The Simulation Argument

"The Simulation Argument is perhaps the first interesting argument for the existence of a Creator in 2000 years."- David Pearce, transhumanist

---

One of the most thought-provoking ideas to emerge and be furiously politely discussed in philosophy departments is the simulation hypothesis. The above quote by David Pearce so humbly expresses the consequences of this idea, which, if you think about it, has such huge ramifications that it not only has the potential to shake the foundation our modern ideas are based on, it can make it easier to question any idea at all... maybe.

The idea itself is simple and you might have heard about it. What if our universe is a computer simulation?

I was walking on the street and thinking well, okay, so if we were in a simulation, what could that mean for us? Hmm… well, on the face of it, it could mean that our reality isn’t what it seems. But, maybe, it could also mean that there could be a God. There could be an afterlife. There could be hell, there could even be Bertrand Russell’s teapot somewhere between Earth and Mars. Every religion could be valid.

I mean, as a person with skeptical leanings, I’m careful not to accept any truth claim with a 100% certainty. That includes the existence of a god–or lack thereof. I really doubt there is a god, but I can’t be 100% certain of it. Maybe 99.9, but not 100, because only a Sith deals in absolutes.

So, if you’re such a skeptic, why does this skeptical idea make you suddenly jump to the god conclusion, I hear you say behind your computer (I’m a psychic as well as a skeptic). And that's a fair point, well done for thinking that, person reading this.

To me, the idea of existing in a physical universe, governed by physical laws, brought about by evolutionary processes that do not require a god anywhere in the mechanism, makes those aforementioned scenarios, when I had my “sudden clarity Clarence” realizations on the street earlier, seemingly unlikelier. Whereas, in a simulation, you don’t know what the creator or creators intended, nor do you know the rules of the game.

Genesis in a religious context suggests a momentousness to the very idea—the universe wasn’t just created nonchalantly by a nerd somewhere. While a simulation could likely be implemented just for the hell of it or to quickly test something or analyze populations. Maybe I’m biased to think that because of playing computer games, where often the point of the game is to simply have fun. Breaking the rules or accessing the console to mess with NPCs is often the most entertaining part and—unlike the New Testament god, for example—rarely do I feel any kind of paternal love for the NPCs generated when I start up a game. If we’re in a simulation, should we expect our creators to have any love or care for us?

Thanks to the Matrix movies, most of us are at least somewhat aware of the simulation hypothesis, but I thought I’ll take you on a quick ride and give you an idea of, not what popular culture has to say about it, but what dudes with coke-bottle glasses, b.o. and two-hundredfold IQs think about it. You know, the boffins.

But first, I must mention that this is not really a new idea and is probably old as philosophy itself, going back to the ancient Greek skeptics, who kind of doubted everything. The sophist Gorgias doubted that anything exists, he himself included. That’s a nice one to pull on the boss after you show up to work late and drunk. "How can I be late, when I don’t exist." René Descartes at least postulated with his famous “evil demon” thought experiment that while everything else might not be real, he himself has to be real. Cogito ergo sum.

The more contemporary brain-in-a-vat argument by Gilbert Harman seems like the simulation argument, except the brain in that scenario is still established and operates in the real world. The simulation hypothesis idea goes a little further than even the Matrix movies and suggests the possibility that your brain, along with the rest of your body, is also a simulation. Nowadays, many physicists, like Max Tegmark, believe that the universe on a fundamental level is mathematical and computational, which could give additional weight to the argument.

---

Are we living in a computer simulation?

The modern version of the simulation argument, bolstered with everything we’ve so far learned about computers and cognitive science, was expressed by mega-boffin Nick Bostrom in a paper in 2003¹. His argument is commonly misunderstood to mean that he poses a simple question, asking if we are in a simulation. He actually proposes 3 likely scenarios and states that at least one of them must be true. This is an example of a trilemma.

So, what are those three scenarios?

1. The human species will not reach a posthuman stage that is able to simulate our current selves. While it doesn’t necessarily mean humanity will go kaput, it might mean that we’ll end up staying at our current technological level indefinitely and will never reach posthuman stage.
2. We end up surviving the future, but we’ll have, for whatever reason, no interest in running any simulations of us as in our current primitive form.
3. We’re living in The Sims. Every time you’re taking a dump you’re being watched by someone.

His argument boils down to this–if computational power keeps on growing, like it kind of has done for the past 50 years, we’ll end up by, let’s say, year 10 000 with computers that have such massive amounts of processing power that it will be trivial to run simulations of us in year 2018. It would be so trivial, that it would be possible to run quite a few of such simulations. Because there would be so many simulations, and subsequently simulated humans, that the likelihood of being in a simulation is very high. If it’s true that we aren’t in a simulation, then that must be explained by his first two premises. While the first premise is pretty straightforward, the second one necessitates that there should be a strong convergence between civilizations to lose interest in ancestor simulations, meaning that civilizations that reach a technologically capable stage, won’t generally decide to try simulating their historical selves. We’re reached this technologically wonderous age and simulating a bunch of glorified Cro-Magnons doesn’t seem that exiting to me. I’ll rather go explore the galaxy in my space ship.

Bostrom relies on a few assumptions to make his argument. He assumes the validity of substrate independence, an idea that harkens back to the multiple realizability thesis of the 70s. What these sexy-sounding words (no? okay) mean in plain English is that brains don’t have to be made out of biological carbon-based cells. We could have brains that house conscious minds on silicon chips, or even potato chips. As long as it’s some sort of machine that can compute all the nitty-gritty neuronal and synaptic activity that occurs inside the brain, it will result in a thinking sentient being. Note that there are some philosophers, who reject this idea. Me? We’ll, I quit eating potato chips. Just in case I accidentally commit some sentient ‘tater genocide (‘tatercide? No? okay).

Bostrom also assumes technological progress and raw processing power will continue to grow mostly unabated to exorbitant levels and simulating human neuronal activity will be feasible sometime in the future. So, how much computing power is needed? Like ten GTX 1080 Ti-s? Well, he goes into that a bit as well. He estimates that the processing power required to simulate the entire human brain lies somewhere between ~10¹⁴–10¹⁷ calculations per second. To do a more fine grain simulation and simulate the molecules that make up the neurons would obviously require a whole lot more.

To simulate an entire universe and all the particles in it would take an inordinate amount of power, which is why omitting some of it for optimization purposes would make sense. Like the interior of the Earth on a microscopic level. The stars and galaxies could be “low res” versions that to our crude telescopes look real enough to fool us. He even suggests that the simulation could keep track of our conscious states and intervene when nosy sims are getting a bit too curious.

Simulating the entire mental history of humans (all the mental calculations of everyone ever) would, by his estimate, take about ~10³³-10³⁶ operations and, before you tell me to get outta here, numbers don’t go up that high, hypothetically it would be possible for posthumans to build a massive computer the size of a planet, perhaps one like this, that could have the processing power of ~10⁴² computations a second, which could simulate the entire mental human history by using less than a millionth of its processing power for one second. Or it could simulate 86.4 billion human mental histories every day. That’s not 86.4 billion people, it’s 86.4 billion human civilizations–each of which consists of 100 billion people that have ever lived. But can it run Crysis? (sorry, I’m still stuck in 2009)

Bostrom goes on to propose that a simulated civilization itself could also reach a technologically advanced enough level to run their own simulations. He compares these sub-simulations to virtual machines that we run on our computers today. Reality may consist of numerous levels upon levels. That has some interesting implications. He wrote:

For example, if nobody can be sure that they are at the basement-level, then everybody would have to consider the possibility that their actions will be rewarded or punished, based perhaps on moral criteria, by their simulators. An afterlife would be a real possibility. Because of this fundamental uncertainty, even the basement civilization may have a reason to behave ethically. The fact that it has such a reason for moral behavior would of course add to everybody else’s reason for behaving morally, and so on, in truly virtuous circle. One might get a kind of universal ethical imperative, which it would be in everybody’s self-interest to obey, as it were “from nowhere”.

But despite all that he suggests there is no reason to go crazy. The implications of this realization wouldn't be all that radical in the day-to-day sense:

Our best guide to how our posthuman creators have chosen to set up our world is the standard empirical study of the universe we see. The revisions to most parts of our belief networks would be rather slight and subtle – in proportion to our lack of confidence in our ability to understand the ways of posthumans. Properly understood, therefore, the truth of (3) should have no tendency to make us “go crazy” or to prevent us from going about our business and making plans and predictions for tomorrow.

---

So, what do you think? What of the 3 scenarios is most likely to you? Will we die off before reaching technotopia? Will we in the future not care about running simulations of earlier humans? Or are we in a simulation?

And let’s be fair, Bostrom himself assigns a likelihood of only 20% or so to the probability of the simulation idea being true.

---

Responses

Some philosophers have added their own thoughts to the conversation. David Chalmers takes an optimistic position, arguing that the simulation argument is not really a skeptical hypothesis, like the ones that drive us to begin to question everything, but ultimately a metaphysical one, like the ones that we’re all so accustomed to and don’t freak out about. It’s just a hypothesis about the fundamental nature of reality, which doesn’t necessitate we throw out everything we know about the world, because, damn it, we’re living in a simulation, nothing is true! There is no spoon! The end is nigh! I will put my underpants on my head! No, what he says is that there is a spoon, but the fundamental nature of the spoon is different. Just because the world is virtual, doesn’t mean it isn’t real.

When quantum physics came along and made us question our Newtonian worldview that had atoms as the smallest fundamental particles, we still accepted that the universe we live in is still there and real, even though the fundamental understanding of the structure of the world was different. One day we may find out that underneath quarks there is another layer consisting of bits. Why wouldn’t a digital world be just as legit as one that runs on physics? And wtih a religious worldview, everything is a creation of a deity, but you don’t doubt the existence of the chair that you’re sitting on neither.

Chalmers has written a lot about the subject. So if you like his ideas, then check out his website consc.net.

Another guy, Robin Hanson, responds to Bostrom’s argument with a paper of his own and says that the possibility of being in a simulation should make us adjust our behavior. He proposes that, while some simulations would be created for academic purposes, the vast majority—if these simulating beings are anything like us—would be for the entertainment value. He brings up the holodeck from Star Trek: TNG and asks how sure can we be that we’re not currently in some sort of future holodeck scenario. If we are in one, then how should we act?

To run a simulation of an entire universe just to simulate one specific time and scenario that interests the simulators would be computationally wasteful. One example of one scenario that might interest our future versions, he proposes, could be a virtual party that involves simulated beings and one of these future humans, who would like to attend the party. It would make sense to omit the unessential people, that aren’t in any way related to the scenario i.e. delete all the people leaving the party. If you had suspicions that you were a simulated being, what would you do? He suggests some strategies:

If you knew that you were a simulated person in this party simulation, and you wanted to live as long as possible, you might want to discourage anyone from leaving the party. If the simulation might end early were the future guest to become bored, you might also want to make sure everyone had a good time. And your motivation to save for retirement, or to help the poor in Ethiopia, might be muted by realizing that in your simulation, you will never retire and there is no Ethiopia.

Also, it would be unlikely that the simulation controllers would keep a simulation running as long as possible. Most likely it would be terminated at some point. So, if you thought you were a simulated being, worrying about long-term problems like global warming, wouldn’t make sense if the world was going to be ended anyway at some unknown point in the future. Instead you could have fun and live for today.

Another idea he poses is that if the people inside the simulation (the sims) end up getting wise to the whole thing, the simulators could terminate the simulation prematurely. So, maybe it’s not such a good idea to go around and spread this whole simulation hypothesis around? In that case… sorry. Forget everything you’ve read so far. Pretend you’ve been reading some awful fanfic. However, if the simulators want the sims to act as realistically as possible, then it should be expected that some sims happen to possess the simulation idea (like some of us do now) and it wouldn’t look weird to our simulators.

If the entire purpose of the simulation is entertainment, then we should strive to be entertaining, lest our bored overlords “change the channel” to something more interesting. How should we be entertaining? Well:

So one should emphasize widely shared features of entertaining stories. Be funny, outrageous, violent, sexy, strange, pathetic, heroic, ... in a word "dramatic."

I love this idea. Hell, you know what? I’m gonna go entertain the overlords and do a whole song and dance routine outside my house, in the style of 60s musicals. Be right back…

Whew, I’m back. Wow, that was fun. What do you mean you don’t believe me? We really need to have a talk about your trust issues. Now, let’s move on.

---

“Glitches in the Matrix”

Many people have written Bostrom with claims of seeing glitches, which could prove the simulation hypothesis, but he is doubtful.

I don’t buy that. We should expect to hear such reports occasionally even if we are not in a simulation. Even if we are in a simulation, the most plausible explanation for such reports is not that they result from any real “glitch” but rather that they originate in the ordinary frailties of the human mind (hallucinations, psychiatric problems, visual illusions, self-deception, fraud, and so forth).

---

What I personally think

Ultimately I think this question will be solved by theoretical physicists, but personally I think it's unlikely we’re in a simulation. Bostrom’s suggested optimization strategies aside, simulating a universe down to the quark level (and the 3.28 x 10⁸⁰ quarks that are theorized to exist) requires so much more computational capacity than what would be minimally required to simulate a life-containing universe, that I think it unlikely. That is a biased statement, however. The 3.28 x 10⁸⁰ number looks big, but compared to what? Other universes? What’s a good number for a universe? How could I possibly know? The number of quarks in our universe might be computationally a comparatively manageable fraction of all the quarks that exist in a lower level reality. Hell, our lower level reality might not even consist of quarks at all. Food for thought.

One idea I had when I read Hanson’s paper was this. Remember how silly the Greek gods seemed to you when you first learned about them? How fickle and petty they were? How goofy, those ancient Greeks. Surely their religion can’t be real. No way god or gods would behave like that! I remember thinking as a kid. But in a simulation argument it would be more likely that the gods wouldn’t be perfect all-seeing all-knowing beings, personally, in their world. They could be like us—vain, moody, arrogant. It would make more sense. We’ve never seen any being that looked anything close to a god. We see plenty of beings that are just straight up dingleberries and douche-canoes.

Second, if you think it’s impossible to create a computer inside a simulation, then I’ll just show you this video of an 8-bit CPU made in Minecraft, using the game’s redstone logic gates.

It only runs at 3.2 hertz compared to the 4 770 000 hertz of the 8-bit Intel 8080, so yeah it’s a little slow, but it works exactly like a CPU, Turing complete and everything. If the person, who did that went a little further and created an entire computer (RAM, storage, etc) inside Minecraft, it would be possible to run Minecraft inside Minecraft. It won’t be playable because it would be incredibly slow to us, but it should be possible to run.

If you create a computational representation of a human brain in Minecraft, then you could also hypothetically host a sentient being in Minecraft. Better make sure you have a fast computer for that, though.

All emulation software runs slower than “close to metal” code, but that wouldn’t be a problem for the simulated beings, who, to an outsider, would just live in a slower world. It would be like some sort of time dilation effect. That wouldn’t impose any extra computational requirements to the bottom-level simulators. Eve Online does something like that to balance the extra load when a server gets overcrowded. For the players time slows down and the servers have more time to process everything without crashing. It might be that we ourselves, if we live in a simulation, exist in a down-clocked universe. That the physical laws that govern our universe are slowed down. If that were the case, how could we know?

So, what would be the tick rate for our universe? Someone on reddit suggests it’s 1.86x10⁴³ hertz, which is how many times a second light can travel the Planck length distance, which is “the smallest physically meaningful unit of distance”. Other’s argued that time is not discrete but contiguous and Planck’s time is meaningless. Could time appearing contiguous be an argument against the simulation argument? Maybe? I’m not a physicist nor a computer scientist, so I don’t know. It could be that to sims, time appears contiguous, even though the simulation runs on discrete ticks.

Another question that should be asked is whether these posthumans were ever humans at all. Right now, we simulate all sorts of realms and creatures in video games that have little similarity to anything in the real world. How could we be sure that our simulators themselves aren’t entirely alien beings and live in a universe that looks nothing like our own? And if they themselves look nothing like us, then it should be wrong to expect that their motivations are recognizable to us too. They might create simulations not for entertainment. They might create them to get the flaggaroobizer to niquiflabate the voobfoooloos so that they rudufudize in the right ploogs.

Let’s be clear, our universe is still a part of their universe and if their universe, in turn, was simulated, it would be part of the universe that simulated them. All these sub-universes would be still part of a single bottom-level universe, so if you think all these simulations are an example of parallel worlds—technically they aren’t.

But the idea of a enormous regress of simulations is interesting, something like Turtles all the way down. This reminded me of a little poem.

Siphonaptera by Augustus De Morgan (slightly modified):

Big Sims have little Sims on computer racks to byte 'em,
And little Sims have lesser Sims, and so, ad infinitum.
And the great Sims, themselves, in turn, have greater Sims to go on;
While these again have greater still, and greater still, and so on.

---

Thank you. Please follow and resteem if you enjoyed this post.

@sildude

Photos used: Pexels.com

Thanks @alexdory for just being a positive role model

---

References

¹ Bostrom, Nick. 2003. Are you living in a computer simulation? Philosophical Quarterly, 53 (211), 243-255.
² Chalmers, David J. 2005. The matrix as metaphysics. In Christopher Grau (ed.), Philosophers Explore the Matrix. Oxford University Press. pp. 132.
³ Hanson, Robin. 2001. How To Live In A Simulation. Journal of Evolution and Technology, Vol. 7.

---

Additional links

Link to Nick Bostrom answering frequently asked questions.

Here be a video that you might like. Many of the ideas you just read come up. That’s Chalmers on the left there. The woman on the right is Zohreh Davoudi, who co-authored a paper about how the simulation idea could be compared with lattice quantum chromodynamic simulations, which is a technique that scientists use to simulate universes (well, kinda) down to the quantum level. Problem is that right now the simulated areas are only femtometers across. That's 1x10^-15 meters.