Your complimentary articles

You’ve read one of your four complimentary articles for this month.

You can read four articles free per month. To have complete access to the thousands of philosophy articles on this site, please

SUBSCRIBE NOW

Articles

The Simulated Universe

Brent Silby asks, is this the real life, or is this just fantasy?

The Simulated Universe Argument suggests that the universe we inhabit is an elaborate emulation of the real universe. Everything, including the people, animals, plants and bacteria are part of the simulation. The simulation also extends further than the Earth: all the planets, asteroids, comets, stars, galaxies, black holes, nebula and other space debris are also part of the simulation: in fact our entire Universe is a simulation running inside an extremely advanced computer system designed by a super intelligent species which lives in a parent universe.

In this article I will provide an exposition of the Simulated Universe Argument, and explain why some philosophers believe that there is a high probability that we exist in a simulation. I will then discuss the type of evidence that we would need to have to determine whether or not we exist in a simulation. Finally, I will describe two objections to the argument before concluding that although it’s interesting, we should reject it.

The Possibility

The possibility that we exist in a simulated universe is derived from the idea that it is possible for a computer to run a simulation of any system that follows a pre-defined series of rules. Since the universe is a rule-following system that operates according to a finite set of laws, it follows that it too can be simulated by a computer.

The proponents of the Simulated Universe Argument further say that if it is possible to simulate a universe, then it is likely that we actually exist inside such a simulated universe. Why? Well, as Nick Bostrom claims in his 2003 article, ‘Are You Living in a Computer Simulation?’ in Philosophical Quarterly, No. 211, if it becomes possible for us to build such a simulation, then we will probably do so at some time in the future, assuming that human desires and sensibilities remain much the same as they are now. He then reasons that any species which evolves within such a simulation will probably build their own simulated universe. We know that it is possible for them to do so, because they themselves exist inside a simulated universe… and it is possible to continue this nesting of universes indefinitely, each simulated universe eventually spawning intelligent species which build their own simulations. Given the near infinite number of child universes, it is more likely that we exist in one of the billions of simulations, all creating their own simulations, rather than in the single great-grandparent universe.

So how would it work? Well, you can’t switch on a video screen to peak inside the simulated universe. The computer running the simulation does not contain virtual reality versions of people who live out most of their lives in their ‘real’ world. It is not like playing a videogame such as The Sims or Second Life. From the outside looking in, all you see are the hardware and numbers, and that’s all the simulation is – a complicated manipulation of numbers. The numbers are stored on permanent storage devices equivalent to hard-drives, and moved into RAM to be operated upon by Central Processing Units. The behaviour of the numbers in a simulated universe program demonstrate the laws of physics in that universe, and they also represent all the matter and energy. As the program runs, the numbers are manipulated by algorithms which stand for the laws of physics. This manipulation yields yet more numbers, which continue to be operated upon by the program. Large data structures are thus moved around within the computer’s memory, and they interact with other data structures. As the simulated universe grows, these structures become increasingly complex, but the laws that govern their behaviour remain constant, as they’re all part of the program.

So from the designer’s point of view the simulated universe contains nothing other than complicated data structures manipulated by the program. But for the creatures who exist inside the simulated universe, it’s all real. They marvel at beautiful sunsets. They walk around in the open air and enjoy the smell of freshly cut grass. They may study the stars in their sky, and dream about one day visiting other worlds. For the inhabitants of the simulated universe, everything is as solid and tangible as it is in the real world; but it’s all reducible to numbers and rules.

It is important to note that the computer is not thought to be simulating every potential subatomic particle in the pseudo-universe. As Bostrom points out, it wouldn’t be feasible to run a simulation to that level of detail. He suggests that the simulation need simulate to a high level of detail only local phenomena. Distant objects such as galaxies can have compressed representations, because the inhabitants would never see them in enough detail to distinguish individual atoms.

We can take this point further. Perhaps the entire universe is compressed in some way, including local phenomena. The simulation could be interpreted by its inhabitants as being comprised of atoms and subatomic particles, while actually it’s not so modelled. We can see that this is a reasonable possibility in modern physics. Consider the indeterminacy principle in quantum mechanics. An observer cannot simultaneously measure eg the position and momentum of a particle. Furthermore, it seems that subatomic wave-particles do not have a definite position or momentum until their position or momentum is observed. This is because subatomic particles do not exist in the sense we are used to experiencing things on the macroscopic level. Given the fact that we do not directly see subatomic particles, we can conclude that what we understand about them is an interpretation of a reality of which we have no direct access. In a simulated universe, this reality could take the form of data arrays which only demonstrate subatomic particles on the rare occasions when the inhabitants are resolving their experience of the universe to that level – for instance, when doing quantum experiments.

The Original Simulated Universe

The Simulated Universe Argument is not new. Frank Tipler put forward the idea of a simulated universe in his 1994 book The Physics of Immortality. He suggested that we may all become immortal when we’re recreated inside a simulation of our universe in the distant future.

Tipler argues that at some point in the future, humans, or some other advanced species, will inevitably develop the technological ability to simulate the universe. According to Tipler, any species which reaches such a zenith of evolution will have an extremely advanced sense of morality. They will recognize a moral problem with the notion of intelligent, self-aware beings dying, so to correct this problem they will recreate every conscious being who came before, and let them live an immortal life inside a simulated reality.

There are fundamental problems with this idea. The first and most obvious problem relates to the alleged moral dilemma that these super-advanced humans/aliens find themselves in. Why should we assume that there is a moral problem with people dying and no longer existing? Sure, from our perspective it seems a pity; but from the perspective of a species with a super-evolved moral sense, it may in fact be more morally problematic to recreate us.

The second problem with Tipler’s idea is one of implementation. In order to recreate humans that once existed, our future recreators would require knowledge of each person’s unique properties, including their personality, their memories, and the structure of their brains. It is unlikely that any future species would be able to gather this sort of precise information. The best they could do would be to create a new universe from scratch, switch it on, and hope for the best. Their simulation will then unfold according to whichever preset collection of rules that they have built into it. Their universe will evolve, and after time planets may form within it. Life could evolve on those planets, and one day become intelligent enough to build its own computer simulations of the universe.

How Would We Know?

If a simulated universe can provide a perfect replica of a real universe, how could we ever know that we exist in a simulation?

One way to find out would be to appeal to probability. As stated earlier, if we accept the possibility that advanced beings can create a simulated universe, then it is highly likely that we actually exist in one, as there will be billions of simulations but just one real original. So statistically there is a higher chance that we exist in a simulation than in the original universe.

Another way to determine whether we exist in the original universe or a simulation would be to look for clues or hints that this is not a real world. Such clues might come in the form of imperfections in the simulation. It is unlikely that we would find an obvious imperfection, such as a fuzzy border on the other side of a mountain. Imperfections in a simulated universe would likely be subtle and almost undetectable. They would be found in the laws of physics.

In 2001, physicists Paul Davies and John Webb published a discovery that has been interpreted by some as just such an imperfection. Their discovery came from observations of distant astronomical structures known as quasars, which pump out energy at the far edge of known space. Now, because information travels to us at the speed of light, looking at quasars effectively means looking far back in time. Thus Davies and Webb observed quasars as they were billions of years ago, and so discovered what could be interpreted as a change in the speed of light. They observed a change in the so-called fine structure constant, which is a ratio involving the speed of light, the charge on the electron, and Planck’s constant (a value integral to quantum physics). Webb admits that they cannot definitely say which aspect of the fine structure constant changed, but it could be the speed of light.

Regardless of which aspect of the fine structure constant has changed, the discovery is significant. This is because such basic physical constants are taken to be universal and unchangeable – they are built into the laws of physics. So, because these are fundamental to the laws of physics, evidence of a shift (or ‘glitch’) in any of these basic constants could be seen as evidence that we live in a simulated universe.

There are other explanations for the Davies/Webb observations, of course. Some theorists believe that the speed of light has been dropping since the beginning of the universe, and that it was once 10⁶⁰ times its current speed. It is possible that this reduction in speed is caused by a cosmos-wide change in the structure of the vacuum (see Recent Lightspeed Publicity (2002) by Barry Setterfield, www.ldolphin.org/recentlight). Or perhaps the space/time continuum is stretching in some way. Alternatively, perhaps the multidimensional gaps between superstrings are changing. There are many possibilities, but the point I am making is that this type of observation is precisely what we should look for as evidence that we live in a simulated universe.

Problems With The Argument

The Simulated Universe Argument relies on the assumption that future humans, or some other advanced species, will have similar desires and sensibilities to current humans, and will therefore want to create a simulated universe. In this section I will examine problems with this assumption. I will then suggest that the Simulated Universe Argument should be rejected, because it unnecessarily clutters up our ontology.

1. The problem of morality

The first problem with the Simulated Universe Argument is related to the point made above in regards to Tipler’s theory of immortality. I suggested that future humans or aliens may not feel a moral obligation to recreate current humans. I would now like to elaborate upon this thought.

Given our current desires and sensibilities, it seems that if we could develop sufficient computing power then we would create a simulated universe. The whole Simulated Universe Argument rests on this assumption. If it is true, then it is likely that we exist inside a simulation. But we need to ask, would a super-advanced species with sufficient computing capacity actually create a simulated universe? If we accept for the moment that it will be technically possible for a future species to do so, we need to decide if they would want to do so. Would it be the morally right thing, to create a simulated universe? We can be very quick to state that it would definitely be the right thing to do – but that’s from our current perspective, and we are nowhere near advanced enough to do anything like it.

Bostrom argues that an advanced civilization would choose to create a simulated universe. He says humanity’s existence (or the existence of self-awareness) is of high ethical value, and as such the world would be a morally better place if an advanced civilization created a universe containing creatures like us.

But like all cultural phenomena, morality evolves. It is simply conceited to assume that our current state of moral reasoning will remain unchanged. Highly advanced civilizations may find it morally abhorrent to create a universe and populate it with living beings who suffer as we do. For instance, we live on a planet full of creatures which have to destroy each other to survive. Humans, who arguably have the highest level of intelligence on Earth, kill animals, and out of greed and jealousy also each other, pollute the environment, torture children, tell lies, commit crime. Would an advanced species think it is a good thing to create a universe that could possibly contain this level of pain and suffering? It’s possible that a future species would rather choose not to create such a simulated universe, because doing so would increase the total amount of pain and suffering.

Furthermore, the assumption that an advanced species will want to create a simulated universe relies too heavily on the idea that they will share our moral standards. We cannot make such an assumption. So the likelihood that we exist in a simulated universe may be a great deal lower than suggested in the original argument. I am not saying that it is impossible: all I’m saying is that more thought needs to be put into what future moral reasoning would look like before we can rest the Simulated Universe Argument on this reasoning.

2. Are we replacing God with a Godlike species?

Another problem with the Simulated Universe Argument is that it suffers from similar problems to an argument for the existence of God – specifically, the Cosmological Argument.

Traditionally, the Cosmological Argument attempts to solve the problem of where the universe came from by stating that:

1. Everything that exists has a cause,
2. The universe exists,
3. Therefore, the universe was caused,
4. The name of the cause of the universe is God,
5. Therefore God exists.

The standard objection to this argument runs: If everything has a cause, then God also has a cause. The cause of God must be something equally God-like. Therefore, there must be more than one God. This does not fit the standard monotheistic view.

Supporters of the Cosmological Argument respond that there cannot be more than one God, so the God who created the universe must be either uncaused, self-caused, or existing forever. But here they run into difficulty. As soon as they allow that at least one thing can be uncaused, self-caused, or existing forever, then they open up the possibility that the universe itself could be uncaused, self-caused, or existing forever. Since we favour economy in our thinking, it is therefore rational to avoid invoking the existence of God to explain the universe.

The Simulated Universe Argument seems to suffer from a similar problem. By allowing for the possibility that we exist in a simulation, we open up the possibility of an infinite number of ancestor universes. Supporters of the Simulated Universe Argument may state that there is an ultimate parent universe, which was caused by a Big Bang or some similar event, or they may claim that the seminal universe existed forever. But these responses are the same as from supporters of the Cosmological Argument. They too stipulate that there is an ultimate, God, and that the chain of causes stops there. However, if this stipulation is unsatisfactory for the Cosmological Argument, it is unsatisfactory for the Simulated Universe Argument too.

Supporters of the Simulated Universe Argument may respond that there is a fundamental difference between their proposal and the Cosmological Argument, claiming that their Argument is different because it is based on the known existence of creatures who have a biology and use technology, while the Cosmological Argument is based on the mere speculation of a supernatural God. But I am not sure there is a difference. From the perspective of our knowledge there is no difference between a supernatural God and a super-intelligent species from an external universe: both are equally difficult to confirm the existence of. We can never know the nature of a parent universe, and creatures in a parent universe are unknowable, and from our perspective they are all-powerful.

Therefore, for reasons of ontological economy, I believe we must reject the Simulated Universe Argument. It creates an unnecessarily cluttered world-view. Why suppose that there exists a virtual infinity of parent-child universes, when we can simply assume that there is only one?

Conclusion

The possibility that we exist in a simulated universe is based on the assumption that if it is possible for us to create such a simulation, then one day we will do so. I have questioned this, firstly, on the basis that it assumes a future morality that resembles our current morality, yet our future moral standards may lead us to view such a creation as a highly immoral act. I have also questioned the argument on the basis that it is a variation of the Cosmological Argument, and so it suffers from the same problems. Accepting the possibility that we exist in a simulation allows for a virtual infinity of ancestor universes. Therefore it doesn’t answer any questions about the origin of the universe – it just shifts the problem. Furthermore, it clutters up our world-view by introducing a multitude of universes when just one is required. Therefore, although the Simulated Universe Argument is an interesting thought experiment, I believe we should reject the possibility that we exist in a simulation, and focus on discovering the origin of this, the actual universe.

Brent Silby is Learning Advisor in Philosophy, UPT School, Christchurch, New Zealand.