This post is an adapted and expanded version of the the talk of the same name that I presented at RefactorCamp 2019.
The material in this post comes largely from two books: David Deutsch’s The Fabric of Reality and Richard Feynman’s QED. The goal of this blog entry is to introduce you to the Basics Mechanics of our shared reality, starting with Quantum Physics. We’ll do this first with a short introduction to quantum mechanics and then talk about two particular aspects of quantum mechanics — observation and time. Finally, we’ll briefly cover what those two aspects tell us about our lived reality.
To accomplish this, my strategy is two fold. First, I need to convince you that the reality we live in is not a singular universe, but rather a multiverse. In this multiverse, everything that can possibly happen, does.
The next part will be to talk about how we end up experiencing this specific version of reality, of all the universes that are possible. What sort of mechanics does the task of universe selection encompass?
Welcome to the Multiverse
David Deutsch gives us a proof for the multiverse in Fabric, that I’d like to revisit for you here. Before we delve too deeply into it, let’s take a moment to briefly reconsider what a ‘proof’ is exactly, at least in the scientific sense
Generally, a scientific proof is done by setting up an experiment, executing the experiment, observing the results and then drawing some conclusions about the world based on those observations.
- Set up experiment
- Run experiment
- Observe results
- Draw conclusions
In other words, we derive our understanding of the world from observations that we make of it.
Deutsch’s Proof of Multiverse
David Deutsch’s proof of the multiverse relies on a fairly famous quantum experiment — the double slit experiment. In order to explain it properly, however, we first need to come to some conclusions about the ‘true’ nature of light, particularly at very small quantities.
Namely, at it’s smallest, is light a continuous or a discrete entity? For a human, it’s hard to observe light beneath a certain ‘density’ of beams, so to speak. If we were to set up a light and walk away from it, at some point it would go dark, as the amount of light reaching our retinas would be beneath the threshold required to register as light. For humans, then, light appears continuous. Either we can see it, or we can’t.
Deutsch, however, claims that frogs can see light at its absolute smallest amount. When a frog moves farther and farther away from a light, at some point the light begins to flicker. As the frog continues to move farther away, the flicker grows farther and farther apart. This is because light is a discrete, or discontinuous entity. There are individual photons of light emitted by the light source — the frog can still see them, however the rate at which they hit the frog’s eyes decreases as the distance between it and the source increases.
This ‘thought’ experiment (more like me telling you how it happens) is an illustration of the truth that light, at its very smallest, is discrete. These tiny light particles are called photons.
As an aside, it’s this very discreteness that gives ‘quantum physics’ its name. Quantum mechanics is a movement away from classical (pre-quantum) physics, where all values are continuous, to a quantized or discrete understanding of how individual particles move and interact. The movements of these ‘quanta’ are what quantum physics deals with.
Ok, so we’ve established that light, at its smallest, is a discrete unit. If you spread it out thinly enough, you’ll get a single ‘unit’ or photon of light.
Let’s get back to the double slit experiment now. If you’re already familiar with it, please bear with me as I walk through the setup briefly.
The idea is that you have a solid sheet of metal with two slits cut into it, and a screen (really an array of detectors) set up behind the metal sheet.
If I was to shine a flashlight through the slits, what would you expect to see on the screen? The answer is that you’d expect to see two beams of light that pass through the slits.
Now, if instead of a beam of light from a flashlight, what if we were to send a single photon through the slits? What would we expect to see then?
Common sense tells us that we could expect to see something similar to the flashlight experiment, except instead of a solid beam of light, it’d be a single flash.
This isn’t what happens. Instead, we see a bar-like pattern of light.
Here’s a more detailed photo, that gives you an idea of what it looks like from a single photon at a time view.
The question is why does what we observe not match up with reality?
One explanation that you might hear is ‘well light is wave-like’, we’re just seeing one of the consequences of its wave-like properties.
This is not a good explanation. It is true that light tends to display results that are easily modeled with equations with waves. But in terms of explanation for what we observe, it’s not very useful.
The problem with this is that we know, from our ‘experiment’ with the frog earlier, that light at its smallest is a discrete unit. In the dual slit experiment, we only send one unit of light through the slits at a time (the bar like pattern above and the illustration both show the expected results from repeating the experiment a lot).
What explains the wave-like pattern that we end up seeing? It appears that something is interfering with the movement of the light particle, something that looks a lot like a wave pattern.
Deutsch tells us that the interference is coming from the single photon that we’ve sent through the slits. That the interference is the photon itself. How can that be? There’s only one photon.
The truth is that the photon runs into itself. If you take the photon and you model all of the possible pathways that it could go through the slits and hit the back screen, you would get an output that looks much like the flashlight shining through the slits.
If you take the photon and you model all of the possible pathways that it could go through the slits and hit the back screen, and then run them all at the same time, and take all of the interference, i.e. places that the possible photon paths cross each other, and let them interact or bounce off of each other, you would get an output that looks much like the bar pattern that we actually observe through repeated attempts of the experiment.
Which is to say, that our observations confirm the model where the photon is interfering with (bouncing off of) itself. We can see the result of this interference in the bar pattern.
To reiterate: the movement of a particle through space interacts with all of its potential pathways through space.
Now, let’s call all of those ‘potential pathways’, as a collection, the multiverse. If you split one off, and just pick one of the collection, one possible version of events that can (and does happen), that one version would be a universe.
Each universe is not equally likely, for example it’s less likely that a photon will intersect itself and end up getting bounced way far away from the light source. In fact, you can see that this is a unlikely event in the bar-pattern photo above, given that those parts aren’t lit up as brightly.
Each of the pathways represent a universe option. We end up experiencing one of them. These different universes interfere with each other, and limit or expand what’s possible across all of them. The conclusion that Deutsch draws from this is that other universes exist and they shape what’s possible in this one, the one that we experience.
Physical reality is not a spacetime. It is a much bigger and more diverse entity, the multiverseDavid Deutsch, The Fabric of Reality
In the talk that I gave at RefactorCamp, I gave another illustration of how what we see is the sum of all the possibilities, using a light experiment from Richard Feynman’s book QED. Rather than going through it here, I’d rather show you the results that Feynman gives us from the double slit experiment.
Feynman is less interested in explaining what the double slit experiment means, however he makes a really good case for how the possibility must be present in order for the interference to happen.
If the universe we experience is one from a set of what’s possible, we can limit the chance of which ‘universe’ we’ll experience by limiting what’s possible.
You can see this in the double slit experiment by adding a sensor to each of the slits.
Instead of a screen, Feynman wants to know how much light that is sent through the slits will end up at a single sensor on the back screen.
Without the slit sensors, we observe anywhere from 0 – 4% of the light reaching a certain spot on the screen.
If you turn on both of the sensors, then there is a steady amount of light that reaches the detector on the screen — exactly 2%.
Why is this? Well, we said earlier that interference happens between the possible paths of the photons moving through the slits, in every possible universe. When there are sensors placed at the slits, however, suddenly there is only one possible way that the photons could have gone through the slits — the one where the sensor was triggered. By adding an observation to the experiment, we’ve removed possibility. We’ve fixed what universe we’ll end up in from a chance to a steady given. There’s only one way for the light to reach the detector, and the results reflect this fact. Consequently, the interference disappears.
What happens, however, if we were to use a faulty sensor? If sometimes it goes off when a photon moves past it, but sometimes it doesn’t. What would you expect to see?
It turns out that the oscillation of observations flattens out, to exactly match the incidence of the faultiness. The faultier your sensors, the closer the wave will be to the sensor-less version of the experiment; the more reliable the sensors, the flatter the line.
This tells us that observation has an effect on what possible universes we might experience. By making observations about the system, we remove the possibility for interference and narrow the result set.
So what? Why does it matter that there’s a bunch of possible ways that the world might go, and that we end up experiencing one of them?
Well, if every possible version of reality happens, it’s interesting to think about our odds of ending up in any one of them. Further is there a way to expand the set of universes that we might end up in? What about limiting them?
We know that everything that is physically possible happens. And that we might experience a subset of this. Are there any things that definitely happen in all universes that we might experience?
Sure there are. We know them as ‘universal constants’. One such example is ‘c’, the speed that light travels through space at. Feynman tells us that light can (and does) travel faster or slower than ‘c’, but for the sake of calculating probabilities of where and when light will end up, it tends to average out to ‘c’ speed.
In fact, this job of ‘what is generally constant between the universes’ is the realm that physics, as a field, deals with.
Observation and Time
I’d like to propose that there are two mechanisms by which universe selection happens: Observation and Time. What do I mean by ‘universe selection’ though? Well, there’s a set of universes that we might end up in. What might expand the set or reduce that set of possibilities?
Which is to say, I’d propose that time and observation are two mechanisms that have the potential to alter the set of possible universes that are reachable from our present universe.
Let’s talk a bit more about this constant ‘c’. What is ‘c’ exactly? It expresses the rate of movement of a photon of light through space. It tells us how much time will pass as light moves through space.
In some sense, then, time is a boundary for how far that light can move. Like how a pawn’s moves are limited on a chess board, time bounds what’s possible in the next moment.
Time is a boundary, it dictates what’s possible.
What does this time limit tell us about universe selection? As it is impossible for light to travel faster than ‘c’, any universe where a human standing on the moon can instantly know about what has happened here on Earth is not a possible universe that we exist in. It’s not in the set of possible universes.
There is a fun intersection here between time and computer science, as a field, that I’d like to illustrate with your help. I’d like you to solve this math problem using a pen and pad of paper.
Ok. What about this one? Three decimal places should be sufficient.
Now for this last one, go ahead an use a calculator.
Great. Thanks for playing along. You should have noticed a vast difference between how long it took you to solve the problem with a calculator versus the ‘easier’ one with paper and pen.
In Fabric, Deutsch introduces the concept of ‘tractability’. A tractable problem is one who’s answer or solution can be found within a timeframe that renders the answer actionable or usable. A tractable problem is one who’s answer is within reach, which in turn expands the set of possible universes to include one in which the answer to a present quandary is known.
Tractability is an important feature of computer science research. Discussions around ‘runtime’ or ‘big O notation’ are conversations of tractability. In classic computation, this is a counting of the number of steps it will take to reach a solution to a problem, often expressed as a function of the number of inputs to a problem.
What you were doing when you used a calculator to solve a division problem is an example of computational tractability. The computer did as many steps, or possibly more, than you would have to reach an answer, however it did them in a fraction of the time that it would have taken you. It made the solution more tractable.
Our ability to arrive at the answer to solutions more quickly expands the set of possible universes. Computers make problems that were previously intractable, tractable. In doing so, they change what universes are possible.
We’ve already seen how observation can serve to bound what’s in the set of possible universes in the earlier example with the double slit experiment with sensors.
There’s another aspect to observation that I’d like to talk about though, and that’s ‘how observable’ the universe is. Namely, can we observe everything?
What is Observation
First, let’s talk a bit about the history and nature of observation. You may have already noticed the word ‘observation’ or ‘measurement’ appear a few times in this essay. We can observe glass bouncing off a pane of glass. We observe what light passes through slits and arrives on a screen. We can construct experiments and measure the speed of light.
In fact, observation is the root of the enlightened era of scientific discovery. What we know of as ‘Science’ is based on our ability to conjecture, construct experiments, and then measure or observe what happens.
At some point, they added ‘run the experiment a few times’, and then do some statistics. Well, we do live in a provably probabilistic universe. It’s good to check that a result from an experiment wasn’t an outlier or a universe possibility that doesn’t occur very often.
As an aside, I do wonder if this possibility set of universes explains why we require science experiments to be repeatable. Without repetition, it’s possible that the conditions that an experiment were run under existed in one possible universe that is, on a whole, not very likely. Requiring repeatability helps us gain confidence that the result is generalizable across the varying set of possible universes.
Anyway. The process of measuring the results worked all well and good until we reached the quantum level. There, physicists ran into an observational roadblock, called Heisenberg’s Uncertainty Principle.
Heisenberg’s Uncertainty Principle
The general idea is that you have a particle that you would like to measure, or observe. Specifically you would like to know the speed that the particle is traveling and it’s current position.
Physicists discovered, however, that this is physically impossible. It is impossible to observe both the location and momentum of any given particle at the same time.
You can know where a particle is, but you will have no idea where it will be next (as the speed is unknowable). Or you can know how fast it is traveling, but you will have no earthly idea where it is!
This is huge. This is unsettling. Imagine being a scientist in this era, when science was the art of observing things. We believed that we could know everything that there was to know about reality, if only we could observe it. So we set out to build better tools for observing all of the things. And we looked at smaller and smaller particles and eventually we got to a point where we can see the smallest things except, we did some math and we discovered that we cannot actually observe all of the things.
This changed science. This changed what we know about the universe.
There are aspects of reality that are physically impossible for us to know, to observe at least simultaneously.
Which brings us to our next point about observability — there are some aspects of reality that are unknowable. This is classification that Deutsch makes as well.
If I were to roll a dice, cover it up, and then ask you which, of the six possible universes, we ended up in, you could guess and have a 1 in 6 probability of accurately deducing the nature of your present universe. It is easy for us to discover which of the six possible universes that we ended up in — we merely have to uncover the dice. This is an example of an aspect of reality that is knowable but not observed.
An unknowable, unobservable aspect of reality would be both the momentum and position of a subatomic particle. This is unknowable information, largely because it is unobservable.
Observability x Tractability
We can take observability and tractability and make ourselves a 2×2.
The upper left quadrant, of tractable and observable things defines the set of universes that we can predict or expand or contract via finding new answers or making new observations.
We can bring more of the set of universes into this quadrant by building new tools that allow us to measure more things or by improving our computational tools that allow us to make difficult problems tractable.
I’d like to wrap up this discussion of possible universes with a short discussion of these ‘sets of possible universes’ and how we, as humans, relate to them.
I gave a very short talk on ZKP through the lens of quantum rhetoric at the Stanford Blockchain Conference earlier this year and afterwards someone wanted to talk to me about jumping off buildings.
Specifically, they wanted to know that since the multiverse was real and really did exist, wouldn’t they survive if they jumped off of a building?
Sure, they would die in most of the universes that resulted from such an act, but they’d just continue to exist in the one where they didn’t die from the fall. Right?
I would officially like to be the one to tell you that the multiverse offers no such guarantee.
To really drive this point home, I’ve made a Venn diagram illustration that shows the guarantees offered by such a situation. In one bubble stands for the set of possible universes where you jump off of a building. The other is the set of universes where you live to see tomorrow.
Notice how the bubbles do not intersect.
I’m not saying that it’s not possible that you’ll survive a fall. I’m saying that the existence of a multiverse is no guarantee that you will. The multiverse is limited to what’s physically possible! You might jump off of a building where it is physically impossible to survive. Sorry.
Ok, so no guarantee that you will survive a building fall in the multiverse.
But what I think is really powerful about this question, is that it illustrates the marvel that is the mind. Our minds can *imagine* a set of possible universes where you don’t die when you jump off of a building. Here’s a Venn diagram of the universe set that I imagine that man had in mind when he asked about surviving a fall.
Isn’t that pretty wonderful? I think it’s pretty profound that humans are capable of imagining things that aren’t possible.
I think it also tells us something about gossip and online communities and our capacity to build VR and games which have their own internal logic and interesting stuff going on. These things are interesting but ultimately, they are not reality.
Just because you can imagine it or believe it does not mean that it necessarily says anything *real* about our reality. Nor does it necessarily make it part of the set of possible universes!
I’d like to leave you with the following.
There are a lot of possibilities of universes to choose from, or that might possibly, physically exist.
A single photon of light itself contains trillions or more different, potential locations. And our physical reality is constantly changing — this is one of the implications of the second law of thermodynamics, with how particles of matter are constantly pushing themselves towards chaos. This constant movement means that the nature of our particular universe that we have ended up in, presently, is a constant question that we can be asking ourselves.
The task of existing in a multiverse, then, is to discover the nature of the particular universe that you and me and all of us ended up in.
More interestingly, who or what decides what universe we end up in? Is it at all changeable?
I’d like to hear back from you what you find out.