The simulation hypothesi.., p.14
The Simulation Hypothesis, page 14
Some specific rules that a rendering engine might follow include:
Do not render anything that is obscured, or behind other objects (that the user wouldn’t see)
Adjust the field of view by quickly adjusting the pixels that have already been rendered
Pre-cache the scenes adjacent to the current one (to make them quicker to load)
Calculate/render new pixels only when necessary
The basic idea of a rendering engine is to show what is visible from the virtual camera, which is placed at a specific (x,y,z) coordinate inside the virtual world and pointed in a particular direction.
The basic rule of video game rendering engines is the same as what causes the collapse of the probability wave, and you might say it is also the rule of QI: only render that which is being observed.
Multiplayer and VR Rendering
As we move from single-player to multiplayer, the idea of a shared 3D world presents a new level of complexity. Not only are there 3D models of the world, but there are many characters, each of which is making independent choices and may be in the same part of the world.
MMORPGs are still rendered on individual computers—as such, a “shared rendered world” doesn’t really exist. Each computer renders what is happening in the scene. If my character is present and yours is present, then both of our CPUs/GPUs will be rendering the scene based on shared information. Where is this information? It is both decentralized and centralized—it is sent from the client machines based upon every choice you make and then synchronized and sent to the other people that are in the same place that you are.
The same optimization techniques are used by VR headsets, which provide fully immersive 3D environments. It would be too slow to render the entire world (and consist of way too many pixels for current storage), so only the visible, non-occluded parts of the scene are rendered, based upon the position and perspective of the player.
The Simulation Hypothesis and Quantum Indeterminacy
This is where we start to overlap with theories of how video games are put together and how they might be used. Quantum indeterminacy, a fundamental principle of the material world, sounds remarkably similar to optimizations made in the world of computer graphics and video games, which are rendered on individual machines (computers or mobile phones) but which have conscious players controlling and observing the action.
“Quantum Indeterminacy,” says Elon Musk, in one of his many public statements on why he thinks the simulation hypothesis is likely, “is really an optimization technique.”
What does he mean by that? The reason we don’t know which path the electron took through the slits, or whether the cat is dead or alive until we observe it, is because there is a computer that is keeping track of all things in our physical world, and it doesn’t have the resources to render every possibility. It only needs to render those things that we, as conscious participants or players in the game, observe. Similarly, until the players actually make choices and log into the game, the gameplay interactions are only possibilities—like the quantum foam probabilities. In order to observe an event in a video game, we need it to be rendered.
Just as the subject and object can no longer be separated from each other in quantum physics, in modern video games, particularly MMORPGs, it is difficult to separate the player from the rendered world itself.
Philosophical Questions Raised by QI
Note that just because we cannot see everything doesn’t mean that it doesn’t exist. Or does it? When we have a 3D video game, we map out the world using 3D models. In some games, we allow user-generated content that stays in the world even after we log out of the gameplay session so that other players can see it when they log in. This information is stored outside of the rendered world. The virtual world is still considered a “shared reality” between all the players of a particular game.
A philosophical question that comes up in both quantum physics and in video games is that if no one is in a particular part of the 3D world — i.e., no one is observing it or no player is there — does the particular possibility exist?
Just like Schrödinger’s mysterious cat, which is neither dead nor alive until someone observes it, the world of video games relies on a player being logged in to render the world.
For example, what happens if there are no players logged into any of the servers of an MMORPG like World of Warcraft? The servers are running, but nothing generally happens until a player logs in to observe what is going on, not unlike quantum physics.
This is not an easy question to answer for either video games or for superpositioned particles or for Schrödinger’s infamous pet. One way to think about it is that all of the possibilities exist somewhere, as information, and it is a conscious choice by the player (or an automated choice by an AI) to go down a specific path that brings that possibility into actuality.
Let’s delve further into the idea of a game having unlimited possibilities and how we choose which of those to actualize in the next chapter. If QI wasn’t strange enough, the idea of parallel worlds sounds like downright science fiction.
Chapter 6
Parallel Universes, Future Selves, and Video Games
As we have seen in the last chapter, QI, or quantum indeterminacy, is a strange but fundamental feature of particles at the subatomic level. Since these particles ultimately make up our physical reality, they must have some impact on our macro reality. This is difficult for us to understand or visualize intuitively because it’s not how we think about the world at a macro level.
Whereas we are used to thinking of particles as existing as material objects at particular points in space (and time), quantum physics gives us a description that is more akin to a probability wave—corresponding to a set of positions that a particle might occupy; only through observation do we get to a single outcome. Until then, all possible values of the particle exist in some probability wave or set of probable worlds.
In this chapter, we’ll delve into another aspect of quantum physics that seems mysterious and baffling, the Many Worlds Interpretation, or MWI. In this interpretation of quantum physics, every time a choice is made (including an observation of a single particle), the universe breaks off into multiple parallel universes. Like many of the concepts in this book, this might seem like science fiction, but it is taken as a serious interpretation by many physicists.
We’ll also look at another baffling aspect of quantum physics: that something in the future can actually impact something that should have occurred in the past. This is counterintuitive to how we think of the flow of time. But in the world of quantum physics, the observation may not happen until the future, which seems to influence what the particle might have done in the past!
These ideas of future selves and multiple parallel worlds don’t make a lot of sense—unless we come back to the simulation hypothesis and think of the universe as a complicated video game with multiple possible moves and outcomes. We’ll pull together these two ideas by looking at possible (or probable) future worlds that exist in a video game or simulation. We’ll also ask if these parallel worlds actually exist or are simply probabilities in a probability tree that is created by whomever or whatever is rendering the world around us. We’ll revisit how AI is built into video games and how the basic idea is to scan possible futures for the best possible outcome given the parameters of the video game.
The Delayed-Choice Experiment
Let’s start with the counterintuitive finding that QI allows the future to influence the past, at least to an extent.
As we saw in the last chapter, the particle-wave duality was at the core of quantum indeterminacy. This was demonstrated again and again through numerous versions of the double-slit experiment carried out by physicists, starting with Einstein and continuing to the present day.
John Wheeler, a prominent theoretical physicist who was involved in many of the biggest developments in the new physics, decided to go further and added another set of choices after the two slits.
This was a backdoor way to pacify some physicists who doubted that the collapse of the probability wave occurred at the moment of observation. Instead, they said, the particle was already either a solid particle or wave when it went through the slits, and the measurement just told us which one after the fact. This would be akin to Schrödinger’s cat already being either dead or alive—it couldn’t be both! Opening up the box at some point in the future simply “reveals” to us which state the cat was in!
Wheeler described an experiment called the delayed-choice experiment. In this experiment, which builds on the original double-slit experiment, after the particle has gone through one set of double slits, there is a mirror that reflects the individual particles in another configuration. The idea is that if the particle was already in a “particle” state when it went through the slits, then its behavior would be different from that of a “wave.”
Wheeler posited this delayed-choice experiment theoretically, but over time scientists have found ways to implement it. In one configuration of a delayed-choice double-slit experiment, shown in Figure 21, a lens is positioned after the slits so that the path of a particle diverges once it comes in contact with the lens. This means that, if a particle goes through slit 1 or slit 2, it will have a different destination—i.e., telescope 1 or telescope 2.
However, if there is a detection screen, called an “interference screen,” put in between the lens and the telescopes for an instant, it will show an interference pattern that can only occur if the particle goes through both slits—i.e., if the particle is a wave! However, the particle can’t end up in one of the telescopes unless it is a discrete particle that has gone through one slit or the other (not both).
Figure 21: An example of Wheeler's delayed-choice double-slit experiment.31
Wheeler and others concluded that the particles retained their wave-particle duality even after they went through the lens—but the matter wasn’t actually decided until a measurement was done at the telescopes. Two paradoxical results showed:
the presence of an interference pattern on the detection screen, which implied that the particles behaved as a wave even after they went through the lens; and
that the particles arrived at individual telescopes, implying that the particles had to be “particles” and not a “wave”—i.e., they had to go through one slit and hit the lens to end up at one telescope or the other!
Wheeler concluded that quantum particles are undefined until the moment they are measured, even if the measurement happens after the particle needs to make a choice of which slit to go into! Once again, the cat isn’t dead or alive until the observation is made!
Measurement in the Future vs. the Past
The strangeness of the delayed-choice experiment, which might also be called a “delayed-measurement” experiment, was further demonstrated by making the particles in the second part of the experiment travel over a much larger distance before they were measured.
In 2017, a team of Italian scientists did a delayed-choice experiment over a longer distance by bouncing lasers off of satellites and found that the results were consistent with the smaller-scale experiments.32
Even though the particle travels some distance (in this case, thousands of kilometers to the satellite) after deciding it is a wave or a particle, the measurement is done clearly in the future, from the perspective of the time the photon is first sent into the experiment.
However, until the measurement occurs, the researchers found that the photon still exhibited properties of both wave and particle. This means that something in the future (the observation) was influencing something in the past (a choice of whether it is a particle or wave when going through the slits)!
Some physicists have dubbed this concept retrocausality, which means the cause is in the future and the effect is in the past, though it remains an area of debate. Wheeler himself, who formulated the delayed-choice experiment, was hesitant to call it retrocausality because it would mean the future was influencing the past (more on another interpretation, the many worlds interpretation, in a bit).
Multiple Possible Futures?
Some physicists and non-physicists have taken this finding at face value and concluded that it is possible for an event (in this case, an observation) in the future to influence something that has happened in the past (passing through the slits). If generalized, this result would change our understanding of time, space, and our physical universe—something in the future can influence the past!
Physicist Fred Alan Wolf, for example, says that information from these possible futures is coming to us in the present and that we send out an “offer wave” into the future, which is interacting with the offer waves coming from the future to the present. Which possible future we navigate to depends on which choices we make (both in terms of our consciousness and observation) and how these two waves superpose on each other (or cancel each other out).
The implication that future probable selves are sending back information to the present and that we are consciously choosing which path to follow is startling and defies common sense.
Some interpreters who agree with the retrocausality interpretation of the delayed-choice experiment have taken this implication to the macro level, saying that there are a set of probabilities in which we exist both in the present and in the future. This means that the probability waves that we talk about in quantum physics are not just about where a particle (or at a more macro level, a person) might be in the moment, but where it might be in the future.
Figure 22: Multiple probable futures are sending us back information we use to make decisions.
What are these probabilities? Are they really future selves? Do they actually exist or are they just mathematical constructs? Most importantly, which of the possible paths do we follow in our lives with all of these probabilities?
These are difficult questions that are not really answered by the mathematics. One thing is undeniable, though: exactly how the probability wave collapses is still one of the biggest mysteries in quantum physics.
The best answer physicists have come up with is that consciousness somehow determines the collapse and that observation and/or measurement are part of the process. Whether that observation happens in the present or the past or the future, there doesn’t seem to be a way around this conclusion.
Parallel Worlds and the Multiverse
If the idea of the future sending messages to the past sounds like science fiction, then another popular interpretation of quantum physics, the many worlds interpretation, has launched a whole subgenre of science fiction and fantasy adventures.
Hugh Everett, a student of John Wheeler at Princeton, is the man most responsible for this interpretation of quantum physics. It is another way to look at quantum indeterminacy, and, in this interpretation, rather than a probability wave collapsing into a single, measurable reality, there are actually multiple universes that exist (or are created) each time a quantum choice is made. In this case, all of the probabilities are true—in different universes!
Neither Everett nor Wheeler use the term “many worlds.” The term “many worlds interpretation,” or MWI, was popularized by another theoretical physicist, Bryce DeWitt, in popular articles in the 1960s and 1970s. DeWitt wrote, in a famous explanation that could have been written by a science fiction writer: “Every quantum transition taking place on every star, in every galaxy, in every remote corner of the universe, is splitting our local world into myriads of copies of itself.”33
If every quantum decision creates different universes, then there must be an infinite number of universes with slightly different choices made along the way—including those with different versions of ourselves that have made different choices in life.
Now we simply refer to this idea as the “multiverse.” The multiverse theory is taken seriously because it offers an alternative to the idea of a subjective universe that quantum physics seems to imply. There are a variety of paradoxes that are explained by this theory. Schrödinger’s cat, for example, is now both alive and dead in two different universes!
Perhaps the most famous paradox that the multiverse theory was meant to get around was the “grandfather paradox”: if you could travel back in time and kill your grandfather before your parents are conceived, then your parents would never be born, and you would never be born. So how could you have gone back in time in the first place to kill your grandfather? In other words, if information can be sent from the future to the past, then it’s possible this information can change the past. By changing the past, it’s possible then to change the future, such that there was no need or possibility of the information being sent back in the first place!
The multiverse interpretation says that the first timeline is distinct from the second timeline. When you go back in time and change something, like killing your grandfather, then you are spawning off a different timeline, a parallel universe, if you will. While you may never be born in the second timeline, you were in the first.
There are many examples in science fiction that explore the multiverse theory, perhaps none as famous as The Terminator movie. In this film, the general-purpose killer AI, Skynet, sends back an independent autonomous AI (Arnold Schwarzenegger as the Terminator) to kill Sarah Conner so that her son, John Conner, who will eventually become the leader of the resistance, is never born. The long running British science fiction saga, Dr. Who, about an enigmatic time-lord who is able to travel in time as well as space, has explored this idea of alternate timelines in many different storylines.
