I've been writing a Perl
module to play the Game of Life
across a network of peer
s, and this gave me some reasons to believe that we do, indeed, live in a Matrix
Suppose that a technologically advanced civilisation wished to create a virtual world, existing as a software simulation. They could build a single huge supercomputer to do it. But I think it is a defensible assumption that they would be more likely, for reasons of cost, to use a network of mass-produced computers, linked together.
Of course, to create a single world within a disparate set of computers, the computers need to talk to each other. For example, in the Game of Life world, each square needs to know how many of its neighbours are alive or dead. A single computer can handle many squares, but it will regularly need to check with other computers for the life status of the borders to its domain.
This might create a problem. If all the computers in our network had to communicate with all the other computers, there would be a Gnutella-like chaos of messages being passed around which would grow exponentially (not sure if that's exactly correct, but you get the idea) with each new node added to the network.
The Game of Life doesn't have this problem, because each square is only affected by its neighbours. By extension, each computer only needs to talk to computers which have adjoining nodes. Instead of increasing exponentially, the number of messages being passed increases arithmetically with each new computer. For example, if each computer handles a board of 100 x 100 GoL squares, new computer N in the diagram below only needs to talk to computers B,D,F and G. A,C and E can remain blissfully unaware.
- - -
- - -
To put it another way, the Game of Life forbids action at a distance. Sounds familiar? Well, our physics too forbids action at a distance. It's a core rule that two bodies can only interact if they are contiguous, or by the intervention of a third body which passes between them.
However, this itself wouldn't solve the exponential message growth problem for our world. Why not? Well, the Game of Life proceeds at a constant speed. Each cell affects its own neighbours, once every turn. Our world, however, has variable speeds. So even if a computer simulation of our world forbade action at a distance, there would still be the danger that a very fast-moving object might arrive from the space simulated by a non-neighbour computer.
What our future civilisation would need is some kind of maximum speed limit on objects in the simulation. This speed limit would guarantee that no object could move through the entire region of space governed by a single computer, before that computer could inform its neighbours of the object's existence. The speed limit would ensure that each computer could talk only to its neighbours, but still remain certain that it knew about everything that could affect space in its region.
Of course, our universe famously does have such a speed limit.
So putting all the evidence together, our universe looks suspiciously suitable for running inside of a massive network of computers.
- It's divisible into regions of space - so each computer can run the simulation for a particular region
- The laws of physics are invariant between these regions - so the software running on the computers can always be the same
- There's no action at a distance - so computers only need to talk to their neighbours
- There's a maximum speed limit for objects in the universe - so computers needn't worry about objects zipping between neighbours faster than network communication can go on.
Well, you have to wonder, don't you?
IANAP. I Am Not A Physicist. Maybe someone can tell me that "no action at a distance" is all bunk according to the latest quantum experiments.
I can't prove it would be easier/cheaper to use a network of computers than a single supercomputer, to simulate us. It just seems likely.
I don't know if the problems of network communication are artefacts of our physical universe. If the advanced civilisation's universe had no speed-of-light restriction, would infinite amounts of communication between computers be possible?