Why We Might be in a Simulation, and Why this Isn’t Completely & Totally Useless to Think About

[NOTE: this entry has been expanded and updated. See July 2018. This 2015 version has a lovely comment from an internet stranger, so I didn’t want to delete it, but I recommend readers skip this one in favour of the updated version]

“The fact that we live at the bottom of a deep gravity well, on the surface of a gas covered planet going around a nuclear fireball 90 million miles away and think this to be normal is obviously some indication of how skewed our perspective tends to be.” – Douglas Adams

The Simulation Argument

In 2003 the philosopher Nick Bostrom published a paper called the Simulation Argument, in which he puts forth some sound logic that reaches a pretty wacky conclusion. But as I thought more about it, his conclusion seemed progressively less wacky…

Here’s my spin on his thesis (I present it a bit differently – I don’t believe that Bostrom has actually stated that he believes a simulated universe is more likely than not):

• If you believe that humans will reach a stage of technological development where we can create artificial intelligences in complex simulated worlds, then there will be more of these simulated universes than the original universe that gave rise to them. So, once you believe that such artificial universes are possible, simple probability suggests that our own universe is one such simulated universe.

Why even think about it?

The most common response from people who accept Bostrom’s logic as sound is: “Why would that ever matter? We live in our universe and it is the way it is, and we can’t experience what our brains won’t perceive so why think about it?”

Accepting that we may be in a simulation is a firm acknowledgement of our insignificance and ignorance. Such traits are worth encouraging in a world where we have become obsessed with harmful posturing and blind adherence to certainty. Embracing fundamental uncertainty and our overall lack of importance can help emphasize that bare experience (of any conscious being) is the only thing that matters.

Real World Support of the Simulation Argument

I’ve spent some time thinking about how applying information technology to our world makes sense, and vice versa. This is by no means a rigorous logical or empirical argument, but is rather the result of many thousands of kilometres spent alone on a road trip in the western United States. All I hope to do is convince the reader that the Simulation Argument could be true, and it would be an added bonus if I can help lessen the dogma of certainty prevalent throughout the human species (if only amongst my 3 readers).

Why the requisite level of technological development is theoretically possible for artificial intelligences in artificial worlds:

• The brain is an indication that intelligence and emotion (ie. everything required for your sci-fi brand of artificial intelligence) can arise from complex assemblies of matter. It is even possible for complex matter to experience realities without external sensory input, which we do every night when we dream. Although the idea of a full-fledged universe might seem improbable when considering our current technological abilities, remember that we are not even a century into the modern computing era.
• Both the brain and IT use electrical signals and binary to transmit information (IT’s binary = 0’s and 1’s; brain’s binary = neurons firing or resting). The brain is nature’s information technology.

Why felt experience does not rule out a universe made of information:

• All of our experience is first filtered through our brains. Thus, we cannot rely on the seeming solidity of reality to argue that it is some kind of physical “stuff” with inherent properties. Colours, sounds, smells, and even hot & cold don’t exist without brains to project them. The brain has the qualities of an information-processing device, so who’s to say that our experience isn’t an emergent property of pure information processing?

The mystery of logic’s ability to describe reality:

• Logic (especially mathematics – which arguably is actually a form of logic and not science) has an uncanny ability to describe reality. No one has ever offered a satisfying explanation for the precision with which maths can predict and model our universe. If our universe were an emergent property of some kind of supercomputer, that supercomputer would have been created through applying logic to the world. And if we were built upon a logical (and empirical) construction, it makes perfect sense that logic applies to the world.
• Of course running this argument back would seem to imply that the “one” original universe was also a logical universe, which nullifies the point that I just made. I haven’t been able to resolve this tension, but have tried to loosen it with the following idea:
  • Perhaps instead of the ancestor-simulation that Bostrom suggests, our universe is actually one of many attempts to create superintelligence (ie. a computer that can run more complex calculations than the calculations required to produce it). To create such superintelligence may require the development of technology that can assemble information more complexly than could be done without such technology. So just as we are attempting to create superintelligence in our universe (replete with hyper-universe-al complexity and thus logical possibilities), maybe we are just the forefront of a cascade of relative “superintelligent” universes, successively more logical than the last. If this is confusing, that’s because it is (and may very well make no sense).
    • Here it is simplified: Universe¹ (low complexity, but ability to accumulate information and thus produce information technology) –> Universe²/Simulation¹ (increased complexity, increased logic) –> Universe³/Simulation² (increased complexity and logic) –> …ever increasing complexity and logical ability, perhaps with no upper limit…
    • Maybe the fortuitous laws of physics were actually designed to give rise to a universe conducive to the accumulation of information (which would require some kind of spontaneous element, because if it were all pre-planned presumably the creators would derive no new information from it)
• Mathematical sequences abound in nature: Fibonacci sequences and other fractal patterns appear to result from nature always perfectly calculating paths of least action (read Richard Feynman on why paths of least action in nature are mathematically ‘miraculous’).

E=MC²:

• The speed of light is the cosmic speed limit. Light has no mass, and is essentially as distilled as information can get. The fastest form that information can take is in the form of light. In a universe made of information, the limits on the speed of that universe would be the limits by which information can be transmitted. In our universe this information-speed-limit is the speed of light. Our cosmic speed limit may actually be the limits on the speed by which information can be transmitted because of technological restraints.
• Light accelerates instantaneously (ie. it reaches light-speed without any acceleration time). Information, not being a physical property, also requires no acceleration time.
  • Matter and energy are two aspects of the same property, but we experience matter and energy as being fundamentally different. Matter and energy might actually both be made of information, which appears in its numerous forms when masked through our experience of the world.
    • Any change in our universe is just different arrangements of what has always been here. It’s like a computer that is limited by its hardware but whose software can be continually updated.

Quantum weirdness (re. uncertainty principle):

• Bostrom suggests that the universe would only need to render itself where experience actually occurs. In this way information processing power could be conserved to some degree. This might be an explanation for strange happenings at the quantum level. Maybe peering at the smallest building blocks of our universe (ie. the smallest chunks of information) forces them to be rendered, and information that generally dictates momentum or location must be diverted towards our experience of the quantum realm, such that there is insufficient information to render momentum and location at the same time.

Phew. I have more to say about this but I’ll leave it there for now.

 

UPDATE: I said that no one has offered a satisfying explanation for maths’ ability to describe reality, but many say the answer is simple: it’s because the basis of mathematics involves counting, and the empirical roots of the logic of counting are easily identifiable. I’m not yet sure that I am fully satisfied by this answer, but acknowledge that I may have been too absolute in my statement about the mystery of mathematics to map reality.

UPDATE 2: For fun, here is Elon Musk – who has had “so many simulation discussions it’s crazy” – talk about the low probability that we are in base reality.