B A programmer's point of view of 2 quantum physics principles

[anothergol]

First, I must say I'm in no way a physicist, I barely understand the basics of quantum physics. And what I will talk about is mostly philosophical anyway.
I don't believe myself that "we are in a simulation", it's just food for thoughts. As the title says, it's about 2 quantum things that struck me, as a programmer.
If I was asked to "program" a simulated universe, from basic rules (so I'm not talking about simulating a world for existing human minds, ala Matrix, which is not very belieavable. I'm talking about simulating a full universe out of a few basic rules, a game of life, that is), there are things I would do.First, it's easy to assume that I'd be limited by memory. Doesn't matter how much memory I have access to, it would never be enough. We today have access to millions times more memory than a few decades ago, and yet it's never enough, because we always want to get the most out of what we have.
So what am I talking of? Entanglement. That, to me, looks like a pretty effective way to do data compression.
Instead of storing a set of (how many?) properties per particle, I would make them share/point to a set of properties. Isn't that what entanglement more or less is?
If my universe is gigantic, I don't think that a few, or even billions of particles far away from each other, being linked in some way, would have a big effect on the outcome of my simulation. I would get very similar results out of much much less memory. If I'm interested in results like life (which frankly would be pretty hard to spot in the result of such gigantic simulations, but that's another story), I don't think that entanglement would affect this too much.Second, tunneling. It's funny because it reminds me of someone's first videogame. When you make a videogame in which you shoot bullets, yes you may handle bullet collision by checking if their path crossed a target within a time/space frame, but generally, because it's much simpler (& thus costs less), you simply move your bullets in.. quantized steps.. and you check for collisions (perhaps interactions would be a better wording).
And when you're a beginner, and you picked too large steps for your bullets before checking for collisions, what do you get? Bullets that go through walls or enemies.
So let's assume my simulation would check for interactions in quantized steps (for computational economy), the result would be possible tunnelling. Annoying, but again, not very visible at a bigger scale, might not affect the outcome of my simulation that much. And interestingly, there would be a way to minimize this tunnelling, by adding random jitter to "particle positions". That way, yeah from time to time a particle would tunnel through a wall, but another particle flowing the same path, because of the added random jitter in its position, would not necessarily tunnel. I would then reduce this unwanted tunnelling quite a lot through this added jitter, which would ressemble quantum uncertainity, no? (this said, I keep finding 2 conflicting explanations of the uncertainity principle, so I'm not sure I understand it).
Noise is often a good way to mask imperfections.Anyway, that was just food for thoughts & just that.
I have no correlation to make between programming & the interaction of a particle with itself. And I would personally be a believer of the many-worlds interpretation (because it's the simplest), which in no way ressembles a computer simulation.
I don't think this is the wrong forum to post this, but sorry if I'm polluting.

 

[Gigaz]

Proper numerical quantum physics requires quite a lot of memory. ;)
For a real ab-initio simulation of an N electron system you need to solve a 3N-dimensional partial differential equation. For two electrons and a space discretization into 100 points per dimension that's 10^12 quantum states. Now you need to diagonalize a 10^12 times 10^12 matrix (Ok granted, it's a sparse matrix, but still not funny. Matrix diagonalization usually scales like N^3)
In laymans terms: Classical simulations require us to calculate one path of the system from start to end. Quantum simulations require us to calculate and store all possible paths.

 

[anothergol]

Indeed (I don't fully understand but I get the idea), as I wrote myself, even the interaction of a particle with itself wouldn't make much sense in a simulation, computationally speaking.

But don't you agree with the "pointer to shared properties" analogy?
I generally hear about entangled particles "communicating with each other" ("at spooky distances"), but can't we imagine them not communicating directly with each other, but with that shared structure of properties, "somewhere in its own dimension" (unlinked from space)? Like the particles don't own their properties, they get them from.. somewhere else? I don't know if that makes sense in "reality" but well, entanglement doesn't seem to make much sense.

I assume we'll never know, but are there theories for the ways & reasons for entanglement like there are interpretations for quantum mechanics more generally?

 

[fresh_42]

I'm afraid we don't support blog like contributions, neither do we discuss philosophical topics, as they rarely deserve to be called as such. We have plenty of threads about the nature of quantum mechanics, often heavily disputed, which can be found via our search mechanisms, so I recommend to have a look at them.

This thread is closed.