Nick Bostrom calculated the odds of our universe being a simulation, you can read about it here. Bostrom makes claims about AI that follow the common ignorance about what it takes for something to be intelligent, even what that actually means. When it comes to whether our universe is a simulation he does a similar thing, going into some debt, but keeping to a level that does not reveal enough to generate a true answer.
Bostrom makes a common mistake made earlier by the so called logic positivists, who thought the world and all thought about it could be described by logic. It can’t be, because logic is a special case of human intelligence, which very clearly excludes the recognition processes, so you would have to state in your logic inference somewhere that “tomato(green)”, meaning “the tomato is green”, but how do you determine this assertion ? The proponents of logic positivism simply did not get to that question in the exploration of their approach. Today one can fix that particular problem (say while building an artificially intelligent robot) by having hardware that produces a description in terms of logic expressions as a result of sensing (camera) and autoclassification of the camera output. Even then logic is intrinsically a model, so can never ‘simulate’ reality.
Fundamental particles, ones we think we should observe, but sometimes can’t
When asking what the chances are we live in a simulation Bostrom goes to his usual exponential realm of super computational powers. This is kind of a intellectual trap, we see simulations that clearly amaze us and we extrapolate our idea about those simulations into the future. Surely this means we can simulate everything? If we can drive a virtual car off a cliff in 3d today, who knows what we can do in a thousand years?
We are introducing a strong quality filter here though : ourselves. We live in a reality that is defined by our senses. We evolved to survive in reality, which causes us to focus on certain aspects of it and have certain plans with it. A good way to understand our brain is as a simulation system evolved to internalize our environment to such an extend that we can model and predict the outcome of our actions, so that we can identify risks before we experience them. This in my opinion is the origin of time to begin with, we create it so there is a now to make decisions about a future, but that is another story (can’t find the post I wrote about it).
Bostrom forgets or ignores a thing that is enormous, which is the apparent computational power of our universe (if we assume it is non deterministic). The universe is a fascinating thing, because it demonstrates a property we call the conservation of energy. If the universe would not conserve energy (meaning stuff keeps changing into other stuff with the same overal energetic value) it would implode or explode immediately. Every nanosecond the finest particles that make up our universe instantly find a new configuration to make that rule true. This happens even though we know there could be many alternatives according to quantum physics. That instant matching of the energy conservation rule with every change is a feat of problem solving that the quantum computer builders try to capture and use.
Quantum mechanics is not exotic and out there, f.i. a single quantum of light, one photon, can cause a response in our eyes, even though we can not see it. To see light we need about 9 photons. The number of them hitting 1 cm2 in daylight is about 100.000.000.000.000.000
What we seldom realize is that in order for the universe to find the new ‘now’, for any change to occur, many options have had to be considered. This is if we assume that there is not only one outcome. There is an argument for that, because quantum physics is probabilistic because we don’t know what the particular outcome in a particular instance is. That is just like saying a role of a single dice has an outcome of 1 to 6 each with 1/6 probability. However the sequence of outcomes could be fixed, we would not know it because our process of roling dice does not allow us to observe it, because we too are dice being rolled in a way (or not : we are coevolving).
There are subquantum energies that we can’t observe
But if our reality is a simulation it has to be modelled. That is hard, because we know there are influences we can’t observe. We know they exist from the so called Cazimir effect, because we observe that non-detectable particles outside to thin plates close together will push the plates together because the space in between is too small for some particles to emerge there to push them back. Emerge? From where? Out of the vacuum of space is the answer. We don’t really know for sure but we think there are other dimensions from which particles emanate into our observable dimensions which then quickly anihilate. Can we measure these particles? No, because they anihilate. Why? To make the energy conservation laws true. The only device that seems to be able to capture them is the so called EM drive.
So we have the world we observe, and we know there is a quantum dynamica structure to it etc. and we are puzzled by the ability of reality to keep evolving, but there is also an unobservable substratum of activity we will never be able to measure or model deterministically.
Nick Bostrom in the mean time hypothesizes that there will be future intelligent beings that will have access to more computational power so much that they can simulate us and our universe, and that the question whether this is the case can be answered by investigating the combined probabilities. This is a top down type of reasoning that assumes intelligence and then projects it as a driving force that may be responsible for our universe.
The description of our universe created so we can mentally simulate it introduces options and properties which may not really be there
We can turn this around. How would a computational system ‘simulate’ the sub quantum particle existences that do have effect on our quantum reality (as the moving Cazimir plates prove) because the nature, direction, type of these apparitions follow unknown rules (except that the wavelength has to fit), if the particle can exist in a vacuum, then it may, for a brief moment. Its like the rule of energy conservation quickly deletes it.
A deterministic computational system would have to take the multidimensional reality and inject these particles based on the quantum context for every location in our gravitational matrix (more on my understanding of it in another post) and thus allow new moments to emerge (at least in our simulation that tries to escape its entropy laws). All this while Bostroms simulation computer would be a clocked device, with discrete timesteps and rules. If we let go of our preconceptions on the computational device of our future intelligent beings, and we would assume they mastered quantum computers (symbolizing a mechanism by which one can choose the instant solution to the conservation of energy challenge when its context is created) then we run in some trouble. Because we then claim the simulation is a mechanistic combination of devices that do what? Be the univers in its raw state. Right.
So Either we run into the limits of a time deterministic law driven simulation as it can not meet the demands of creating the reality we observe or we harness this same reality in order to ‘simulate’ it. We would need the universe to simulate the universe. It would be impossible to observe this universe from outside because this would violate the principle that holds it together. Ergo, we can’t be living in a simulation.
That said our reality is weird, but we should recognize our desire to find a structure and the cause of things or its origin is one that is born from our design. This is why Bostrom is so succesfull, he creates a fantasy threat ( super human AI for insance which can be real for sure) that answers our potential to fear. In that sense the theories of spooky action at a distance when considering quantum entanglement is exicting because it promises us the ability to cover great distances possibly faster than light. That’s a like a race car, fighter jet or thoroughbred stallion to our minds : Cool. But there’s no proof for it.
From the height our biased intelligence we may think we may exist in a simulation, but from the fundamental processes we observe and those we know we can never observe it follows we can’t be. If we do take the position that our universe could be computatioally simulated and observed then we let go of our tools of argument and the question can be answered clearly with : “who knows”.