# Universe simulation

Is it theoretically possible to simulate the universe? I know that some philosophical theories say that our universe is a computer simulation, but is this feasible? Would it ever be possible for a computer, classical or quantum, to perfectly simulate a system inside of which said computer is?

If our simulation were an imperfect representation of the original, by induction each following simulation in subsequent universes would be less and less precise.

I just fail to see the foundation of the simulation theory, but I would like to understand the reasoning.

• Aren’t most simulations an imperfect model?
– user207455
Jun 5 '19 at 16:25
• My question is whether physics law allow for perfect simulations to actually be made? Jun 5 '19 at 16:32
• Possible duplicate of physics.stackexchange.com/q/374863 Jun 5 '19 at 17:34
• That duplicate is closed. Jun 5 '19 at 19:03
• Where would you put it? Jun 5 '19 at 19:13

## 1 Answer

The definition of simulation that is usually used includes the fact that simulation involves a certain model (usually, mathematical). The models might vary in their granularity and definition of what is considered key characteristics of the object/system/process they are intended to represent.

Then, during the simulation one is going to achieve (provided perfect implementation) a result that will be a subject to certain approximations being made (modeling error, measurement error, numerical method error, etc.) depending on the chosen model and simulation methodology.

Thus, depending on what you chose to be a "good enough" model of the universe, the answer to your question will be different.

For example, if you insist that the model of the universe must include the fact of the particular universe simulation (the one you are doing) happening in the future, you would arrive in a circular argument fallacy, and such simulations are not possible -> you would have to know the simulation result in advance.

I would also suggest a quick look at the Wikipedia article on simulation and Wikipedia article on mathematical models. It's worth to note that in order for a simulation to be perfect, the used model has to be perfect. If that seems to be your key question, you would have a more fruitful discussion on MathOverflow and Philosophy SE, while Physics might not be the best fit.

• Yeah, that was my key question. I have already tried Phylosophy SE, but they rejected me there by closing my post. Probably at this point of time, the question is not appropriate anywhere. Jun 5 '19 at 19:14
• @JonhSmith your question on Philosophy SE is not the same at all. That one was closed for a good reason. You might search there for existing questions about "model vs reality", "perfection/imperfection", and formulate a concrete question(s), but not the vague one. Jun 5 '19 at 19:23
• What about if the universe you simulate is "running" at a lower speed. You are able to simulate it perfectly from the beginning, but it goes half as fast as your actual universe. The computational resources you require to simulate your simulation within the second universe are just half (per time step) and continuing in this way, if every step in is half as fast, you do not need infinite computational resources (this of course would need you to know perfectly the initial conditions of the universe, which is really unfeasible, but well, it is a thought experiment).
– user137661
Jun 6 '19 at 16:37