Are you a subroutine?

The idea that we may all be living in a simulation, sort of like The Matrix or the world in Dark City, isn’t actually very new. Nick Bostrum published a paper in 2001 in which he made the following, almost-impossible-to-follow argument:

This paper argues that at least one of the following propositions is true: (1) the human species is very likely to go extinct before reaching a “posthuman” stage; (2) any posthuman civilization is extremely unlikely to run a significant number of simulations of their evolutionary history (or variations thereof); (3) we are almost certainly living in a computer simulation. It follows that the belief that there is a significant chance that we will one day become posthumans who run ancestor-simulations is false, unless we are currently living in a simulation. A number of other consequences of this result are also discussed.

Now, I feel like a fairly smart fellow, but I find it very difficult to follow what he’s saying. But what I think he means is this:

One of the following is true: Either we’re probably going to go extinct before we can make universe simulations, or a post-human civilization isn’t going to bother running a lot of simulations, or we are almost certainly living in a simulation.

As much as I can suss out of this argument, I can’t say I even agree with it. It seems to argue a very limited number of outcomes from an almost unlimited set of possibilities. For example, maybe mankind doesn’t go extinct, and maybe they do run a “significant number of simulations” of their history. This would seem to indicate that we almost certainly are living in a simulation, and I can’t quite make that leap logically. Maybe they run the “significant number of simulations” and they never get them quite right. He seems to be arguing that either the future is happening, or it isn’t. As I said, I’m still trying to noodle it through.

However, I am not going to argue his points, but rather just entertain the idea that we could be living in a simulation. I think it’s entirely possible, or at least I haven’t heard any compelling arguments that it is definitely impossible.

Elon Musk, who I respect immeasurably, has said he thinks it’s almost a certainty that we’re living in a simulation. His exact words were “There’s a billion to one chance we’re living in base reality.” I find this an odd thing to believe, but it’s fun to wonder how he got to that conclusion, especially since I consider him to be one of the smartest and most dynamic people alive. Imagine how that could color your thinking. Musk is a guy who certainly seems to be able to do whatever he sets his mind to. Does this success come from a combination of extraordinary intelligence combined with the reassuring thought that this is all a big game and none of it has any real consequence?

This also brings up some pretty interesting ideas, not all of which are very pleasant. Are we all simulated? What if only some of us are simulated – that is, simulated to have consciousness – and the rest are just filler, people put into the simulation to increase the population? If only some of us are simulated, and others are not, what does that make the others, some sort of “sub-simulation”? If they really are only “sub-simulations” then there is no problem in us seeing them that way, but if they are real people (that is, full-scale simulations, not sub-simulations), then we essentially have become Nazis, consigning a group of people to subhuman status who are really just as sentient as the rest of us. It brings to mind the name of a story written by Robert Heinlein in the 1960’s, titled “–All You Zombies–“. That’s not a typo, by the way, those M-dashes are in the title of the story. If I’m simulated at a higher level than you, even simulated with some consciousness, or higher consciousness than you, aren’t you somehow subhuman? It doesn’t take much imagination to envision that such ideas could lead to horrific abuses. Unless we’re all simulations, in which case there is no abuse because there is no real person there, or even a real simulation.

So let’s suggest that there are three possible levels of consciousness, or “human-ness”. The highest level is that we’re all really human beings. The second level is that each individual is a simulation, or at least some of us are. The lowest is that some individuals are sub-simulations. This actually brings up another interesting idea that is literally thousands of years old, going back at least to Plato, that only “I” am real and all of “you” are simulations. At any rate, whether we are all simulated, or some, or only me, or nobody, is beside the discussion here.

We seem to be coming up on a wave of understanding. 50 years ago, when computers took up entire buildings, the idea of simulating all of the intricacies of the universe would have been unthinkably absurd. The first computers were built to calculate artillery shell ballistics, something you can do on your iPhone now without even closing Pokemon Go. Even the largest computer was incapable of even simulating a cockroach, let alone a human, let alone billions of humans.

Then, in 2001, this Nick Bostrum fellow looked at how computational power was growing at massive rates – not only the power of an individual computer, but the use of millions, then billions of computers, most connected by the internet into one large, distributed computing device – and thought the idea seemed less absurd. Now, in 2016, it’s becoming an idea that is not only not absurd or even unlikely, but with the exponential growth of computing power every year, seems entirely possible. If we can keep up this pace of growth in computational power, simulating a universe begins to seem even likely. That is the wave of understanding – from being far behind the wave, with no concept that a Simverse could be true, to cresting the wave and coming down the front of it on the knowledge of what computers may some day be able to do, and how inevitable a Simverse may be. From Absurd, to possible, to likely, to inevitable.

So let me address some of M. Mahin’s points in detail.

Reason 1: The technical feasibility of such a simulation is very low.

Bostrom’s paper glosses over the many huge technical hurdles of creating a computer simulation in which the end result is many conscious minds who think they are living on a planet when they really are just living in a computer.

This is where I’m going to invoke the “World of Warcraft Argument.” Or I might call it the “Grand Theft Auto Argument.” I’m still undecided on that. World of Warcraft is an online game where thousands of humans control characters of various kinds, and go on adventures, kill zombies and elves, and dress up in silly costumes. It is, on a basic level, a simulation of a world. It’s a small world, to be sure, and the simulation is very basic, but it’s still a simulation on some level. Grand Theft Auto is a similar kind of game, but instead of killing elves, you rob drug dealers, run from the cops and drive around like you always wanted to.

The games both simulate “worlds” with people in them, in different ways. The World of… the World of Warcraft is larger, has more simulated creatures and thousands of real players in it, while Grand Theft Auto simulates a large city with, generally, one human player. Also, Warcraft does its simulating on centralized servers, and Grand Theft Auto does it on your computer.

I bring these games up for two reasons. First, they are simulating worlds now, in 2016. Warcraft has been simulating its world since 2004, and the first Grand Theft Auto game (at least the first 3D version) came out in 2001, and it ran essentially the same way it does now. So you’ve been able to simulate a world on your own home computer for 15 years now.

Now obviously, the “worlds” simulated in these games are nothing like our world. They are crude, simplistic, and limited. Further (and most importantly), the denizens of these worlds don’t “learn”. They have a strict and unchanging set of behaviors. But these are simulations we’re doing now, and have been for 15 years. In the case of Grand Theft Auto, we’re doing them on basic computers that exist in the millions on our planet.

He goes on:

Bostrom’s first gloss is to ask us to suppose that in his alien-generated simulation of human experience “these simulated people are conscious (as they would be if the simulations were sufficiently fine-grained and if a certain quite widely accepted position in the philosophy of mind is correct).” We have no idea whether it is actually possible to produce human-like consciousness in a computer, so this is a very questionable assumption.

True, we have no idea whether it’s possible to produce human-like consciousness in a computer. This is certainly no evidence that, in the future, people will not be able to. Or, as the case may be, in the present, in the real world, not our simulated world. Not knowing if something is possible does not make it impossible, and history is certainly rife with examples of things that were not believed to be possible, but were done anyway. A short list would include flight, breaking the speed of sound, landing on the moon and computers in your own home. Humans have a bad habit of assuming that things that are impossible now will always be impossible.

Concerning Processing Power

Here are some of the ways people reduce the computational load of simulations now.

1. Simulate only what the user can see. If you’ve ever played Minecraft (another game, yes, I know, I have a problem) you may have noticed that sometimes, you’ll run to a new area of the world, and it isn’t quite there yet. In fact, it doesn’t exist yet. Your computer got a little behind, maybe it had to drop the kids off at school, and it didn’t get around to generating that part of the world. It isn’t far behind though, and as soon as it figures out that you are entering a new part of the Minecraft world, it quickly generates it right in front of your eyes, and then goes on like nothing ever happened.

Minecraft worlds are, for all functional purposes, unlimited. You would probably die of old age before you could explore a Minecraft world to the limits of the game’s ability to generate new land. Normally a world that large just takes too much computing power to maintain. Minecraft gets around this problem by only creating parts of the world you actually see. And once you leave a part of the world, it essentially “unloads” it and doesn’t bother simulating it. It has saved the particulars of that part of the world in memory somewhere, but it is no longer actively simulating it. The sheep and cows are no longer wandering around being eaten by wolves, but the lay of the land is preserved.

So a simulation of our universe does not need to constantly simulate every person, planet, galaxy and molecule. It only needs to simulate what we’re looking at. And I find a very interesting reflection of this in the Schrödinger’s cat thought experiment.

I can’t go into too much detail about this, it is beyond the scope of this article, but I have to give you the basics. Let me start by saying I may get parts of this wrong, so if I do, please contact me and tell me what I screwed up. I definitely am ready to be proven wrong about this thing that I barely understand.

The most widely held school of quantum mechanics (the behavior of very small objects in our universe) is that quantum systems can exist in two states at once. That is to say, they can be both “on” and “off” at the same time, a state called “superposition”. We never experience superposition in our world – our car is either in our driveway, or it isn’t. It is never “there” and “not there” at the same time, but apparently the quantum physics folks (who are far smarter than I am) have found that this can actually be true on the quantum level. Now, if we ever actually look at the particle, that is observe it somehow, it collapses from two states to one. The particle is no longer “on” and “off” at the same time, it becomes only “on” or “off.” The act of observing the particle causes it to collapse. This is a very weird, and very important, point.

Erwin Schrödinger thought this was hogwash, so he came up with a thought experiment: Say you put a cat in a box. We’ve all done that, right? But then, say you put some cyanide in the box, connected to a mechanism that either releases the cyanide (killing the cat) or does not release it (leaving the cat alive). That mechanism includes a quantum particle, which will only fires when you look in the box and observe it. That is to say, until you open the box and look inside, the cat is both alive and dead at the same time.

This is quantum behavior, in a nutshell. Hey, I didn’t write the rules here. I’m just playing the game.

So this quantum behavior, it seems to me, could be evidence of a simulation system that is saving processing time by only simulating things we are observing. Strangely enough, it may be that the world of quantum mechanics and the world of Minecraft both do the same thing: They only build the world when we are looking at it.

If the universe is a simulation, and we are the only observers, the programmers don’t have to simulate every molecule, or atom, or even planet. They only have to simulate the parts we are looking at. Pluto? It was maybe 1000 bytes of information before New Horizons explored it. Up until July 2015, that was the extent of our knowledge about this planet (or dwarf planet). So in a simulated world, up until July 2015, the “sim load” of Pluto was maybe 1,000 bytes. Maybe 10,000. That’s trivial even to modern computers. Heck, it was trivial back when I was using 5-1/4″ floppy drives, and that was a long dang time ago.

But now, the information we have about Pluto may be something like 10 terabytes (10,000 gigabytes). But the info we have for, say, Callisto, is still less than a megabyte.

In a simulation that is programmed to show only what is visible, and indeed what we are looking at, What we can see is computationally trivial.

Bostrom seems to have made the mistake of calculating the computing cost of a passive human experience (one in which we are like roller coaster riders without any branching options), rather than computing the infinitely higher cost of a free-will human experience (in which we are like car drivers who can drive anywhere and get out of the car anywhere and interact with people anywhere).

In fact, the likely computing cost of a simulation creating something like the consciousness of all human beings (with complete free will and a full range of choices) would in all probability greatly exceed the maximum capability of a computer that had all of the mass in a solar system (the sun, all the planets, and all the asteroids).

I don’t agree, on two points: First, he must be assuming how much computational power we can create using the techniques of today. That is, a computer with the mass of the earth, using today’s processors. This doesn’t take into account the possibility that the simulating civilization has access to much better computational technology, which is a certainty.

The second point is this: How much free will do you actually have? Sure, there are an infinite number of possible actions you might choose to take, but how many are you actually going to do? As I sit here typing this on my computer, I can do any number of a thousand things – I could get up or go outside, or throw my mouse at the wall, but of the many actions I could take, how many will I actually even possibly do? The actual number of actions that I might do in the next minute is pretty low. I’m not going to throw my mouse, and I’m probably not going to get anything to eat because I’m not hungry. The number of actual choices I might make is pretty low in reality.

What does this all mean? It means that we literally have evidence that we are living in a simulation. This is no longer wild fantasy. We now have evidence that the universe that we are not currently observing is not real.

https://blogs.scientificamerican.com/observations/should-quantum-anomalies-make-us-rethink-reality/

https://en.wikipedia.org/wiki/Fine-tuned_Universe
Last edited 4/21/2018