5 ms·

> If the universe can be evaluated in a simulation, then the computer carrying out the simulation is not actually necessary. In fact, even defining the rules of

> If the universe can be evaluated in a simulation, then the computer carrying out the simulation is not actually necessary. In fact, even defining the rules of the universe is not necessary, because the contents of the universe with any given set of rules are not changed by the act of writing the rules down. Mathematical truths are true, even if they're never discovered. If a universe can be simulated, then it has a mathematical description, which means all of its configurations, and all of the events within it, are determined from the mathematical description. I don't think that's true. There is much more substance to the universe than just its mathematical description, and you need the computer to run the simulation to provide all the flesh and character the math does not.

jstanley 3y ago
What does the computer do that you can't describe mathematically? The flesh and character are just emergent properties of enormously complicated mathematical structures.
- PrimeMcFly 3y ago
  > The flesh and character are just emergent properties of enormously complicated mathematical structures. They are not just emergent properties, they are one of many possible combinations of possible emergent properties, just lie you yourself are. The math is always abstracted and doesn't contain the full detail, the meat and flesh. I think you are dismissive of that, but you shouldn't be. It's very significant. Crucial, even.
  - jstanley 3y ago
    How are you implementing a simulation using constructions that can not be described mathematically? If it can't be described mathematically, it can't be simulated mathematically. And if it can be simulated by a computer (i.e. mathematically), then it can be described mathematically, and the computer isn't adding anything to the experience of any observers inside the universe.
    - PrimeMcFly 3y ago
      > How are you implementing a simulation using constructions that can not be described mathematically? They could be instructions not constructions and rely on any number of data not in the descriptions themselves. > If it can't be described mathematically, it can't be simulated mathematically. You're missing the point. You can say everything running on your PC right now can be described mathematically, and that's fine. But it's also highly dependent on how you have used it, configured it, what you have installed at. The initial mathematical description wouldn't account for the things that are a result of environment and your input.
      - jstanley 3y ago
        I mean, I think you're just wrong here. The entire state of the computer can be encoded and manipulated using mathematics. If that weren't possible, the computer wouldn't work.
        PrimeMcFly 3y ago
        I'm not wrong, you are misunderstanding my point. > The entire state of the computer can be encoded and manipulated using mathematics. If that weren't possible, the computer wouldn't work. I don't disagree, but 'can' is not the same as 'is'. Absolutely everything can be modeled and described mathematically with enough variables, but that isn't always feasible, and describing all modeling all instances of something with practically infinite permutations probably isn't. Besides, if they already had that information they wouldn't really need to simulate anything. The point is that the simulation gives the parameters for the simulation but then lets it go its own way and doesn't already have all that information of the path it ended up taking. The path it ends up taking is the result of numerous complex variables, (i.e. all the choices we make in the simulation) which were not and could not be modeled ahead of time (unless you want to argue that infinity would not be so in some kind of higher dimension). Thus, actually running the simulation, to generate that flesh and meat, is the point of running the simulation, and why I think your original point that anything that can be simulated need not be simulated is wrong.

jstanley 3y ago

What does the computer do that you can't describe mathematically? The flesh and character are just emergent properties of enormously complicated mathematical structures.

PrimeMcFly 3y ago
> The flesh and character are just emergent properties of enormously complicated mathematical structures. They are not just emergent properties, they are one of many possible combinations of possible emergent properties, just lie you yourself are. The math is always abstracted and doesn't contain the full detail, the meat and flesh. I think you are dismissive of that, but you shouldn't be. It's very significant. Crucial, even.
- jstanley 3y ago
  How are you implementing a simulation using constructions that can not be described mathematically? If it can't be described mathematically, it can't be simulated mathematically. And if it can be simulated by a computer (i.e. mathematically), then it can be described mathematically, and the computer isn't adding anything to the experience of any observers inside the universe.
  - PrimeMcFly 3y ago
    > How are you implementing a simulation using constructions that can not be described mathematically? They could be instructions not constructions and rely on any number of data not in the descriptions themselves. > If it can't be described mathematically, it can't be simulated mathematically. You're missing the point. You can say everything running on your PC right now can be described mathematically, and that's fine. But it's also highly dependent on how you have used it, configured it, what you have installed at. The initial mathematical description wouldn't account for the things that are a result of environment and your input.
    - jstanley 3y ago
      I mean, I think you're just wrong here. The entire state of the computer can be encoded and manipulated using mathematics. If that weren't possible, the computer wouldn't work.
      - PrimeMcFly 3y ago
        I'm not wrong, you are misunderstanding my point. > The entire state of the computer can be encoded and manipulated using mathematics. If that weren't possible, the computer wouldn't work. I don't disagree, but 'can' is not the same as 'is'. Absolutely everything can be modeled and described mathematically with enough variables, but that isn't always feasible, and describing all modeling all instances of something with practically infinite permutations probably isn't. Besides, if they already had that information they wouldn't really need to simulate anything. The point is that the simulation gives the parameters for the simulation but then lets it go its own way and doesn't already have all that information of the path it ended up taking. The path it ends up taking is the result of numerous complex variables, (i.e. all the choices we make in the simulation) which were not and could not be modeled ahead of time (unless you want to argue that infinity would not be so in some kind of higher dimension). Thus, actually running the simulation, to generate that flesh and meat, is the point of running the simulation, and why I think your original point that anything that can be simulated need not be simulated is wrong.