See part 1 here.
So maybe we’re living in a simulation, maybe we’re not. Why does any of this matter? If we can’t actually tell any way, we are just arguing over the likelihood one way or the other and why should any of us care? It doesn’t change the taste of coffee; it doesn’t affect the colour of the sky. This rhetorical question is posed time and time again to “theoretical” or “non-practical” philosophers, when they ask questions that don’t have immediately testable answers.
Hearing an argument like this is deeply saddening, as it implies the only thing that matters to you, in your life, is the actions you take and the consequences of said actions. There is an implicit premise in the asking of this question, which is that something doesn’t matter unless it changes how you behave, or effects some practical set of physical actions you take in the world. Debating about the simulation argument matters simply because it can affect how you see the world. It may not change your daily routine, but it can change the way you think and feel about your life. Thinking about problems of scepticism matter because they can make you appreciate what is going on in the world in a new light. Caring about the truth, about metaphysics and epistemology is important because it can affect you in a way that still matters which western culture has seemingly forgotten about. Also, the simulation argument specifically means that there are sound naturalistic arguments for a higher power, which its deeply interesting in itself. As well, if we can eventually simulate universes, it gives us a way to partially test physicalism in a way most philosophical theories could not ever be tested.
Now onto the next part, is there any way to actually discover if we’re in a simulation? This is not a simple question, but I believe there are ways through which we may be able to discover this, or at least uncover that it‘s likely to be the case. The easiest would be looking for non-causal events, things inconsistent with naturalism, which literally break the natural order of the world. These unexplainable things would certainly be evidence of a creator in general. However there is another interesting more subtle way, which may help us discover if we are in fact being run on a computer, if we assume the creator of our simulation lives in a universe with similar or identical physical laws. As I’ve mentioned before, there is no reason this need be the case, but without this assumption there is very little we could conclude at all.
This alternative method of detecting if we are in a simulation is provided by the observation of chaotic systems. We know through observation that there are physical systems (such as a double pendulum, the weather) which while deterministic, are almost never predictable. These systems all share a common property, in that varying the initial condition of the system by very little, results in drastically different states of the system at later times. This is what it means to be chaotic, to be incredibly sensitive to initial conditions. An example is that a fully extended double pendulum released while being held up at an angle of 45 degrees will follow a wildly different path than one released from an angle of 44 degrees. This is not the case for the single pendulum.
People can and do simulate double pendulums all the time on computers, in order to try and predict their chaotic behaviour. However any calculation of a double pendulum’s motion on a computer will ultimately fail to remain faithful to the mathematical rules the pendulum’s motion follows. Why is this? Precision. A computer, no matter the size, will inevitably have a finite precision. You can have 1 digit of precision or 100, but you cannot have infinitely precise calculations done in a computer. Which means that when computing the movement of a pendulum, you have to cut off the precision at some point. Consider the following simple example. Suppose your computer can only hold 5 digits after the decimal place. In reality, you drop a pendulum from an angle of 50.123456789 degrees. When you plug this into your computer you are forced to enter only 50.12346 (if you’re rounding). Because these initial conditions are not identical, eventually the calculated, simulated pendulum’s path will diverge from that of the actual path the real pendulum will take. The crux of this difference is the (well founded) assumption that space and time are continuous in our baseline reality; that is to say, you can zoom in and out forever and space will never be made up of some discrete chunk. The example opposite to this is of matter, where everything in our lives looks like a continuum, but is in fact made up of enormous quantities of discrete atoms. Space on the other hand can be subdivided infinitely. This makes it the case that you need an infinite number of digits to specify something’s position in space, particularly once it is in motion. This fact that matter moves continuously through space, whereas coordinates on a computer are forced to jump from one number to the next can be realized in the above example in seeing that the pendulum would be forced to jump to position 50.12345 and avoid the infinite set of numbers between 50.12345 and 50.12346.
This divergence of the simulated pendulum from the real pendulum which exists in a space where all spatial measurements are continuous (in the mathematical sense) tells us that the pendulum on the computer is not one in baseline reality. This means if we could ever show that a pendulum in our universe consistently diverges from the continuous mathematical form we assume it follows, or if we could show that the double pendulum (or any chaotic system) could be properly modelled with a finite amount of precision, we would have definite evidence that space is fundamentally discrete, and so either our understanding of physics is horribly flawed, or we have a great deal of new evidence that we are in fact being simulated on a machine.
The Third Triumvirate 