In Defense of the Simulation

In recent weeks, there has been a little bit of discussion about the universe as a simulation.  I wanted to toss a few pennies into the mix and give a defense of the idea.  This is not to argue that I believe the world is actually a computer simulation or part of the Matrix or anything like that, but simply to point out that the idea has more going for it than a lot of people first realize and it is wise to consider rather than just brushing it off as if it has no merit.

The first line of reasoning is something I’ve mentioned before.  It actually fits Elon Musk’s argument on why he thinks the universe is a simulation, and I’ve also heard Neil deGrasse Tyson make a similar claim.  It deals with the statistical likelihood of us being in a simulation and if materialism is true, I believe it is an airtight argument.  The line of reasoning goes like this:

We have already advanced our computer technologies to such an extent that we can do very complex simulations right now.  The rate of computing is increasing so quickly that it seems that in a very short amount of time, we will be able to do such things as fully replicate a human mind.  In fact, we will be able to replicate more than one mind at one time.  The instant we are fully able to replicate a human mind, given materialism, the simulated mind will be self-aware and thinking.  Ultimately, there will be no difference between the simulated mind and our own physical mind.  In fact, in 2013 researchers were able to simulate one second of biological brain processing time using 82,944 processors over 40 minutes.  In theory, again given materialism, that network would have been identical to a normal human being’s brain with the awareness of one second of time passing.

So we are on the cusp of replicating human beings in a digital environment.  If materialism is true, there will be no difference in the mental space of the machine and the biological components.  In fact, materialism stipulates that we are all biological machines as it is.

But here’s where the math comes into play.  If the minds we mimic are indistinguishable from our own, and our own minds are the types of minds that design simulations, then it stands to reason that the digital minds we create are also going to create their own simulations.  And if we are able to simulate ourselves to such an extent that we are indistinguishable from our simulations, then so too our own simulations will be able to create their own simulations that are indistinguishable from themselves.  In short, if it is possible for us to do this, then it is a statistical certainty that we actually have already done this, as have the minds that we have created, and so on.

Given the fact that billions and billions of simulated worlds have therefore been created, the math game is simple.  There is one “real” world and billions and billions of simulated worlds that are indistinguishable from that “real” world.  Thus, the odds that you are in the one “real” world are billions and billions to one against.  Therefore, you are statistically certain to be in a simulation.

As stated throughout, the above line of reasoning only works if materialism is true.  Additionally, it relies on us actually getting to the pivotal point where we can actually replicate a mind exactly like our own.  But this line of reasoning is not the only line of reasoning leading toward the simulation conclusion.  So let me look at the second line of reasoning, which is the quantum world.

The world “quantum” comes from the fact that energy is dispensed in indivisible packets called quanta.  Similar to the concept of atoms, there is only so far you can subdivide until you reach the foundational limit.  But in the realm of pure number, the quanta does not make sense.  You can easily divide numbers on and on and on, until they reach an infinitesimal value, and even at that point you can still divide it an infinite number of times more.

But while you could do that in mathematically theory, in the real world you cannot.  The real world is quantized.  At first glance, it doesn’t make sense why that should be so.  However, those who have studied computers understand this quite easily.

If you use computers for math functions, you quickly realize there is a limit to the size of values.  You cannot divide an infinite number of times because at some point the computer runs out of the resources to hold the data needed to divide further.  Therefore, there is a fundamental limit in place, the absolute smallest amount of data you can go to, which is of course the bit, 0 or 1.

This means that computers, fundamentally, are quantized.  It is impossible to get around this.  The quanta is built into the foundation of how computing works.  So, while it is not apparent why such restrictions would be so in the “real” world, it is extremely obvious why they would be so in a simulated world.

We can even extend this metaphor further when we consider the role of observation in the way that experiments play out.  The most famous example of this is the double-slit experiment.  Briefly, for those who may not know it, when trying to determine if light was a particle or a wave, researchers shone a beam of light through a single slit and onto some photosensitive paper (i.e., a film plate).  What showed up is what you’d expect if light was made of particle-like photons.  However, when light was shone through two slits set a short distance apart, what appeared on the photosensitive paper showed interference zones, which happen with waves.  This held up even when the light was dimmed down so that it was only releasing one photon at a time.  Each photon would make a single distinct point of impact, much like a particle would, but over time the design that built up showed the interference pattern of a wave.

Even this is not truly the most remarkable aspect of this experiment, however.  When scientists tried to determine which of the two slits a single photon went through, the simple fact that they were observing the experiment meant that what appeared on the photosensitive paper no longer showed interference.  In other words, by trying to observe which slit a photon went through, the photon no longer acted like a wave at all: it behaved as if it had always been particles.

Why this happens is still not understood.  Clearly, the act of observing the experiment interferes with the experiment and changes things, but it is not clear why putting a measurement device on one slit and not the other changes the behavior of particles going through the slit without any measurement device too.  And if you really want to melt your brain, look up the quantum eraser experiment, wherein by “erasing” the data that one would have learned by the observation so that you cannot use it, the interference pattern reemerges.  Yes, this means that whether or not the interference pattern emerges is completely dependent upon whether or not you can know which slit the photons pass through.  If you can know, there is no interference; if you can’t know, there is interference.

Again, this behavior seems quite confusing and not intuitive at all, at least in our “real” world.  But if you were designing a simulation, it would make perfect sense.  I know this because at one point I was going to design a simulation that would pit two countries in a war against each other.  A player would control one of the countries, controlling the army’s budget and things of that nature.  But because I was thinking of countries with populations of millions, it quickly became apparent that it would be too calculation intensive to replicate millions of units for each country.  Instead, I could use statistics to compress the data.

Each country would have a birth and a death rate.  Therefore you’d have a certain number of citizens at a certain age, with a certain number who could have children, and a certain number in the work force, and so on.  One could easily just make up a bunch of actuary tables to accomplish this.

But the problem is that it would be a very boring game.  If you’re playing the game, you want something rendered on screen.  But what would render?  Easy: what the player is looking at.  So if you zoomed into a specific city looking at specific people in a specific building, then those individuals would be actualized.  Where they come about was based on the statistical likelihood of what would be there, but once rendered they would become definite objects.  At least up to the point when the player was no longer looking at them, at which point they would go back to being nothing but a probability.

What struck me when I was musing on this was that’s exactly how sub-atomic particles seem to behave.  When you’re not looking directly at them, they behave in a statistical manner, having a probability of being in a specific location at a specific time.  But once you “zoom in” and look at them directly, they become definite, specific entities.

This even gave rise to a form of the Heisenberg uncertainty principle, because you could only see what was on your screen.  What wasn’t rendered was only probability.  And that meant that if you tracked a single entity with enough precision that it was on your screen, you had no knowledge of what was near it that could affect it (such as an enemy soldier that might “kill” that entity).  And if you zoomed out enough to see what was near it, you would lose the specificity of the location of that entity.

In designing this simulation, this behavior was needed because it would save computer processing time in not having to keep track of millions upon millions of individual entities.  It seems reasonable that if we are in a simulation, the same thing would be in effect now, even if the fundamental computer running our software is trillions of times more advanced than what we have today.

My final stream of evidence involves a similar concept, which is the relativity of time, mass, and length.  We know that time slows down (relative to stationery objects) the faster that an object moves, up to the speed of light where time stops completely.  Equally, the mass of an object increases up to infinity when it is at the speed of light (this is why photons, which move at the speed of light, have to be mass-less, because if they had any mass it would have to be infinite).  Additionally, the length of an object shrinks, relative to a stationery object, the faster that it moves.

All of these things are linked together, therefore, and it’s almost like there is a governor attached.  Again, we see similar things in simulations.  Calculations have caps in place to keep the simulation moving smoothly.  If you use more resources in one area, you have to free them up in another area, or else the whole thing gets bogged down.  This sort of limitation is exactly what we see in computer systems today, put in place to keep things from blowing up and becoming useless.

In conclusion, I’ll just say this.  I’m not advocating that we actually are in a simulation.  By no means.  But if I were going to design a simulation of the universe, the way I would have done it leads toward the very things that we see in the universe around us.  So make of that what you will.