Hacker News Reader

Comment by zw123456 | Hacker News Reader

And perhaps someday we'll get to meet the great programmer in the sky the developed the simulation we are living in and we will ask him why he designed it that way and he will say something like "oh I just wanted to optimize the code so I simply excluded reality subroutines when there were no beings looking. I just never thought you guys would notice. As soon as I saw that you guys noticed the flaw, I was going to load a patch but then the confusion it was causing with the simulated beings became interesting and so I just left it in as an accidental feature of the game." Meh, but probably not.

View on HN

I always find it fascinating at just how much of our fundamental physics ends up being constraints on information movement. The laws of thermodynamics are about entropy, general relativity puts constraints on the movement of information (for example, Spooky Action as a Distance(tm) is faster than light, but you can't transmit information with it), the Uncertainty Principle puts limits on how much information you can have on a given system.

Information information everywhere you look in fundamental physics. It does make me wonder why.

You can reconstruct every mathematical theory as information flow and restrictions. That's a property of mathematics (Turing found that), not physics.

I'm fascinated by such things also. I often can't help thinking of God as a programmer, though I don't know how much of that is confirmation bias of my priors as a theist and a programmer. But I've always thought the periodic table and DNA both felt a lot like code...

The masons thought about God as the great builder, I believe generally we make gods in our own image.

As do the high-level masons.

Physical constants look alot Like magic knobs defined in some header file.

I think it isn't a coincidence that there is so much order and elegance behind the building blocks of our physical world and the universe but the result is chaotic, and unpredictable.

Maybe it's just our brain discovering patterns that don't mean anything.

It's hard to imagine a universe with infinite resolution. You'd effectively have an infinite number of bits in an infinitely dense space, which would potentially require infinite energy to flip.

But I suspect (on no basis whatsoever, except the fact that all previous metaphors have been wrong) that the "universe = Turing machine model" is wrong in some fundamental ways.

I have no idea what the universe is, but I'm open to the possibility that it isn't just an information processing system.

Agreed, it just feels inadequate to me, too. Information theory is a hammer ... and now everything we can think of (~ information...) looks like a nail.

Another hint that maybe the big Turing machine isn't quite cutting it might be the experiments showing that gravity isn't quite like 'just' an entropic force but seems to behave genuinely different.

Quantum loop gravity theory suggests a resolution of 1.61619926 × 10^-35 meters, the Planck length.

Pretty much, but consider the alternative - hypothetically infinite universes, inifitely divisible particles, turtles upon turtles...

Because you need information exchange on the fundamental level to get an arrow of time on the macro level.

Sharding. =)

In other words, reality is lazily evaluated and is written in Haskell ?

Not really, most of it is just hacked together with perl: https://xkcd.com/224/

Perl 6 has lazy evaluation, doesn't it?

Lazy lists, anyway. As far as I know there's no built-in support for lazy scalar evaluation.

What does Perl 6 not have?

Users.

Hmm, I figure it's more the the optimizer depends on code not exhibiting undefined behavior, like in C. It's not really the case that it's lazily evaluated, just the the causality gets screwy when you try to observe things in an asynchronous system without any kind of coherent memory model.

The fastest anything can travel is the speed of light, which is the speed perceived when a result is already calculated and held in a CPU cache. For everything else, it must be loaded from RAM, Disk, Network and perhaps also operated upon - which can never be as fast as the speed of light.

Reality wouldn't type check.

Reality lacks a compile stage entirely - hence all the runtime errors.

Reality is interpreted.

Of which, I am one.

That would quite possibly be the worst optimization ever.

"Yeah, I could have tracked and updated n values pretty cheaply, but instead I decided to exponentiate an exponentially huge matrix and use that to update a vector containing 2^n complex numbers associated with the possible assignments of the original n bits. Also, you should use bogosort. It's the best."

Maybe the 2^n version is necessary to support consciousness?

"Lol. And you little sims still think you're conscious. So cute."

/jk

"Think"?

Yeah but maybe they know how to make P=NP

That is assuming that whatever reality our simulation was running in followed the same rules.

Of course. I used that assumption because it was also implicit in the original claim. If we don't have any idea what the top-level rules are, then claiming quantum mechanics is an optimization artifact falls a bit flat.

If the top-level rules are classical-ish, then quantum mechanics is expensive instead of cheap. If the top-level rules are quantum-ish, then that begs the question of why we thought quantum implied simulated. If the top-level rules are totally different... then there's not really much to be concluded.

Off to be the Wizard by Scott Meyer is a very funny fiction book that is based on the concept of reality being a computer program that hackers can manually manipulate to do cool stuff. It mostly occurs in King Arthur's Court, which happens to be the chosen time and place all these hackers wind up.

Reality frustrum culling?

^ This

I kind of love the idea that it's an optimization because the universe would have a shitty frame rate if light's quantum state wasn't lazily evaluated based on whether anyone's viewport was aimed at it.

Surely, "As soon as I saw that you guys noticed the flaw, it became part of the ABI and had to be maintained."

Well, if you want to prove the simulation argument, all you have to do is simulate a reality for an individual at any time-dilation you like. With the parallel compute capabilities we are increasingly growing, simulating a small section of the universe at a large time-dilation should be within our grasp sooner than we think.

There is a limit to the total computation that can be done in the universe or with a certain limited energy/volume of space. It's caused by the slowness of signals traveling at light speed. They say probably a black hole is the best computational engine there is.

https://en.wikipedia.org/wiki/Limits_to_computation

Right, but you don't need to simulate an entire universe, just what's observable.

It's all fun and games until your solution blows up.

> developed the simulation we are living

Well, the way I understand this, is that this result proves that we do NOT live in a simulation. Because if reality does not exist until measured, then we apparently cannot compute reality ahead of time. And if we cannot compute reality ahead of time, it does not exist yet.

If the purpose of the simulation is to simulate us, then it would make sense for it to not bother computing anything we don't measure.

https://en.wikipedia.org/wiki/If_a_tree_falls_in_a_forest

Quite sure there's a clever quip about this exact thing in terms of bits and bytes, I just haven't walked into that part of the forest yet.

True in principle, but you run the risk of premature optimization. Spontaneous changes of will would be impossible.

> Spontaneous changes of will would be impossible.

We don't even know what (free) will is, yet alone whether anyone has it.

The equivalent of a memory leak being spooky action at a distance?

No, I'm quite sure that would be black holes.

Edit: I think you're right. Black holes would be the equivalent of a segfault.

But probably though.

Reply on news.ycombinator.com