Assuming our simulation is not designed to auto-scale (and our Admins don’t know how to download more RAM), what kind of side effects could we see in the world if the underlying system hosting our simulation began running out of resources?
Simply put.
We wouldn’t notice anything.
Our perception of the world would be based only on the compute cycles and not on any external time-frame.
The machine could run at a Million Billion hertz or at one clock-cycle per century and your perception of time inside the machine would be the same.
Same with low ram, we would have no indication if we were constantly being paged out to a hard drive and written back to ram as required.
Greg Egan gave a great explanation of this in the opening chapter of his Novel Permutation City
Clearly wrong .
Running out of ram happen all the time. We see something, store it, and that something also gets stored in ram. But if that second storage gets reaped by the oom, the universe reprocess it.
Since it’s already in our copy, it cause weird issues. We call it Déjà Vu!
We can see that already when something approaches the speed of light: time slows down for it.
This simplification horribly misunderstands what time-dilation is, and I love it.
My vm is running out of ram.
I have a running theory that that’s also what’s going on with quantum physics, because I understand it so poorly that it just seems like nonsense to me. So in my head, I see it as us getting into some sort of source code we’re not supposed to see, and on the other side some programmers are going “fuck I don’t know, just make it be both things at once!” and making it up on the fly.
Data in memory will be offloaded to swap space. I doubt we’d notice any fluctuations since we’re part of the simulation, but externally it could slow to a crawl and basically be useless. They might shut it down, hopefully just to refactor. But again we probably wouldn’t notice any downtime, even if it’s permanent.
That would be the most pleasant way to go :)
Not sure you’ve experienced the end of many SimCity games if you think this is the case. 😂
If anything, the earth lately kinda feels like someone’s gotten bored with the game.
12 meteors, 8 volcanoes and 10 tornadoes incoming you say?
A landscape full of Arcos and waves of boom and bust?
iirc this is a plot point in the book “Fall; or, Dodge in Hell” by Neal Stephenson (sequel to Reamde). At some point the virtual world slows to a crawl so much that people outside of it cannot really track what is going on but it’s transparent to those inside the world. I might be misremembering exactly how it was implemented.
An automatic purge process will start to prevent this. It happened several times in the past. Last time between 2019-2022. It removed circa 7 million processes. With regular purges like this it is made sure that the resources are not maxed out before the admins can add more capacity.
Maybe we’re already there and death is just the garbage collector freeing up more space.
I love this concept
Could make a good book
The OOM killer goes on the prowl.
If our entire universe is a simulation so are our laws of physics, in the parent universe running our simulation the universe might be powered by pure imagination and the concept of memory or CPU cycles or even electricity might not even exist
These answers are all really fun but I didn’t see anyone point out one thing: why should we assume that our creators’ “computer” architecture is anything remotely similar to our technology? I’m thinking of something like SETI—We can’t just assume that all other life is carbon-based (though evidently it’s a pretty good criterion). The simulation could be running on some kind of dark matter machine or some other exotic material that we don’t even know about.
Personally I don’t subscribe to the simulation theory. But if it were true, why would the system have any kind of limitation? I feel like if it can simulate everything from galactic superclusters down to strings vibrating in Planck Time, there are effectively no limits.
Then again, infinity is quite a monster, so what do I know?
all other life is carbon-based (though evidently it’s a pretty good criterion)
The short version is that the only other element that allows 4 covalent bonds is silicon, but nobody has been able to find a solvent that allows complex silicon-based molecules to form without instantly dissolving any structures they form.
I remember reading about how silicon is theoretically possible, but I had (erroneously) assumed there were more potential candidates. Thanks for the additional info. This stuff is so fascinating!
That’s why history repeats itself. It’s doing that more frequently these days because there’s more people remembering more things.
One word
Alzheimers
Have you ever noticed when you look into a telescope that it takes a little bit to position yourself right to see what you’re looking at? And it seems like you used to be able to do it a lot faster? That’s not age, that’s actually lag time added to cover decompressing the data.
That would only be a problem if you need dynamically allocated memory. It could be a statically allocated simulation where every atom is accounted for.
Given the whole “information can neither be created nor destroyed” aspect of atomic physics, taken literally, this theory checks out.
Limitations of hardware resources show up as “Natural Limits”, like the speed of light, in the simulation. The amount of RAM consumed translates to the Hubble Bubble, or the greatest distance light could have traveled since the beginning of our universe, and moreso to the amount of matter and energy contained within it, which is a constant. Energy and matter cannot be created or destroyed, only changed forms allowed, so a set amount from the beginning.
They take some users offline to free up some memory for everyone else
Why would we run out of RAM? Is there new matter being created? It’s not like we’re storing anything. We will keep using the same resources.
The nature of quantum interactions being probabilistic could be some resource saving mechanism in a higher order simulation.
New human instances are being created, and as our society’s general education keeps going up, they demand more processing power.
As our tech goes up, this has to be simulated as well. Not only things like telescopes and the LHC, but your computer who’s running a game world doesn’t actually exists and it’s the super computer who’s running it.
Obviously, this is just a drop in the bucket for an entity that can make a fully simulated universe but the situation quickly becomes untenable if we start creating hyper advanced simulation as well, we are maybe only a few decades away.
Human instances still run on the same underlying physics. No further RAM is needed.
As our tech goes up, this has to be simulated as well
Everything is made up of atoms/photons/etc. If every particle is tracked for all interactions, it doesn’t matter how those particles are arranged, it’s always the same memory.
Atoms and photons wouldn’t actually exist, they would be generated whenever we measure things at that level.
Obviously, there’s many ways to interpret what kind of simulation it would be. A full simulation from the big band is fun but doesn’t make for good conversation since it would be indistinguishable from reality.
I was thinking more of a video game like simulation, where the sim doesn’t render things it doesn’t need to.
where the sim doesn’t render things it doesn’t need to.
That can’t work unless it’s a simulation made personally for you.
I don’t follow. If there are others it would render for them just as much as me. I’m saying it wouldn’t need to render at an automic level except for the few that are actively measuring at that level.
Everything interacting is “measuring” at that level. If the quantum levels weren’t being calculated correctly all the time for you, the LEDs in your smartphone would flicker. All those microscopic effects cause the macroscopic effects we observe.
If it was a simulation, there would be no need to go that far. We simulate physics without simulating the individual atoms.
None of it would be real, the microscopic effects would just be approximated unless a precise measurement tool would be used and then they would be properly simulated.
We wouldn’t know the difference.
