I used to believe that the world is lazily loaded. That means the first moment I see something (e.g., a world map) is the point where all the past history to consistently explain the observation gets "locked in".
Another angle on it would be showing that the universe has multiple possible levels of detail - that objects behave like newtonian point masses most of the time, that digestion is replaced by a simple hunger meter when nobody's paying attention.
From this direction, time's main function is compressing "now" so there's more space for something else later.
I'd be willing to believe that everyone has a god-given amount of universe rendering time that can be allocated in different ways, and that other people's attention stacks with it in a complex, layered way.
It's the same as "If a tree falls in the forest and nobody is around to hear it, does it make a sound?"
But then again, that information has to be stored somewhere right?
Like, if a tree falls in the woods and nobody is around, sure you can say it isn't being perceived so it won't be "rendered" but the moment someone walks in the forest and sees that fallen tree, it would need to be rendered from some information about the fall and information of how it looked before the fall. That information would have to be sitting somewhere in limbo waiting to be rendered. So why not just actually let it sit IN the rendered "realm" and now you have one place for data to live in. You save on data transfer fees too.
As a collection of probabilities stored as a wave.
If a bank account has a 50% probability earning a penny every second, the bank doesn't need to store all of the intermediate values for every "clock tick", but instead calculate how much has accumulated since the last time someone checked the account and store that value with a timestamp.
Not existing is the cheapest.
Then, broad strokes, location and orientation
Then pieces, level of decay
etc,
and moving between the layers is you spending your attention budget to force the universe to do more work.
Coupling this thought with the simulation theory, it makes me wonder how the simulation would respond to increasing entropy over a larger volume than just our local system. That is, if we send a bunch of biological/artificial agents in all directions throughout the cosmos and let them wreak havoc, would it crash the simulation? Or what if we just smash a bunch of asteroids out of their stable orbits and let their disruption cascade throughout the system? Or maybe we've already messed everything up by broadcasting highly entropic radio waves throughout a spherical volume with a radius of ~100 light years?
But there's a difference between "screaming into the void", like broadcasting radio waves, that presumably have 0 effects, and other things.
I have seen creationists in web forums trot this exact argument out (only the creator was God of course).
https://www.scientificamerican.com/article/the-universe-is-n...