Would the world be less complex and enticing if reality were a mere simulation? It may well be that the simulation hypothesis is all wrong and of no import. We all know what computers can do given their vast computational powers and the onset of mind-blowing artificial intelligence. We have come to appreciate fabulous hyper-realistic deepfake images and astounding videos. Why not expect simulations to appear soon and enrich this synthetic mix?
So many advancements and so much hype under the bridge so far! Questions about simulated worlds are always followed by issues of actually living in them. Odd arguments in conjectural articles offer evidence or proof if you can believe it. Take a look at “15 Reasons Why we Might be Living in a Simulation“, “It is Confirmed: We Live in a Simulation” and see multiple examples on speculations on a brave new world for mankind.
Dissecting the Simulation Argument
No matter what these articles say, there may or may not be some actual science lurking behind the hypotheses. One would assume at least some degree of expert examination. As a particle physicist, Tim Lou, PhD has reasons to doubt it. He finds most premises for supporting the simulation hypothesis are rather dubious. In fact, they often contradict what we have come to accept as fact.
Lou outlines the most common arguments in favor of simulation theory in a nutshell. One is Moore’s Law. It states that computational power is undergoing rapid and exponential growth.
Next, an argument of note states that our apparent world operates according to simple, logical laws. These are easily imitated by powerful computers. From this you can extrapolate that a simulated world would resemble our own if created by future computers.
A final top argument postulates that there are more simulated worlds than “ordinary” ones. In fact, our current existence is part of a simulation.
There they are: the most common arguments. In Lou’s opinion, “They hold no water. I can counter each and every one.”
First, let’s take the exponential growth proposition. Systems exhibiting such growth are bound to slow down. Just consider virus transmission or population migrations.
Second, Moore’s Law may be empirical by nature, but why should the trend march on if it lacks adherence to fundamental laws.
Third, there is no reason to assume that simple physical laws are automatically amenable to simulation.
We see that premises one and two are false so we can conclude logically that computers cannot mimic our world in any real way at all. Not even close! Going on in this vein, one to three propositions are inherently weak and the entire argument fails.
As a physicist, Lou turns to premise three, as it has been underrated and ignored. I like the fact that it is related to nature’s physical laws that can co-exist with the concept of simulation. Simulated worlds couldn’t possibly copy or create such a complex construction as our world reveals. Some are, in fact, hidden from view.
Living in a Complex World
We have to ask: how do we know about our world? We know things from empirical observation as we walk around and notice objects. We live by our human-centric tasks that entail human scale, macroscopic things. If we are reading, we do not pay attention to the ink in the book or pixel patterns on the screen at hand.
We ignore such things in favor of what matters. In short, we zoom in on what is relevant. Of course, these other things are right there in front of us all time. The importance of things is therefore relative. They could matter to microscopic forms of life.
We are constantly inundated by stimuli so we have to block most of them out. Simulated realities work the same way. In games and situations of virtual reality, only macroscopic details are emphasized. If things seem accurate in a simulation, why complain? It has done its job enough to satisfy the average human eye. It is an efficient way of doing things since we don’t care about what we don’t notice.
In short, we get the macroscopic level of reality as dictated by the computer. This is one huge distinguishing difference between simulations and true reality. You need a magnifying glass in real life to get deeper into things, like ink patterns on a book’s page. If more details are needed for some reason, use more magnification such as a microscope. Do as a physicist would and put the item in an accelerator to expose its subatomic world.
A computer might be asked to do this same job. It can’t look at the atomic scale of things yet, but who knows….
Physicists can do a lot in zooming in on reality. They can go beyond it to the furthest reaches of space, where nothing is really observable: perhaps more than one hundred billion lightyears out. On the other hand, they can go in on an infinitesimal scale too. The tiniest subatomic particles are “visible” in a way at one quadrillionth of a meter, a quark’s size.
At this level, close to 10⁴⁰, there will be no sign of pixelation or glitches. This dwarfs the imagination of the layman, if not the physicist himself or herself. Computational power is improving and will surely delve deep soon. We will come to understand the magnitude of the universe on a laptop, maybe.
It is pretty hard to believe life as we know it is a mere simulation given what physics tells us from the greatest magnitude to the smallest subatomic particle. Then there is the matter of dynamics and the thirteen billion years of its evolution in our universe. It makes the simulation hypothesis wane in importance or plausibility.
In terms of dynamics, which defines the world, changes are constant and immeasurable at every level from the microscopic to the greatest macroscopic. Unpredictability is predicted so to speak. It is built into things like weather patterns, gambling results, the ups and downs of financial markets, and more. Uncertainty is the hallmark of modern society.
Maybe it is just that we don’t know much at all. Our uncertainty stems from our ignorance. Do we understand human psychology enough to know how it impacts buying and selling decisions on the stock market? We can’t possibly fathom the dynamics of billions of molecules that impact the weather. We can’t track such complexity.
Chaos Theory for a Chaotic World
We may not know it all, but we recognize patterns of unpredictability. Chaos theory talks about what is irregular and in flux. Most complicated systems exhibit chaos. No sparse mathematical equation can explain it.
We can only track and look at special properties or symmetries within them and try to find a physical law that applies, like energy and momentum conservation. We then come to see certain characteristics: patterns do not repeat, a possible configuration is likely to be reached, if only approximately, and even the smallest disturbances will eventually lead to some rather big changes.
We realize that “chaotic events” scramble information, just like the pseudo-random number generators in your computer.
the Butterfly Effect comes to the fore in this context. As suggested by its name, the theory advances the idea that small changes (like the flapping of a butterfly’s wings) will lead to impressive outcomes of a dramatic nature such as storms or hurricanes.
A creator of a simulated reality would want to try their hand at mimicking this. If they do, then the simulation will have to reflect chaotic systems and their exponential growth. It is an inevitable “error”.
Any error, even the tiniest, will be devastating as a result. We think of computers as enjoying precision, but it is actually quite finite. In no way can a computer predict the outcome of a chaotic system. There is a kind of built-in uncertainty as in weather systems.
There are some conclusions to be drawn. Even the most powerful machine will fail in its attempt at a simulation of a chaotic system. The errors of course continue to multiply endlessly. In short, you cannot simulate the real world, including the flapping wings of a butterfly and their impact on life.
All the resources in the world would be needed to get a grip on both the microscopic and macroscopic levels of a system. The results can never be perfect! So why go on with such simulations?
We do what we must and rely on statistical estimates and generating random numbers. The layman will never know what is wrong or left out. The layman will accept “probabilistic outcomes and some degree of quantifiable uncertainties (as with weather and stock market predictions). Simulations are a long way from reality and will likely remain incomplete for some time.
Taking a Simulation Perspective
You can argue all day long for and against simulations, but there will likely be some flaws. For example, in using physics, perhaps physical law at present is wrong. We assume that it doesn’t apply to simulations. A simulated reality would follow some other version and any laws that would have been tested are no doubt approximately true.
The physicist can measure numbers with finite precision such that he or she can expect errors waiting in the wings. Any and all physics experiments have glitches or deviations no matter how hard we try for full precision.
Let’s take one example from 2015 when LIGO made a startling discovery about gravitational waves from the merging of two blackholes. It garnered the Nobel Prize for this in 2017!
It was about measuring tiny distance variations to the tune of 1/10000th of the width of a proton. Gravitational waves from over one billion light years of distance cause these variations, fully in line with Einstein’s equation.
There were no glitches here, but they may be detectable at some time in the future if money is to be spent looking for them.
We have to admit that the laws of physics are not perfectly compatible with experiments, much as we would like them to be. They can be theoretically elegant and mathematically consistent, of course, but not more.
According to Lou, we will fail if we apply these laws to a “simulation clause”. There is no real need to do so in any case. It wouldn’t help with predictability and it could wreak havoc with the value and simplicity of physical theory.
Looking at it from the Perspective of Omnipotent Simulation
We can counter the above arguments, stating that physics experiments may be simulated. We can zoom in all we want and take a close look to arrive at detailed observations and it is all a fake. It just mimics real time in the lab.
We are not heading outside the territory of simulation hypothesis here. We can’t explain away inconsistencies in physics experiments by saying they are simulated! Even thinking about simulation is a part of a simulation.
Those deep into simulation theory won’t be able to trust their own thoughts. You are in a perpetual state of questioning and confusion. Are physical laws absolute or to be abandoned?
If we let them go, what are we going to gain by it? A better simulation? We will also lose predictability and the ability to examine facts, looking for falsification, which are part of explaining natural phenomena.
Lou would call this an omnipotent version of the simulation hypothesis and finding it lacking in scientific value. What does have value is quantum mechanics, beyond classical physics. It is the answer to the nature of our world and the magnificent hydrogen atom.
Classic Quantum Physics
The arguments posed above are not invalidated by it by any means. Classical physics holds sway. What is different about quantum physics accepts incomplete answers. We yearn to know more, so why not about a quantum simulation?
What is holding this understanding back? For one, we lack a working and scalable quantum computer. Second, we fail to fully comprehend how it all works – such as gravity.
After so much time and effort, science has failed to come to a clear explanation of the quantum world. Thus, why bother to ask about simulations?
Instead, why not mention what we know and accept for sure? We can agree on a few things:
For various reasons, quantum mechanics contains hidden complexities. It is difficult to simulate this with a traditional computer. This could be mitigated, of course, by a quantum computer.
They are not the same animal. In fact, they are restricted by mathematical theorems. Take the no-cloning equivalent of no-deleting theorems. Again, the concept of simulation seems paltry and absurd.
Drawing Some Conclusions
Arguments may come from the die-hard side of the issue who maintain that many of these intricacies about quantum mechanics work with the simulation hypothesis. Lou’s question is this: given our definition of a quantum simulation, would it be apropos? Physics is a natural science that requires things to be specific and predictive to have real value. Can we say this about simulations?
The simulation hypothesis may not be amenable to scientific scrutiny at all, hard as we try to make it fit. We are relegated to comparing the natural world to computer simulations.
It doesn’t matter that our computational power is impressive. We have made enormous progress, but it in no way enables the kind of simulations we envision. It can’t manufacture the same scale, dynamics, complexity, inherent chaos, and adherence to quantum mechanics. As a candidate for a scientific hypothesis, simulation theory is not tenable.
Go ahead and call our world a simulation all you want. You would have to admit that some omnipotent power created it or the simulation would likely be meager at best done by a less powerful host computer than we would imagine.
The physics governing the simulation have to be significantly more complicated than the real world laws of physics to bear any scientific value.
Many published peer-reviewed fundamental physics papers have appeared on the topic. We can’t go so far as to negate the research in theoretical physics now and in the future. Some areas of simulation theory will surely inspire work of value.
But the current simulation hypothesis might remain inspirational, and not much more. It surely will not change our understanding of the universe as we know it. It takes a lot to gain new insight. Think of the LIGO measurements that try to capture the content of the universe in a classical sense. You have to contend with about 10⁴⁵ orders of magnitude.
What if each location could store at least one of the required quantum bits? there would be (10⁴⁵)³ = 10¹³⁵ bits. Lou estimate a lower number of bits in a simple way: “use about 10⁸⁰ number of atoms, with a precision of 10⁴⁵ for each location variable, requiring at least 100 bits to yield 10⁸² bits”.
We can look at it in various ways such as through the lens of quantum gravity. We can compute the number of quantum bits if the universe is in fact observable and an upper bound. This is on a par with black hole entropy, another area to examine in and for itself.
You can quickly calculate 10¹²⁰ quantum bits without absolute accuracy. No, don’t try to equate quantum and classical bits here! But we can see that the universe efficiently stores data. With the tool of black hole physics, we can compute this bound, but we will not arrive at how the bits are arranged for best efficiency. Quantum gravity still holds many secrets.