# A concept which "shows" why are we limited by speed of light?

These are my thoughts about the above:
So, lets say at a given time t, I am in a room, and it has a state, which describes all the info about that state. Assume for simplicity that the room and its contents are the complete universe (the following reasoning can be extended for the actual "world" too).

So lets say that we are in a simulation, and the whole state information is stored somewhere in a "memory" (of the "computer" the simulation is running on). Now to change any part of info, the time taken should be lower bounded by a function depending upon the speed of "electrons" in "parent" world.
In physics, we often assume time as the 4th dimension. There is also a "principle" that says that no "info" can be sent faster than the speed of light. So lets say there is a description of the 3D world in an array. Now, if the distance between any two points is more than ck meters, where c is speed of light and k is an integer, then, before t=t0+k, at the first point, there can be no change at info about the second point.
Coming back to the memory description, lets say there is some smallest "scale" or unit of time called d. So there is a state at t=t0. And the "next" one at t=t0+d.Now in that state, lets say there needs to be some change which corresponds to a change at point p0 in the simulation space. So the info about that change cannot propagate faster than some function of the speed of those parent world "electrons". (Propagate here means change in values corresponding to the info of state)
Note that this means that the structure of our space (simulation space) and the way the info is stored in the simulation computer memory has to be similar

So I jut wanted to know how does this sound, and what more can we think in this direction. (Has anybody ever thought about anything in this direction before?) Can this explain some more things?

Note: I know there is a theory called simulation theory, but I don't know if people have dug any deeper into it.

• This might get some useful feedback on physics SE. Sep 27 '20 at 14:20
• @DonBranson - I think it's too far removed from any real physics theories for physics SE, they aren't likely to accept questions premised on the possibility that our world is a simulation unless it's in the context of a well-defined mathematical hypothesis about what that would imply about some specific experimental observations. Sep 27 '20 at 16:47

So the info about that change cannot propagate faster than some function of the speed of those parent world "electrons".

You're making the implicit assumption that we need to update the simulation in real-time.

Suppose we have a program that simulates a universe. The universe is pretty big, so we'll use a distributed supercomputer to do it. Each compute node has a set of particles that it's responsible for, and it communicates with the other nodes whenever it requires information about the particles they manage. Supercomputers use very high throughput cables to connect their nodes, so this communication is fast. Once all the communication is done, and each node has computed the states of its particles for the next iteration, they signal that they're ready for the next update, and the simulation advances by one Planck-time interval.

But let's say we want even more nodes. We link together several supercomputers, and have them communicate over TCP. The communication phase will take a lot longer, since the nodes are kilometers apart and need to talk over the internet instead of just the local cluster, but we add enough nodes that it still completes the computations in about the same amount of time. And even if it slows down a little, it's fine, since the simulation doesn't advance until all the nodes report that they're done with the current step.

Of course, the internet protocol doesn't specify anything about the hardware used to transmit the packets. Let's say we don't want to use the existing internet infrastructure for some reason, so we replace all the cables with carrier pigeons (per the specification in RFC1149). The code doesn't change at all; the differences are all further up the network stack. Throughput can be kept at a fairly high level, depending on how much weight the birds can lift, but latency will go way up. We'll probably need quite a bit more time to compute each tick of the simulation. At least a few hours, but probably no more than a month. It'll depend a lot on how many birds we have. But as with the first change, the simulation itself doesn't care (or even notice), because it remains frozen in time until the nodes all hit the synchronization point and perform the update for the next Planck time interval. The simulated universe remains exactly the same as it was in the first version.

From the perspective of those in the simulation, it makes no difference if it takes a second or a year to compute each update. To them, it's a Planck-interval; no more, no less. The physical limitations of the real universe need not limit the physical laws of the simulated universe in any way.

Unless the latency is so high that funding runs out before the simulation is complete. Then things would be identical right up to the point where the universe suddenly stops existi

For any given physical phenomenon, a great many theories which are able explain it may get dreamed up. The acid test must always be, do they make testable predictions about the real world?

Simulation theories suffer from two problems; arguments for them invariably fall victim to deep misunderstandings about infinity, and they are unfalsifiable and therefore at best pseudoscience.

Unless and until these barriers are overcome, the question remains speculative metaphysics bordering on fantasy.

I sort of see where you are going with this but I don't think that it really works.

Re a concept that 'shows' objects cannot move faster than light - clearly there are already arguments that show why things cannot move faster than light, which work very well, and are not in need of support from additional arguments based on a simulation hypothesis (which are inevitably far more tenuous). So I presume that the point of the argument here is to add support to the simulation hypothesis by showing that one can derive aspects of real world spacetime geometry from it, which brings me to the next point...

One of the problems I have with a lot of arguments made around the simulation hypothesis is that it seems to me that, assuming this hypothesis, we know more or less nothing about what you refer to as the 'parent world' in which the simulation is running. You sort of acknowledge this by using quotes on 'electrons' etc., but I don't think you are fully facing up to this. We have no reason whatsoever to assume that there is any structure analogous to electrons in the parent world, or that any other features of our physics apply, or indeed anything other than, e.g., basic logical and mathematical principles the falsity of which we struggle to conceive. Considerations about constraints on information propagation are certainly not in this category.