Simulation Theory Debunked

The most common formulation of the simulation theory (or more accurately "simulation hypothesis"), as coined by Nick Bostrom goes like this:

1. "The fraction of human-level civilizations that reach a posthuman stage (that is, one capable of running high-fidelity ancestor simulations) is very close to zero", or

2. "The fraction of posthuman civilizations that are interested in running simulations of their evolutionary history, or variations thereof, is very close to zero", or

3. "The fraction of all people with our kind of experiences that are living in a simulation is very close to one."

This trilemma tries to force us into concluding that #3 is true (we are most likely in a simulation), because #1 and #2 are most likely false.

Option #1

Option #1 of the trilemma assumes that high-fidelity ancestor simulations are scientifically equivalent to our current universe, in that all of our scientific discoveries could be equally explained by a high-fidelity ancestor simulation (a simulation that is super realistic and mimics the life of our ancestors). If there is a scientific gap between our universe and a simulated universe, then #1 is true (there are no advanced civilizations that can reproduce our universe in a simulation), and the conclusion that we are in a simulated reality is unfounded.

The Science Gap

• The Quantum Gap

• The Consciousness Gap

• The Brain Gap

The Quantum Gap

Quantum physicists will refute this point in that nothing about the way our universe works at the quantum level looks like a simulation. Simulations require a function based on systemically ordered computing to calculate an image for how reality should be rendered. Rather than systemic order at the base of reality, there is quantum chaos. Perhaps the simulation hypothesis apologist would move the goal post to saying that the quantum units are being simulated and programmed to be randomly chaotic. If this is the case, there is no guarantee that their simulations will randomly produce an "ancestor simulation", since our ancestors are the products of chance. Option #1 fails if this is the case, in that only a small number of random simulations will produce anything similar to an "ancestor simulation".

The Consciousness Gap

There is no guarantee that we can ever produce consciousness within a simulation. The philosophic hard problem of consciousness highlights this issue. If simulation theory requires the production of conscious entities, then the veracity of option #1 rests on whether or not it is ever possible to create conscious entities in a simulation. If simulation theory requires advanced races from base reality to inject their consciousness into the simulation, then option #1 fails in that only a small fraction of simulated entities will be able to be sustained by an injection of consciousness, as the number of conscious entities in base reality is not scalable with the number of simulations that can be run.

The Brain Gap

From neuroscience, we know that brains produce consciousness, rather than mediate consciousness. Consciousness is not a phenomenon external to the brain. This can be proven by a variety of experiments on the brain. We know, for example, that when you damage the brain, it alters your consciousness. We know when the brain absorbs certain drugs or chemicals that it alters consciousness. So, consciousness seems like an emergent phenomenon of the structure and state of the brain. Consciousness even turns off when sleep or anesthesia are applied. If consciousness is sourced externally to the simulation (consciousness from base reality is injected into the simulation), then we have a problem for explaining why external consciousness seems to be wholly dependent on the materials of the brain (the materials generate consciousness, not receive it). For more thoughts on the brain, check out my exploration of consciousness here - "Cosmic Consciousness" Analogy (Hard Problem of Consciousness) | TranscendentPhilosop (wixsite.com).

Option #2

Option #2 tries to pressure the reader into assuming that advanced races will be interested in producing "high-fidelity ancestor simulations". But there are many reasons to doubt this assumption.

The Interest Gap

• The Cost Gap

• The Videogame Gap

• The Why Gap

• The Moral Gap

The Cost Gap

The simulation hypothesis smuggles in the assumption that simulations are infinitely easier to run than finding universes in the wild. It could be very possible that the energy resources needed for running massive simulations are so costly that running large quantities of "high-fidelity ancestor simulations" is not desirable. High costs make simulations less desirable.

The Videogame Gap

There are a whole host of attributes that we associate with simulations. The most common type of universe simulation is a videogame. If you examine our universe, it doesn't meet any of the attributes of a videogame - 1) we aren't aware of our original selves, 2) we aren't aware of our purpose, 3) our lives end brutally and arbitrarily without any narrative to justify what you did wrong and how you should change in order to win the game, and 4) there are no points. We know that humans have a natural interest in videogames, but if our universe is not a videogame, then where does the interest in massive simulations come from?

The Why Gap

If our universe is assumed to be a "non-videogame simulation", then an important question to ask is why? Why simulate entropy? Why simulate consciousness? Why simulate an insane amount of empty space? Why simulate an insane numbers of dead planets? Why is alien life so rare in this huge simulation (Fermi paradox)? Why did they design it to be almost as boring as possible, despite committing a seeming infinite amount of resources to running the simulation?

The Moral Gap

The morality of this universe as a simulation is very suspect. I tend to think that morality is best defined in terms of collective meta-wellbeing. Wellbeing implies a bias to long-term sustainable happiness over short-term self-destructive happiness. Collective implies that the wellbeing of everyone in the group is considered (fairness). Meta implies that this is referring to a higher order agglomeration of wellbeing factors than include all possible psychological needs and values, resolving conflicts between utilitarianism and deontology by including the need for "rights" and "duties" as a psychological factor within wellbeing. Morality, defined in this way, necessarily requires that something is only good if the benefits to collective meta-wellbeing outweigh the costs. Collective meta-wellbeing is essentially based around conscious entities and their sensory experiences. When a simulation randomly tortures individuals with disease, this violates collective meta-wellbeing. When a simulation randomly allows children to die from accidents, this violates collective meta-wellbeing. Only an immoral intelligence would design a simulation that randomly violates collective meta-wellbeing for no reason. That begs the question as to what kind of telos for the simulation is necessary to justify the morality of the decision to run the simulation.

Option #3

Option #3 of the trilemma assumes that with 1) technology, combined with 2) interest, that it necessarily follows that 3) we are most likely in a simulation. This argument fails in the following ways -

Probability Gap

• The Detail Gap

• The Selection Bias - Cosmic Dust Analogy

• Natural vs Simulated - Manufactured Rock Analogy

• Occom's Razor

• The Law of Non-Contradiction

• Thought Experiment Testing

The Detail Gap

If the universe was a simulation, you would expect the level of detail generated by the simulation to match it's purpose. If there is a gap between observed detail and the purpose of the simulation, then the probability of that explanation being true is lowered. Video games don't program atomic level detail since that isn't relevant to the purpose of the game. Video games also don't program unnecessary planets. We should be able to posit potential teloses for the simulation and then compare observations against the telos for probability estimations.

The Selection Bias - Cosmic Dust Analogy

Assuming that the number of simulated universes is greater than the number of real universes, you could make the same argument about cosmic dust. Within the universe, cosmic dust is infinitely more common than organic matter, therefore when I open my fridge, I will see cosmic dust. The problem with this conclusion is that "opening the fridge" is not a random sample of matter in the universe. The fridge is an entity that contains selection bias - we have selected the only planet with life in the universe, we have further selected a home of an animal at the top of the hierarchy of life, and we have selected an appliance within this creature's home. These limiting factors make the result of "opening the fridge" incredibly biased. The contents of the fridge do not represent reality across the universe. Similarly, our universe is not a random sample of possible universes, including simulated universes. We are selecting a very biased sample from the list of possible universes. This is the universe that we evolved within. Simulated universes will vary in quantity and quality. We know that this universe is not a 2d videogame simulation, nor a 3d videogame simulation, nor a VR videogame simulation, nor a digital videogame simulation. We know it's not a philosophic zombie simulation (because consciousness exists). We know that it isn't a small simulation (based on the size of the universe). We know that it isn't a low detail simulation (quantum levels of detail everywhere). There are so many filters that can be applied to our current universe that it isn't necessarily true that we could be having our experience within any possible universe. It's quite possible that our experience is only possible within this set of filters. This is in fact the survivorship bias - one of the subtypes of selection bias. Our dataset (this universe) is the only "surviving universe" if you consider "the ability to carry consciousness" as the trait that is necessary to survive. All of the universes where consciousness failed to generate are not being included in our sample, making it less representative. If 99% of universes fail to develop consciousness, then the probability of consciousness is low, not high. If consciousness is rare in a natural universe, then it should be even more rare in an artificial universe. If most simulations fail to develop consciousness because of this compounded difficulty, then it is unlikely that we are in a simulation. This issue is further fleshed out in the Anthropic principle, and has been equally applied to the issue of why we exist on a goldilocks planet of seeming perfect attributes for life (Anthropic principle - Wikipedia).

Natural vs Simulated - Manufactured Rock Analogy

Imagine if you took a hike and found a rock in the trail, and someone told you:

• "The fraction of human-level civilizations that learn how to manufacture high fidelity rocks is very close to zero", or

• "The fraction of civilizations that are interested in manufacturing rocks is very close to zero", or

• "The fraction of all rocks that are manufactured is very close to one."

This argument seems plausible at first, but then when you go hiking, would you conclude that the average rock on the hiking trail was manufactured by humans? No, you would not. Because different types of rocks have their own probability clouds - a context in which they are more likely to appear. A manufactured rock is more likely to appear in a human's front yard. A natural rock is more likely to appear in a natural area. Also, this analogy highlights the issue with assuming the fraction of simulations to real units will necessarily be high, merely due to psychological interest. The fraction of simulated rocks to natural rocks is NOT high, because the quantity of natural rocks is infinitely greater than human capacity to manufacture. Just because it is easy to manufacture rocks doesn't mean it is easy or desirable to create infinitely more manufactured rocks than are currently naturally in existence. Similarly, there is no guarantee that it is easy or desirable for advanced races to create massive simulations in a quantity that far exceeds natural universes. We don't even know how many universes exist in total! It could be that natural universes are as common as natural rocks on a mountain. Simulation theory smuggles in an assumption that natural universes are hard to find randomly, whereas they have no evidence that natural universes aren't as common as rocks on a trail.

Occom's Razor

Occom's razor reminds us that the more claims that need to be simultaneously true in order for the hypothesis to be true, the less likely it is true. The probability of two things happening together is calculated by multiplying the probabilities together. For example, the probability of one coin flip resulting in "heads" is 50%. To get the probability of two coins resulting in two heads, the 50% is multiplied against 50%, resulting in 25%. If you add a third coin, you multiply the 25% by another 50% coin toss, reducing the total probability to 12.5%. So, the more factors that need to happen together, the lower the probability. So to run the actual probability of the entire "simulation hypothesis" claim, you need to combine the probabilities of 1) having massive simulations, with 2) consciousness, with 3) a telos that justifies suffering, with 4) a reason that justifies the lack of videogame original self-awareness, with 5) a reason for being so boring and arbitrary, with 6) a reason for making the simulation infinitely big, and with 7) a reason for it being worth so many resources to run, when a significant percentage of the conscious entities in the simulation hate it so much they would risk self-harm and pain to self-terminate and face an unknown hell or void just to escape a "simulation" that randomly tortures those within it.

Law of Non-Contradiction

Additionally, the more logical inconsistencies inherent within a theory the less likely it is to be true. The law of non-contradictions states that contradicting statements cannot both be true. People can try to solve the "why" gap and explain all the reasons for the simulation, but based on my experience, most explanations for "why" are bound to be contradicted by some aspect of the nature of the universe. The more contradictions, the less likely to be true. The more gaps, the less plausible.

Thought Experiment Testing Organon

An organon is just a logical tool to understand truth - an epistemological tool. Thought experiments are when you imagine a certain set of conditions to help you walk through the logical implications of those conditions. A thought experiment could be reduced to a formulation of "if x, then y". So, thought experiment testing would be first figuring out "if x, then y", and then subsequently looking at the evidence available to compare it with "y" to see if the evidence available is consistent with "y" or contradicts "y". If the "y" matches observations in reality, then the thought experiment has predictive value. Then you might want to dig deeper into the thought experiment to generate more predictions for testing. "If x, then y1, y2, y3, and y4". If every single prediction matches observations in reality, then that gives credence to the plausibility of "x" being true. If there is at least one contradiction, then "x" is don't have full explanatory power. Perhaps "x" will need to be paired with an "x2" to make the thought experiment more robust in order to handle potential contradictions.

Simulation Hypothesis Thought Experiment Test 1

If "this universe is a simulation that has a purpose of being interesting" [x]

then (->), "we should find interesting life on every planet" [y];

• Predicted reality = "Lots of interesting life everywhere"

• Observed reality = "There is lots of evidence that life (especially complex life) is rare in the universe"

Reality does not match the prediction, therefore this thought experiment test failed.

Simulation Hypothesis Thought Experiment Test 2

If "this universe is a simulation that has a purpose of testing human morality" [x]

then (->), "we should find that the entire universe is organized around humans and their morality" [y];

• Predicted reality = "Everything should exist for the sake of humans and their morality. Planets without humans should not exist since it isn't related to the purpose of the simulation."

• Observed reality = "There is lots of evidence many planets without humans exist."

Reality does not match the prediction, therefore this thought experiment test failed.

When we look at the observed reality, it becomes clear that it takes quite a large amount of delusionally narcissistic hubris to conclude that the entire universe revolves around humans. Conversely, a naturalistic view that things in the universe just naturally evolved to be what they are because that is the way nature works makes a lot more sense.

Naturalism Hypothesis Thought Experiment Test 1

If "planets are a side effect of the big bang, things happen randomly, including life" [x]

then (->), "life would be scattered randomly throughout the universe at the rate of its likelihood" [y];

• Predicted reality = "life would be scattered randomly throughout the universe at the rate of its likelihood"

• Observed reality = "life is an unlikely phenomenon and it exists at a low rate in the universe"

Reality DOES match the prediction, therefore this thought experiment test PASSED.

Conclusion

Is the "simulation hypothesis" plausible? My judgement -