table in any number of different configurations, change the position, angle, and power of the snooker cue in any number of ways, but the complexity of my snooker table simulation doesn’t change, since the variables themselves are only changing, not becoming more complex. Since I am not adding any new snooker balls, the complexity doesn’t change.
The same should be true for nested simulations running within a parent simulation.

Another problem with this method of simulating the universe would be the uncertainty principle. The uncertainty principle states, essentially, that it is not possible to know with certainty both the position and velocity of a subatomic particle. If this were to hold true, it would not be possible to simulate even a single atom with absolute certainty, since the position and velocity of it’s component parts could not be known with certainty.
However, I believe that if a civilisation had advanced to the state where a computer powerful enough to simulate an entire universe could be built, that civilisation would have access to a level of knowledge of physics that our society can only dream of. It is quite possible that the reason why we are limited by the uncertainty principle is due to our current physical models being incomplete, or incorrect. Scientists have only recently begun exploring the quantum world and it’s many exciting discoveries, and it is a strange place. Theories are often found to be wildly inaccurate, and are thrown out, or modified. Sometimes impossible things become possible, and perfectly reasonable things turn out to be impossible. I don’t think that anything “Banned” by our current knowledge of physics, such as the processing power required by a computer, or the inability to precisely know the position and velocity of particles, should preclude us from considering the possibilities of a world without those limits.
It would be quite possible for a simulation to be run by a civilisation with knowledge of a “Theory of Everything” which simulates a society like ours, which does not have that knowledge. It is quite possible therefore, for a simulation using an inaccurate or outdated physical model to be run within a simulation which uses an accurate, complete one.

I have always been interested in virtual worlds and artificial worlds of all types, and so the simulated reality theory naturally captivated me. We will not advance to the stage where we can create our own realistic universe simulations in our lifetimes, of course. However, could it be possible that we are just variables being modelled in a computer somewhere? Could the very nature of our reality be just a simulation? If it is, does that make our reality any less “real”?

I have manged to avoid mentioning a certain popular Science Fiction movie in this article so far, but I feel that a quote is appropriate here:

“What is real?
How do you define real?
If you’re talking about what you can feel, what you can smell, what you can taste and see, then real is simply electrical signals interpreted by your brain.”

That is of course a quote from the Matrix.

I would go even a step further than that. Reality, at it’s most basic level, is just a particle simulation. We already have simulations to predict weather, the effects of a nuclear explosion, and the outcome of complex physical processes. I see no reason why a sufficiently powerful computer, armed with an advanced knowledge of quantum physics, could not simulate an entire universe, and the intelligent beings therein.

We may all be no more than variables in a vast computer, our lives nothing more than the execution of a computer process. our reality just a simulation, possibly among many, worlds within worlds, each creating worlds of their own, each believing that they are the creator and not the created.

Facebooktwittergoogle_plusredditpinterestlinkedinmail

Comments are closed.

Looking for something?

Use the form below to search the site:

Still not finding what you're looking for? Drop a comment on a post or contact us so we can take care of it!