Hacker News Reader

Comment by eloff | Hacker News Reader

Simulation theory speculation:

If the universe is a simulation, when there gets too much matter in an area to simulate all the interactions, a black hole is the way programmers fixed that - it gets turned into a singularity and isn't simulated anymore except as one point source of gravity (or in a more complex way if black holes preserve entropy - jury is still out on that one.)

Silly and untestable, but fun to think about, like the rest of simulation theory. It would be required to do something like that if you're running a simulation with finite resources. You couldn't just keep piling matter into a finite area without bound, eventually you'd overwhelm what you can compute. Same reason one might want a hard speed limit like the speed of light.

Interestingly this is a not a discrete function, time slows near large masses, which would allow the computer to keep up as matter in an area increased, much as you'd expect in a simulation.

View on HN

Agree! This is my favorite crazy pet theory: spacetime is simulated on a flat 3d substrate with constant compute and memory per volume, which runs at a fixed rate in time (c) [1], and to account for a high density of mass requiring more compute the simulation is slowed down proportionally. Then, black holes are just the extreme end of this slowing down of the simulation.

[1]: You always move at a constant speed: the speed of light. But that speed can be broken down into two component directions which you can trade off: time and space. You typically move near the speed of light in the time direction, meaning your speed through the space direction is comparatively low.

It's because the stronger the forces and the higher the velocity in a cell of the simulation, the more iterations the numerical integrator needs to perform to maintain acceptable accuracy. It's obvious, really. ;)

Also the universe isn't actually expanding and the red shift we see in distant stars is just a numerical error due to the enormous number of steps between here and there. :)

My crazy pet theory: computational bandwidth is constant for any spacetime coordinate. According to Bekenstein bound the lower the temperature of the system the cheaper it is to flip a bit of information. As time progresses the CMBR gets colder, so it's cheaper to flip a bit, but conversely, as time progresses our Hubble volume gets smaller, limiting our total available energy budget to perform a bit flip.

None of these theories are crazy at all. Special relativity is directly derivable from energy causing a fixed amount of change (action) per unit time.

When the information/computational changes that the energy is allocated, is used for non-local movement, then we consider this to be an aspect of velocity, and the amount of changes leftover for internal changes, which effectively are internal time, is decreased by a ratio equivalent to the standard time dilation formulas.

Movement is change...change requires energy allocation, movement diminishes other interactions, hence time dilation, hence special relativity.

Wolfram talks about this in his new graph based model of physics. Mass is an emergent metric of the amount of internal changes occurring in an object per unit time. Essentially, causal loops in the graph. Feynman had a toy model called the checkerboard model where the amount of bouncing in an area represented mass.

https://youtu.be/xQafZ3CvBUs

Movement does not require energy, acceleration does.

You're wrong. What you probably meant is that movement does not require external/additional energy. It uses already allocated energy.

Once you grasp this, you will understand why action is measured in units of Energy x Time.

https://en.wikipedia.org/wiki/Action_(physics)

Let me clarify my feelings here.

>>You're wrong. This is no problem. I'm happy to be wrong, I'm happy to be told I'm wrong.

>>What you probably meant is [...] This is problematic to me. I'm a grown adult, I own what I said, and I said precisely what I meant. When it's wrong, it's wrong, and no, I don't want to retcon that in some feeble attempt to save face, and I certainly don't want others to do that for me.

Obviously, this is not my field. Is movement and motion here the same thing? Motion is relative, so this energy "use" here is relative - in some reference frames where an object would be observed as at rest, is it still "using" energy? You see these quotes? That's because I don't really understand what "use" means here. An object in motion is in motion as a consequence of acceleration, which requires energy, and certainly possesses unrealized kinetic energy, but it isn't consuming energy. So what does "use" mean here?

> in some reference frames where an object would be observed as at rest, is it still "using" energy

There are numerous ways to measure relativistic mass (total energy) of a system from outside of it. As long as there is a measured energy, it is still using energy. Observation tricks usually just juggle energy from potential to kinetic, they don't change the total energy, so they don't change the usage.

> I don't really understand what "use" means here.

Energy causes change. The total amount of energy of a system measures the total number of changes happening every second, essentially actions per second. Action = Energy x Time = (Action / Seconds) * Seconds = Action

Energy is really only transformed, it is never created nor destroyed, you probably already know this basic tenet that Einstein espoused. The energy of mass is commonly thought to be "at rest" but in actuality, mass is just the phenomena of localizing the changes that the energy must cause. When this localization is upset, you get a nuclear explosion, a lot of change that once was concealed.

We might want to know how many changes per second a single Joule of energy causes. We can calculate this by taking the inverse of the Planck constant. Since Planck constant, h = Joule / Hz [1], the inverse would give us Hz / Joule. Which yields a value of about 10^34 Hz per Joule of energy.

I wish I knew all of this earlier in my life, as a lot of my _energy_ was spent on trying to do Cellular Automata simulations of our universe. Knowing how fast our universe computes, makes any simulation attempts with our current computers seem quite foolish if magnitude matters at all. Some might still argue the rules are more important, but sometimes, as you see in strength sports, even a 50% difference in strength beats any technique advantage a fighter can muster.

[1] https://en.wikipedia.org/wiki/Planck_constant#Value

You do have a way with words. Congratulations.

An object possessing kinetic energy continues to move unless acted on by an external force/energy. It's literally Newton's first law. It doesn't move for free though. Energy bound to the object is being allocated every Planck second that it moves. That energy could have instead been used for internal interactions, local time for that object, if it were at rest with the same relativistic mass.

Time dilation is a consequence of energy having to be divvied up between external and internal changes.

Nothing I am saying is controversial. But yes, I am making it much easier for everyone to see that movement is just change of state/information, and that all energy causes a constant change of information per unit time.

The universe is computational, not really a surprise...

Although I'm skeptical of the correctness of Wolfram's grand new theory/hypothesis of everything, there are many intuitively appealing aspects of it, and I wouldn't be surprised if some aspects of the theory may help find or understand future discoveries. (Though maybe not; I'm definitely not an expert.)

Here's a detailed explanation of the hypothesis in a Lex Fridman podcast episode: https://www.youtube.com/watch?v=-t1_ffaFXao

Thanks, I'm also a big fan of Fridman. He's Rogan 2.0.

Why would gravity follow an inverse square law in this model? Wouldn't you expect there to be a hard cutoff as you moved away from a busy region, instead of a gradual decrease?

The steps of granularity would likely be the Planck length.

His proposal does have some unresolved issues. E.g. distance is quantized but the interactions of those quantized variables produce continuous things that affect other things that are quantized. We don't have enough energy to see the limit yet, but it is possible.

It's simply some kind of easing function, and if the universe does have a designer, it's obvious to see that said designer prefers smooth transition and curves for most macroscopic phenomenon.

Also by defining the inverse square law you can maintain a cohesive interaction model between different objects and regions. Their sphere of influence (which can be described by gravity waves) will now just be superimposed on top of each other without extra work.

Or even crazier, maybe it's a combination of both, maybe there is a hard cutoff after all once you pass the event horizon.

Because that's how the simulation is configured?

> meaning your speed through the space direction is comparatively low

But you do move in space at a great speed if you account for the rotation of the earth around the sun, the movement of the solar system in the galaxy, and the movement of the galaxy relative to other galaxies. It's not like there's a coordinate system with an absolute origin somewhere.

Leaving aside that this completely ignores how relativity works, even if there were a fixed reference point (e.g., the CMB rest frame), our motion through space relative to that frame is ~ 0.07% of c.

Wouldn't it have to be a 4d substrate for c to be the "time" step-rate of the simulation? What does flat mean in this context?

This is fun to think about. What is the finite resource though? Time inside the simulation can be simulated so isn't 'real' time. If simulation results were desired quickly that could be a thing, or limited storage of state.

Another good hack is inflation where parts of the universe disconnect and so can be tossed out if you only care about continuing to compute the parts you keep.

In this case I'd say time is the constraint, as you want it to run consistently through your simulation. So you can't let one part get ahead or behind another in a way that breaks causality.

You are likely not running the universe on a Von Neumann machine, so compute and storage would be combined and spread out. You can't just have opposite ends of the universe interacting instantly in that case.

It doesn't have to be that way, but you can see why it would be desirable and seems to fit observations.

However, I'm at a loss how you would explain quantum entanglement if it truly is instant over any distance. That's not really known, but it's at least a lot faster than light.

However, I'm at a loss how you would explain quantum entanglement

You seem to be applying the rules of our simulated universe to the universe outside the simulation. There's no particular reason to think the speed of light or any other rule applies outside.

This is true. The physics outside could be anything and we're unlikely to understand it.

There's a chance us humans are simulating our past for some unknown reason - in which case one would expect the laws to be similar.

It's also not unreasonable to think that there will be many similarities between the laws of the simulation and the outside reality because they'd have what they know to draw on for inspiration.

Entangled particles are stored in the same quantum register. When there are two references and it’s read, the systems says nope can’t do that and it unlinks them.

My point about time in a simulation is that it can be simulated. It's more a computation of a directed graph of causation. 'When' the computation happens in the host environment isn't material within the simulation. Trying to do things in any sort of 'to scale' time is an optimization rather than a requirement.

I don't see the problem with entanglement at all unless it was so prevalent as to direct the shape of the simulation model. If it's a lesser exceptional quantity, they can be handled as special cases, whenever one is evaluated, the universe could even in the most crude implementation, be briefly stopped, the entangled partners updated, and resumed. Pretty much like running in a VM that's doing gc if your real time clock is arbitrary information provided by the host environment.

And why bother simulating the quantum state of everything? Just generate a random one if a simulated entiry observes it.

That would be visibly inconsistent with what we observe.

Make the entity ignore it :)

You're heading towards the Problem of Evil[0]. I'd hope the simulators aren't jerks.

[0] https://en.wikipedia.org/wiki/Problem_of_evil#Logical_proble...

I think it's pretty clear they treat us the same as we treat NPCs in our video games heh

VRAM, of course, can never have enough of that :)

It's all fun and games until someone underflows velocity.

In all seriousness, finding such a bug in the software would be the best possible evidence of living in a simulation. The hard part is demonstrating that it is a a bug an not a feature.

Delayed choice quantum eraser.

you might like "Fine Structure," a novel based on this premise! https://qntm.org/structure

Alternate theory. In its core, the universe is a giant one-dimensional loop, which starts "curling" to form multidimensional forms as energy lowers and it's insufficient to keep the string straight (think about small magnets forming a line, but if not strong enough you can form looped shapes with them).

The universe has potentially infinite dimensions defined as n-dimensional Lorenz space: a space where every next dimension is perpendicular to all previous dimensions. The number of spatial dimensions depends on the overall energy level of the Universe, as well as local energy topology.

Black holes represent these inflectional shifts of spatial dimensionality. The surface of a black hole encodes a closed (no beginning, no end) space of a 2D universe, and inside that 2D universe, high-energy regions form two-dimensional black holes, circles, on the surface of the 3D blackhole. The surface (circumference) of those circles is just the original 1D string unentangled, and this is where the set of nested black holes end.

Our own universe is the surface of a 4D blackhole.

We mathematically determine black hole is just a point through which structure doesn't survive, and truly there's nothing at the center of a black hole (i.e. the black hole itself as we define it), but hovering just above the Schwarzschild radius is a whole another set of nested universes, nested in one another, and each of those nested universe is rich with emergent structures, just like in ours 3D+time universe.

We simply can't access it, because the building blocks are of a different category. Basically the building blocks for our particles are made up of what is the particles in a 2D universe (refer to string theory and how vibrating 1D strings in N dimensions form our perception of particles).

So maybe black holes aren't just "we give up, nothing happens" here. Maybe they're rather "we give up, let's unravel one spatial dimension here in order to find new energy-stable configuration".

This actually doesn't contradict your simulation theory, but it adds some unexpected twists to it. Because what looks like just spacetime eating holes may in fact be brimming with life, and maybe our entire universe is one, as well.

> a simulation with finite resources

If the physical stuff of our universe obeys conservation laws, then our universe is finite in resources by definition.

So the "simulation" bit in your statement is not needed and facilitates no understanding, only clouds the discussion with emotional baggage.

Yet the bounds of space are ever increasing, while the amount of matter is not.

This would only make sense if there is some weird arbitrary limitation for matter concentrated in an area. And then the solution would be uncontrollable and happen to fit nicely with the rest of the rules.

Simulation theory is pretty much akin to believing in god as opposed to science. It doesn't lead us to a better description of reality. Just passing off what we don't know one level up.

Not necessarily. Religion, simulation theory, multiverse theory and other currently untestable things could potentially make falsifiable predictions and lead to a greater understanding. I don't think religion specifically is going to achieve that, but neither can I say that with 100% certainty.

Anytime we unlock a new branch of science we potentially refine our knowledge of the physical world. Computer Science was the last branch we unlocked, it might lead us to a better understanding of the universe.

That's besides the point though. In fact, you can treat science ~as~ a religion if you want - this is just semantics. That doesn't diminish that in science, we utilize the simplest models that most accurately represent reality. We don't needlessly suggest an alternative wherein we are left with more complications.

This is the case when we suppose for instance, that the Big Bang is not the beginning. Be it god or a computer simulation, we then wonder how those things came to be. Ergo, more complexity that does not resolve our understanding of the world.

Geocentric for instance works, but heliocentric is simpler because then you don't have wild erratic orbits of other planets around Earth. General relativity over newtonian because while more complex, it affords a more accurate depiction of reality.

> Be it god or a computer simulation, we then wonder how those things came to be. Ergo, more complexity that does not resolve our understanding of the world.

Not necessarily. It could totally explain or universe but leave us clueless as to what lies beyond. To be fair, I expect we'll never be able to answer what lies beyond - so that's fine.

But be it a god or a computer simulation, it needs to make testable, falsifiable predictions to better or understanding of the natural world. Otherwise, true or false, it's useless to us.

> But be it a god or a computer simulation, it needs to make testable, falsifiable predictions to better or understanding of the natural world. Otherwise, true or false, it's useless to us.

true

> Not necessarily. It could totally explain or universe but leave us clueless as to what lies beyond. To be fair, I expect we'll never be able to answer what lies beyond - so that's fine.

You're right. It could totally explain or universe but leave us clueless as to what lies beyond. As does science.

An oracle might solve the halting problem, but it has nothing to say about itself: https://en.wikipedia.org/wiki/Oracle_machine#Oracles_and_hal....

It's turtles all the way down ;)

Simulated jet physicists running black-hole simulations in simulated computers in a simulated universe. Imagine the horsepower. It could be 'simulations all the way down' !

I have a theory that the "singularity" will occur when, for purely economic reasons, somebody at one of the big cloud providers decides that it's cheaper to use the other cloud provider's services for their service, not realising that the other provider is already doing that.

If the universe as we know it is a simulation, we know absolutely nothing.

I know I exist. But that's it.

I'm less sure.

And how do they simulate us having consciousness?

You just need to simulate a bunch of Keith Frankishes and Daniel Dennetts (illusionism theory of consciousness).

Emergent behavior :)

There is no 'they'. We are the simulators!

Reply on news.ycombinator.com