They argue that the universe isn’t mathematically computable, and therefore not possible to simulate. It’s not about physical computers.
We know there’s a class of ”uncomputable problems” for which there’s no algorithm (most well known is halting problem). If the universe rely on any of these uncomputable problems, then no computer - no matter how advanced it is - can simulate the universe. Something else other than pure computation is needed.
However, their argument rely on that ”quantum gravity” is what makes the universe uncomputable. I’m not sure how valid this statement is.
When someone claims something isn’t computable, it is instantantly sus, especially from math nerds and not compsci nerds. Imagine the universe is indeed uncomputable, but each measurement is. The number of measurements you’d need to sim (at various scales/resolutions) is vastly smaller than the universe as a whole. This is morally equivalent to occlusion pruning in 3D games. If you aren’t looking at it, it isn’t being rendered.
When it comes to theory of computability, you don’t need to account for optimization techniques. No need to consider the practicality of getting an answer from the algorithm, like how long it takes or how much memory it requires. Either you can get an answer in finite amount of time, or you can not.
But I agree it’s sus when it comes to making such strong statements about the compatibility of the reality. I don’t trust this paper makes all the right assumptions.
Going to circle back around on uncomputible in “our” version of reality. I mean it’s kind of lazy in its way but it seems like the possibility that the “real” universe is a fundamentally different kind of place throws out most if not all methods for “proving” it’s not. I’m not even a fan of the matrix theory but still, to acknowledge it.
This doesn’t really address the idea that our simulation is a simplified version of the “real” universe though does it?
They argue that the universe isn’t mathematically computable, and therefore not possible to simulate. It’s not about physical computers.
We know there’s a class of ”uncomputable problems” for which there’s no algorithm (most well known is halting problem). If the universe rely on any of these uncomputable problems, then no computer - no matter how advanced it is - can simulate the universe. Something else other than pure computation is needed.
However, their argument rely on that ”quantum gravity” is what makes the universe uncomputable. I’m not sure how valid this statement is.
When someone claims something isn’t computable, it is instantantly sus, especially from math nerds and not compsci nerds. Imagine the universe is indeed uncomputable, but each measurement is. The number of measurements you’d need to sim (at various scales/resolutions) is vastly smaller than the universe as a whole. This is morally equivalent to occlusion pruning in 3D games. If you aren’t looking at it, it isn’t being rendered.
When it comes to theory of computability, you don’t need to account for optimization techniques. No need to consider the practicality of getting an answer from the algorithm, like how long it takes or how much memory it requires. Either you can get an answer in finite amount of time, or you can not.
But I agree it’s sus when it comes to making such strong statements about the compatibility of the reality. I don’t trust this paper makes all the right assumptions.
Going to circle back around on uncomputible in “our” version of reality. I mean it’s kind of lazy in its way but it seems like the possibility that the “real” universe is a fundamentally different kind of place throws out most if not all methods for “proving” it’s not. I’m not even a fan of the matrix theory but still, to acknowledge it.