|
I have had questions about this topic for a long while now, and I'm mostly curious about the limitations of it. When I have brought the topic up before, the most common response is "Heisenburg Uncertainty Principle". Will that completely nullify the pipe dream that is full universe simulation, or are there ways of circumventing it? Also, if it was somehow possible for a 1:1 simulation of the universe to be accelerated and observed, and the observer has gathered that in the next thirty seconds he will pick up a black pen, but instead decides to not pick up anything at all, what happens?
|
|
#
?
Jul 25, 2016 19:19
|
|
|
|
| # ? Apr 25, 2024 02:27 |
|
|
A perfect simulation of this universe would have to run in another universe containing a computer bigger than this universe. A computer built inside this universe to simulate this universe would also have to simulate a simulation of itself simulating itself simulating itself recursive unto infinity.
|
|
#
?
Jul 25, 2016 19:25
|
|
shame on an IGA posted:A perfect simulation of this universe would have to run in another universe containing a computer bigger than this universe. I understand the second statement you made, but this one still confuses me a little bit. What if the simulation wasn't accelerated and was taken at a much slower pace? I have pretty much thrown out the idea of the machine being an oracle, enabling visions of the future, but what if it simulated to the exact moment before the simulation began?
|
|
|
#
?
Jul 25, 2016 19:30
|
|
|
Buttmeister posted:I understand the second statement you made, but this one still confuses me a little bit. What if the simulation wasn't accelerated and was taken at a much slower pace? I have pretty much thrown out the idea of the machine being an oracle, enabling visions of the future, but what if it simulated to the exact moment before the simulation began? Data storage. To simulate the behavior of every single particle in the universe, first you'll need a database with initial states for every particle in the universe. Let's keep it simple and say the universe only contains one proton. You'll need some kind of physical storage media that can encode at the very minimum the fact that this particle is a proton, its position, and it's velocity. And that coordinate system has to be an insane number of bits to keep subatomic precision over universal distances. So there's a bunch of atoms to record data on a single hydrogen nucleus, never mind a processor or RAM to do anything with that data. now multiply that by every particle in the universe. e: and yeah Heisenberg means you cant have the position and velocity at the same time, but there are PLENTY of other reasons this is impossible. shame on an IGA fucked around with this message at 20:04 on Jul 25, 2016 |
|
#
?
Jul 25, 2016 20:02
|
|
|
shame on an IGA posted:Data storage. To simulate the behavior of every single particle in the universe, first you'll need a database with initial states for every particle in the universe. double edit: holy poo poo it'd have to be a 256 bit coordinate system you're looking at almost 1 Kb/particle just for the position.
|
|
#
?
Jul 25, 2016 20:12
|
|
|
Given distances, 256 bit encoding is not quite right, you would need continuous encoding or reference encoding to objects around the object by distance and mass. Kinda like Vector graphics vs JPG Graphics. Vector graphics can scale while JPG can only zoom to a specific bit depth. After that you would need A) Collections engine and B) Elements engine as some things like gravity are weak at the particle level but hugely strong at the planetary level. Stuff like perfect spheres break down, so we would need to rewrite math to deal with subatomic probability and wave interference. Needless to say you cant simulate the universe with out knowning (or having really good idea) how it all works. This makes even the most basic process cost huge amounts of processing power (say figuring out how elements would combine).
|
|
#
?
Jul 25, 2016 21:11
|
|
|
For the sake of discussion, let's say processing power is no longer an issue. What if there is a fundamental component that both matter and energy share? Would one be able to simply program the behavior of the most basic component of all things, and all results are just macroscopic interactions between these components? For example, if I had baking soda and vinegar and I mix the two, the simulation isn't worried about how the chemicals interact with each other on an atomic or molecular level, but rather how these fundamental building blocks of matter interact en masse. I may be rambling, but if one were able to define the behaviors of the fundamental components of matter and energy and recreate the Planck epoch 1:1, then just let the simulation run, would that make any difference?
|
|
|
#
?
Jul 25, 2016 22:18
|
|
|
Buttmeister posted:For the sake of discussion, let's say processing power is no longer an issue. Well, just off the top of my head, since you're simulating the entire universe you'd need to account for things at the most basic level. Like, you could simulate (to use your example) the result of baking soda + vinegar without simulating stuff at the subatomic or molecular level, but since you're talking about the whole universe you also need to simulate what would happen if individual molecules of those substances encountered one another in any number of circumstances (while being bombarded with different amounts of radiation or any other circumstances that might require information at the subatomic level). It's not enough for something to work only in certain circumstances - it would have to work in literally all circumstances.
|
|
#
?
Jul 25, 2016 23:22
|
|
Ytlaya posted:Well, just off the top of my head, since you're simulating the entire universe you'd need to account for things at the most basic level. Like, you could simulate (to use your example) the result of baking soda + vinegar without simulating stuff at the subatomic or molecular level, but since you're talking about the whole universe you also need to simulate what would happen if individual molecules of those substances encountered one another in any number of circumstances (while being bombarded with different amounts of radiation or any other circumstances that might require information at the subatomic level). It's not enough for something to work only in certain circumstances - it would have to work in literally all circumstances. Well, my hopes of seeing into the future to see if I will ever find love have officially been stomped out. I have no more questions regarding this topic. Thank you shame on an IGA, Ytlaya, and PyRosflam.
|
|
|
#
?
Jul 26, 2016 00:51
|
|
|
A universal simulation is inconsistent with relativity as well as quantum mechanics.
|
|
#
?
Jul 26, 2016 21:13
|
|
|
Not an expert, take with shovel of salt: - Universe simulations are already performed all the time in computational chemistry, physics, biology, engineering, etc. But the resources required to simulate even a single molecule for nanoseconds are completely ridiculous and impractical. Even the nanosecond simulations that we can do right now are only possible due to making a lot of mathematical abstractions and using statistical "tricks" to reduce the number of necessary computations. - I don't really see the problem with Heisenberg uncertainty, the Standard Model (the current accepted model of that describes elementary particle interaction)is a quantum mechanical theory and considers the wave nature of particles. - We don't have a complete, working model of elementary particles and their interactions yet. So we simply can't simulate many aspects of the universe right now. Most famous example is dark matter, which is not yet included in any accepted model. - Numerical simulations are always abstractions with finite accuracy, so in the long run you could not perform a 100% accurate simulation with a current computers. e: about your questions of the predictive power of such a simulation: yeah, it's very limited. Fundamentally, the universe is probabilistic, not deterministic. You could make extremely good predictions, but they would never be perfect. The larger the scale of the prediction, the better the results since quantum mechanical effects often become negligible at large scales. If you ran an accelerated simulation on yourself, that simulation would predict the results for a scenario where you have never seen the results. You would of course behave differently after you have seen them. Simulation != Time travel paradox GABA ghoul fucked around with this message at 15:43 on Jul 27, 2016 |
|
#
?
Jul 27, 2016 15:27
|
|
|
PyRosflam posted:Given distances, 256 bit encoding is not quite right, you would need continuous encoding or reference encoding to objects around the object by distance and mass. Kinda like Vector graphics vs JPG Graphics. Vector graphics can scale while JPG can only zoom to a specific bit depth. 256 bit encoding is a fair statement. If we make the assumption that the planck length is the smallest unit of position, the known universe is approximately 5.4×1061 planck lengths across. 256 bit encoding can store values up to 1.1×1077, so it'd be able to store every possible position across the universe. You'd need three of those, for each axis X/Y/Z. (Of course this assumes the universe can be adequately described as a normal cartesian space; which is probably not true at all since relativity alone throws that concept under the bus. It's a bit challenging to design a simulation for a system we don't even understand at its most fundamental levels.)
|
|
#
?
Jul 27, 2016 15:50
|
|
|
The reality in which the universe-simulator is running may have different laws of physics to our (simulated) universe, so it is difficult to make definite pronouncements about the properties of the universe-simulator.
|
|
#
?
Jul 29, 2016 13:30
|
|
|
shame on an IGA posted:Data storage. To simulate the behavior of every single particle in the universe, first you'll need a database with initial states for every particle in the universe. The problem isn't just storage but also getting that data in the first place. At this point in time we are completely unaware of anything or anyone who has enough knowledge about this universe to simulate it with real accuracy, entirely regardless of the logistics of running said simulation.
|
|
#
?
Jul 29, 2016 14:14
|
|
|
What if the universe is actually continuous, and some super-being decided that they'd make a simulation by just quantizing everything at some ludicrously small scale?
|
|
#
?
Jul 29, 2016 17:49
|
|
|
waitwhatno posted:- I don't really see the problem with Heisenberg uncertainty, the Standard Model (the current accepted model of that describes elementary particle interaction)is a quantum mechanical theory and considers the wave nature of particles. Of course, we're typically more concerned with macroscopic things, so we can probably sweep that under the rug. quote:- We don't have a complete, working model of elementary particles and their interactions yet. So we simply can't simulate many aspects of the universe right now. Most famous example is dark matter, which is not yet included in any accepted model. So basically OP, you can't have everything. - We can simulate incredibly small-scale physics really accurately, but you're limited to small amounts of particles and minuscule timescales. - We can simulate incredibly large-scale physics, but you miss the details of things like the formation of planets / life, and other things you might be interested in. - You can simulate pieces in between, but you're always going to have to make some simplifications and sacrifices, regardless of how good your technology is.
|
|
#
?
Jul 29, 2016 20:01
|
|
|
Buttmeister posted:Also, if it was somehow possible for a 1:1 simulation of the universe to be accelerated and observed, and the observer has gathered that in the next thirty seconds he will pick up a black pen, but instead decides to not pick up anything at all, what happens? The pen is not picked up. Your reality is then different from the simulation, which in turn is now entirely useless as it is no longer accurate. You pick up the pen anyways to not invalidate the massive effort it took to make a 1:1 simulation of the universe. Alternately, no simulation is perfect, and you just don't pick up the pen. EDIT: also, one must assume that the universe is contained within itself, that all currently accepted truths are to remain constant. Hypothetically, if we were to devise a way to breach this reality and into an alternate one, the entire simulation would fail. This is mostly science fiction, but there are theories that suggest that such a thing could happen. Disclaimer: I am not an expert. 
Sole.Sushi fucked around with this message at 11:52 on Sep 6, 2016 |
|
#
?
Sep 6, 2016 11:46
|
|
|
I think GTA is a good approximation that gets better with each new version.
|
|
#
?
Sep 6, 2016 22:22
|
|
|
I just wanted to make an apple pie from scratch.
|
|
#
?
Sep 7, 2016 00:52
|
|
|
WAIT A loving MINUTE Saying we need a database bigger than our universe to store the state of a particle is like saying we need a unique identifier for EVERY character of EVERY word in EVERY poo poo post on something awful. It's also like saying we need a unique identifier/representation of every emote, every color, and every sprite, and cannot just use bbcode to call to the emote (which is pushed through the network, to a graphics library or whatever on our computer and then to the screen with same definitions but different instances).  Again, this isn't about simulating the FULL universe and EVERY particle, but rather ONE definition of everything that is the same, then combining those definitions to produce an entirely new result. TELL ME WHY I AM WRONG AGAIN TIA
|
|
|
#
?
Oct 30, 2016 16:02
|
|
|
Just like multiple worlds theory, to my mind it's just baseless speculation. Throughout all of history people have constantly insisted upon this not being the base reality or "all there is", and I don't see full universe simulation as any different than that. We think that one day we will have the technology to completely simulate a universe, therefore we must be in one of those simulations because [blank]. I actually find it less coherent than the idea that intelligent life created whole new "real" universes, because what would be the point of simulating universes? With the amount of colossal processing power and technological prowess required, what possible benefit could they gain from it? What could they learn? What's the end point?
|
|
#
?
Oct 30, 2016 16:49
|
|
|
Why simulate a universe exactly like your own? Data. Why simulate a universe not like our own? Also data.
|
|
|
#
?
Oct 30, 2016 22:52
|
|
|
weak wrists big dick posted:Why simulate a universe exactly like your own? What data could these godlike beings possibly want? If you have the ability to perfectly recreate a universe, don't you already understand the universe in all its minute detail, and any possible extrapolations therein? Seems trivial.
|
|
#
?
Oct 30, 2016 23:49
|
|
|
What if the present is the most recent instance of the simulation these beings are simulating (who could be us, trying to simulate out own universe to gather ~data~) and we are just along for the ride? It's simulations, simulations all the way down! Its really easy to argue for this, you just slip through cracks that get smaller and smaller, but never seem to disappear completely.
|
|
|
#
?
Oct 31, 2016 04:24
|
|
|
Jeza posted:What data could these godlike beings possibly want? If you have the ability to perfectly recreate a universe, don't you already understand the universe in all its minute detail, and any possible extrapolations therein? Seems trivial. why does the theory that we are currently in a simulation necessarily depend on the idea that it's a "perfect" simulation? for all we know we could be in a relatively crappy simulation that's missing out on a lot of the details that are in the real universe. in fact it seems pretty buggy in general and I hope there's a patch soon.
|
|
#
?
Oct 31, 2016 05:09
|
|
|
weak wrists big dick posted:Its really easy to argue for this, you just slip through cracks that get smaller and smaller, but never seem to disappear completely. One of the symptoms of arguing impossible to prove metaphysical questions is that it can never be unproven. Well, some people contest that we may be able to "break" the simulation, but I'd lay a hefty wager that they're doomed to be disappointed. And of course, any evidence that did so would be contestable in its own right. You tangentially bring up another point that the whole process is recursive - if you simulate a universe where they develop the capability to run universe simulations and so on it leads to infinite complexity, which would end up requiring infinite power. Whoever was in charge would have to pull the plug before that happened, so you'd never really reach a simulation of comparable development to your own. One question that has occurred to me that I don't really know how to look up is whether in order to simulate something, you actually need to simulate it. I know that sounds weird, but I can't immediately get my head round it. Do you need to make a working simulation in order to hold process all the theoretical outcomes of every action? Or does the very thought of it create the simulation? Obviously when we imagine counterfactual situations, you would not say that we create real simulations in which the actors live. But when that imagined counterfactual is complex enough, is there any way of distinguishing that from reality? Does a computer create millions of brief simulated chess worlds when it brute forces all outcomes of a move? Or is it all held theoretical? When we simulate the weather with forecasting technology, I assume its progressive and it is working as it goes to predict what will happen next. But if we're assuming near unlimited computing power, couldn't you simply reach the outcome of any given set of parameters for a universe instantly without the need for a progressive simulation? Or does that very instant calculation create a simulation in which we live what appear subjectively temporally normal lives for millennia in the blink of an eye? Earwicker posted:why does the theory that we are currently in a simulation necessarily depend on the idea that it's a "perfect" simulation? for all we know we could be in a relatively crappy simulation that's missing out on a lot of the details that are in the real universe. in fact it seems pretty buggy in general and I hope there's a patch soon. There wouldn't be much point in running faulty simulations if your intention was to gather data, as it has been presupposed that is what they would be for, and obviously Occam's Razor so far would point out that we have yet to find a single "bug". I think the idea that we are some inferior and simple universe being hosted by a vastly more interesting, dynamic and complex universe is a far more compelling one. But again it begs the question why such a civilisation would even make such weirdly retrograde simulations, and it also transgresses one of the key premises of the original argument - that we can imagine post-humans creating simulations for their benefit. Sentient life from a universe fundamentally different and more complex than our own is so far outside of our capability to conscience that applying even remotely anthropocentric logic to them seems totally meaningless to me. Jeza fucked around with this message at 16:31 on Oct 31, 2016 |
|
#
?
Oct 31, 2016 16:27
|
|
|
Throwing out the idea of simulating our universe for a moment, is there any reason why we could not simulate a different universe with rules and forces far different than those found in our own? Perhaps a simpler universe with a smaller number of starting particles that could actually be run on a desktop computer? For example, what if we got rid of a dimension and a couple of elementary forces and maybe replaced them with a universe that is constantly contracting instead of expanding? What if we just threw away gravity? Would there be any value to doing that?
|
|
#
?
Oct 31, 2016 17:22
|
|
|
kedo posted:Throwing out the idea of simulating our universe for a moment, is there any reason why we could not simulate a different universe with rules and forces far different than those found in our own? Perhaps a simpler universe with a smaller number of starting particles that could actually be run on a desktop computer? I'm currently using my computer to simulate a different universe where the rules and forces are exactly equal to the rules of the popular computer game Civilization V, yes.
|
|
#
?
Oct 31, 2016 18:23
|
|
|
Jeza posted:There wouldn't be much point in running faulty simulations if your intention was to gather data, as it has been presupposed that is what they would be for I don't think that needs to be presupposed either. For example we ourselves make simulations of weather patterns on our planet for the purpose of trying to predict what the weather will be like in specific locations in the future. These simulations are faulty, we cannot accurately 100% predict weather patterns and the data that these simulations include does not reflect the entirety of every variable that could potentially affect an area's climate or future weather patterns. However, even with these faults, our simulations are still robust enough that they are useful in predicting the rough paths of certain kinds of storms enabling us to issue warnings and evacuate people in time. A simulation doesn't have to be 100% accurate to be built or to be useful, and so I don't see why any "our universe is a simulation" theory would need that simulation to be 100% accurate either. Earwicker fucked around with this message at 19:00 on Oct 31, 2016 |
|
#
?
Oct 31, 2016 18:56
|
|
|
Earwicker posted:I don't think that needs to be presupposed either. For example we ourselves make simulations of weather patterns on our planet for the purpose of trying to predict what the weather will be like in specific locations in the future. These simulations are faulty, we cannot accurately 100% predict weather patterns and the data that these simulations include does not reflect the entirety of every variable that could potentially affect an area's climate or future weather patterns. However, even with these faults, our simulations are still robust enough that they are useful in predicting the rough paths of certain kinds of storms enabling us to issue warnings and evacuate people in time. A simulation doesn't have to be 100% accurate to be built or to be useful, and so I don't see why any "our universe is a simulation" theory would need that simulation to be 100% accurate either. This is true, but weather simulation is a short term and extremely primitive variable. Our ability to predict the weather accurately gets exponentially worse over time, even a week, so you can imagine the margin of error involved in having a faulty universe that exists over billions of years. The two are also somewhat incomparable, as weather simulation is an attempt to mimic our own world's weather patterns, but simulating a universe that perfectly mimics our own would only be obtainable once we already had a perfect understanding of it, rendering the whole process a little bit meaningless. If you understood 99% of the universe you lived in, let's say you make a 99% accurate reproduction, but once you spin up ten billion years of development, suddenly the universe already imploded or doesn't function at all. It's an exercise in extreme accuracy from the very start, beyond our wildest imagining. If the mass of your muons is off by one trillionth or the speed of light is a few femtoseconds miscalculated, the cumulative effect over time could be massive, let alone the fact that everything in the universe essentially co-exists and works in relation to one another. It's a little bit like, either you have the whole picture, or you have none of it. There's a lot of things to engage with, but primarily I think my take-away is that if you have the capability of simulating an entire universe, you can do it perfectly, but I see no recognisable motive for doing so. Simulating universes that work on different principles seems an interesting proposition, but conceptually we are engaging with an idea that our puny simian brains have no chance against. At that point you may as well throw your hands into the air and shout "I'm a Deist!" and rant about how humans are inferior pawns in some greater game.
|
|
#
?
Oct 31, 2016 21:08
|
|
|
ok here it goes if you can get a "consciousness parser" in a 1:1 simulation of our universe, you could possibly "pull" someone (who has already died in real life) and replicate their consciousness (with silicon or otherwise), you have not just figured out how to beat death, but circumvent the concept of death as a whole
|
|
|
#
?
Nov 1, 2016 12:51
|
|
|
weak wrists big dick posted:ok here it goes There's no true death within the simulation, because all relevant information is stored within the simulation and could be restored at-will. I don't know how a "real" universe could reach back temporally to completely recreate their own past, however. There's too much we don't know about the techniques involved, but from what we know now that doesn't seem possible. Death would have been solved far, far before the capability to simulate a universe however. It's child's play in comparison, assuming a materialist view of life/sentience. But being able to drag 100% accurate simulations of people from the past before a certain date seems impossible. The only way I can see it being possible would be to precisely simulate the very beginning of your universe, and I mean exactly perfectly, and then add on the proviso that the universe is completely deterministic and will play out the exactly same way as it did in your past. Two very big ifs. At the moment it seems like too much information is lost in the passage of time, and currently there is no perceivable way to retrieve that kind of information.
|
|
#
?
Nov 1, 2016 13:18
|
|
|
This thread is like watching monkeys do calculus.
|
|
#
?
Nov 5, 2016 02:25
|
|
|
You would only need to simulate components of the Universe which are presently being perceived. Why draw polygons you can't see in the game?
|
|
#
?
Nov 5, 2016 19:57
|
|
tsa posted:This thread is like watching monkeys do calculus. Any human doing calculus is a monkey doing calculus if you really think about 
|
|
|
#
?
Nov 5, 2016 19:57
|
|
DavidAlltheTime posted:You would only need to simulate components of the Universe which are presently being perceived. Why draw polygons you can't see in the game? Because the polygons/whatever that aren't seen can effect the state/attributes of stuff that is seen through time. Also, the point wasn't to "prevent" death, but circumvent it. With the bullshit made-up consciousness parser I was talking about, you could continuously store ALL human consciousness at the moment of their death up to the start of the simulation 
|
|
|
#
?
Nov 5, 2016 20:00
|
|
|
weak wrists big dick posted:ok here it goes When I was a kid, I thought it would be cool if Super Mario could come out of the video game into the real world. We'd get to hang out and it would have been awesome.
|
|
#
?
Nov 6, 2016 01:55
|
|
|
What if the simulation was performed just to farm ideas? Like it runs to completion and spits out a comprehensive list of notable inventions, songs and concepts made by faux-earth, which can then be repackaged and sold to real-earth. Our universe may only exist because "Who Let The Dogs Out?" turned out to be a multidimensional best seller.
|
|
#
?
Nov 6, 2016 08:40
|
|
|
The simulation exists so the 17 year old running it in the master universe can copy and paste only the finest shitposts to 4^256chan
|
|
#
?
Nov 6, 2016 13:31
|
|
|
|
| # ? Apr 25, 2024 02:27 |
|
|
Hi I am actually running this universe as an accelerated sim of my own one right now. All the posting is throwing off my quantum states so you guy guys keep it down a sec please? - The Management
|
|
#
?
Nov 6, 2016 17:04
|
|
