I have to disagree and I can not really understand what is so appealing about the simulation idea to many. It is an idea that will lead nowhere because you are unable to distinguish a real universe from a simulated one. In order to do this you would have to know what at least one of those scenarios looks like and then compare the universe you observe to what you know to be true about a real or simulated universe.
It is common to base such thoughts on analogies between our universe and our computers, i.e. our computers have only finite precision while the universe seems continuous, at least to a very good approximation. Therefore if we would detect rounding errors in the universe we would know that the universe is actually a simulation.
But that is at least naive and most likely not true. How would you know that a real universe does not have finite precision? Why would you assume simulation must have finite precision? And you have to answer the same question for any other feature you want to use to make the distinction.
I have to disagree and I can not really understand what is so appealing about the simulation idea to many.
You can’t see the appeal of an unfalsifiable conjecture that everything including us, is an intentional creation of some manner of vast and powerful intelligence? Aliens, Simulators, Bulk Beings, or you know... Gods.
It’s just new window dressing on an ancient dream.
Well, I can neither see the appeal of gods. I guess many people just like to have uncertainty removed from their life even if that is not actually true. And that is probably the point that I don't get, why or how people get something out of those, well let me call them lies.
The universe is big and scary, and decidedly not anthropocentric or even comprehensible,.. some people want it to be, decide that it is, and believe that.
But the question "is simulation possible, and if yes what it can look like" is new, and can have useful answers.
That is pretty interesting, but now how most people formulate the issue. After all, what we could simulate, or what could be simulated in a universe like ours, in principle is pretty interesting. The question of what another reality simulating us would be like isn't, because it could be absolutely anything at all. Questions without the hope of answers are, to me at least, not interesting.
True, if we go specifically searching for the answer to the second question, we may not find much interesting, but when we know more about the first question we can gain some insights about the second one.
So if by not interesting you mean `this door is closed now, revisit later` i completely agree with you!
Personally I find the peculiar implication of a simple question so appealing. Will we eventually be able to create simulations that one would be unable to differentiate from reality if that simulation is all they had ever known? The answer there is quite self evident, yet it simultaneously somehow implies it is highly likely that we are ourselves within a simulation.
The reason is that we can safely assume that we are within one of n + 1 realities. The n simulated realities and the 1 real reality. Unless you attribute some sort of exceptionalism bordering on creationism to our existence, there's every reason to think that n is going to proportional to the number of advanced civilizations in existence and is presumably going to be very very large. And as you increase n the odds of us actually being in the 'real' reality, which are of course just 1/(n+1), approach 0.
Like you mention it's something that doesn't necessarily go anywhere. Even if you "prove" it after death, that life you'd be experiencing at that point faces the same logical issue - at least unless the argument loses consistency at some point. Something most seem reluctant to admit is that it's also comforting as an atheist. I always expected the infinity of time after my existence to be very much akin to the infinity of time before my existence. That's just disappointing. The seemingly high probability that all is not as it seems is certainly emotionally appealing. But it's certainly not just the emotional appeal driving me there - I could obviously have chosen to invoke Pascal's Wager in the decades of time prior to considering this, but it seemed fake -- this, somehow, does not.
It also answers, or at least compellingly kicks the can, of the question of existence. If I write a program to spit out a random number, I presume that there is no entity that suddenly whisks into existence under the false impression that it is choosing a random number - which it then decides upon, before then being whisked out of existence. So why am I, the 'thing' spectating the outcome of the series of chemical and electrical reactions carried out by this body, here? Descartes would have quite the time if the answer to our little question had been yes during his time. It changes the picture so much more than the dreams that drove his ponderances.
---
From a scientific point of view, it's obviously not a great concept as it cannot be falsified. And like you mention, it also cannot really be physically proven without begging the question. On the other hand, it can be treated as a soft science and if a preponderance of evidence does lean towards supporting it then we can at least say, "Hmm.. well we can't really say anything for certain, but that's certainly interesting."
Do you think a NPC in a game could figure out that he is just a character controlled by the game logic? Current games are obviously far away from having conscious NPCs but I neither see any fundamental problem with having NPCs controlled by artificial brains nor any obvious way how such a NPC could figure out that he lives within a game.
Or what if, while you are asleep, someone essentially detaches your brain from your body and uses the signals your brain sends to control a simulation of your body in a virtual environment and feeds back the simulation output into your brain as what you see, hear, feel, and smell? How would you detect that you are no longer really in the real world when you wake up?
For example, in Minecraft you can build a redstone interferometer to detect things about the computer you're executing on and over time the statistical behavior will detect the physics of the real system.
A character in Minecraft could build a cosmic ray detector, for instance.
So yes, absolutely -- my expectation is that any intelligent structure in a computational "bubble" will be able to detect that it's in a bubble managed by some outside system.
Is your argument entirely from lack of imagination in how it could be done?
>For example, in Minecraft you can build a redstone interferometer to detect things about the computer you're executing on and over time the statistical behavior will detect the physics of the real system.
Only because minecraft does a lot of things to save from doing extra work, like unloading chunks or skipping frames if it's under load. If you remove those optimizations (and add some error-correcting ram or a redundant process to double-check calculations to protect from cosmic-ray-induced errors), then it's easy to make a minecraft-like game that's deterministic across platforms and doesn't reveal details about whatever platform it's on.
Take something like Ethereum's blockchain: it's not possible for an executable turing-complete contract to know what type of computer it's executing on, because the VM is strictly specified, and if there were any flaws that accidentally revealed information from the environment, then every nearly every node in the Ethereum p2p swarm would get different results and the blockchain would fork ridiculously and practically halt as all nodes desynchronized from each other until people updated the software to fix that bug.
Is a redstone interferometer an actually thing, a quick search didn't turn anything up? But even if you could detect influences from the outer reality within the game world, how would you decide between them being part of the game mechanics and them being shadows of the outer world given that you have no ground truth knowledge of what the game is supposed to be like?
> Is a redstone interferometer an actually thing, a quick search didn't turn anything up?
I'm probably using slightly different terminology than other people -- I haven't played much in a while.
The basic idea is to use two quasi-connected redstone wires hooked up to a piston -- one providing an update signal but no power and one providing power but no update signal.
The circuit is such that the two wires are actually branches off of a single redstone wire, using only redstone -- so you can detect if the left or right branch off of a redstone line updates first (within a single update cycle) based on if the piston extends: it only extends if the line carrying the update signal is updated after the line carrying the power signal. (If they come in the other order, it stays retracted because it updates before there's power available.)
> how would you decide between them being part of the game mechanics and them being shadows of the outer world given that you have no ground truth knowledge of what the game is supposed to be like?
You only have an operational model, just like us -- but it's eventually going to look like an extremely localized computation running their world that occasionally gets bumped by outside sources. That sounds like them discovering that their reality is a subregion hosted in another, richer reality.
The question of "game mechanics" only has meaning to the people outside in their language for discussing their actions -- at the end of the day, there is no "game", there's only what the simulation does as a facet of reality. The people inside will accurately suss out a model that includes a structure that looks like them being run as a simulation.
How they interpret that is likely to be different, but no more or less correct, than how we do.
Consider that your perception of reality is just that. What your brain tells you and what is really there are not necessarily the same thing. There was another excellent article on QuantaMagazine arguing that there's even an evolutionary imperative that would drive our brains to beneficially misrepresent reality: https://www.quantamagazine.org/the-evolutionary-argument-aga... That article is a bit of another very interesting tangent.
The point is that reality is little more than what you experience. All that simulating a reality requires is that an individual's experiences are able to be constrained to within the controlled ream. Take a child, 'disable' all senses except sight - and install permanent VR goggles. That child, lacking outside interference, will believe in a reality that is not really real. Should it develop the cognition to wonder it will likely do as we do. Why do large bodies attract one another? We can explain the operation of this through perhaps the curvature of spacetime, but that then again one can ask then why is there spacetime? At some level the answer simply ends up being that because that is the way it is - a inherent facet of our universe.
A child you made deaf like that would still discover sound as the vibrations that distorted the lenses on the eye once detection got sensitive enough.
I dispute the very notion that it's possible to cut someone off from the basic nature of reality, because you must fundamentally have constructed the cage from that reality and thus their interaction with their cage will reveal its mechanics (and hence the mechanics of reality) to them.
You can only make it take more time by suppressing noise, but eventually the "errors" will carry accurate information inside. (And someone inside can accelerate this process by causing noise in certain manners.)
> lacking outside interference
Depending on your interpretation of quantum mechanics, this is either impossible (Bohmian; maybe all of them) or insufficient (Copenhagen, MWI) -- in either case, small "errors" will reveal information about the actual nature of reality.
The Cave itself is only given definition in terms of what we might call the Ultimate (in a Platonic setting); though it's essence is to obscure the Sun (a symbol of the too bright to observe Ultimate), the very ability to do so is fundamentally related to the fact it's also a manifestation of the Ultimate.
Similarly, the Fire and anything used to cast Shadows. By careful observation of the nature of Shadows and Light from the Fire upon the Cave, we observe a manifestation of the Ultimate interacting with itself, because the Light from the Fire is merely a transposition of Truth from the Sun, while the essence of the Cave is revealed in it, and we can see the Ultimate through the lens of the Cave.
It's fundamentally impossible to separate the manifestation of an Essence from an expression of the Ultimate interacting with itself; ergo, any structure capable of self interacting and capturing its "Errors" will find in them the ability to distill a view of the Ultimate.
-- the argument is pretty endemic to any structure capable of expressing existing: what exists cannot be cut off from the nature of existence, because fundamentally it's doing that.
(Also, first time I think those philosophy classes came in handy.)
I am already not convinced that it is likely that there are a huge number of faithful simulations. It seems to me that simulating a universe faithfully would more or less consume all the resources of that universe. So all simulations could at best be heavily simplified versions of reality, like the worlds in our computer games compared to our universe. That of course doesn't preclude our universe from just being a simple game in an outer universe that would be vastly richer or huger or more complex than our universe. But it makes arguments like that all advanced civilizations will surly want to create ancestor simulations or things like that harder to justify.
The argument that we are most likely in a simulation because there are many of them and only one reality doesn't sit well with me either. Nobody at any level of nested simulations would be justified to believe that he is not simulated but actually lives in the outermost real universe. So why is it more likely that someone else lives in the real universe than us? Someone has to be the exception, why not us? I am unable to formalize it and maybe I am just committing a logical error, but something seems wrong to me if the conclusions everyone draws are, when taken together, obviously wrong, i.e. everyone concludes he is almost surely not in the outermost level but someone certainly is.
I do not think it's necessary to speculate on the nature of reason for simulations in terms of the logic. In terms of recreation, it's very interesting though. And oddly enough, I think it also adds some circumstantial evidence to simulation. We now live at what will likely be one of the most important moments in the history of our species - the internet, becoming an interplanetary civilization, automation (and the probably change to a new economic system it entails), are all happening with our lifetimes. Any of these events would be a monumental once in a civilization type event. To have them happen all at once is just unbelievably fortuitous. Beyond that interesting coincidence, this is also one of the first moments in history that would be able to be compelling reconstructed. Think about the massive data collection happening all the way from private corporations up to the NSA. If that data is not erased, it will serve as a completely untainted and objectively perfect time capsule of all true views, ideals, desires, and so on of our time. I'm not fond of this data collection, but in terms of societal impact - it is hugely beneficial in the very long term. An advanced AI would be capable of creating an incredible replication of a society, perhaps with its emotions somewhat exaggerated, by combing through and simulating this information. Of course that's just one of an infinite number of possibilities, but like I said - I find them interesting and fun to consider.
There's a logical slip in your math. It is not more likely that someone else lives in the real universe than you. It's equally improbable. And the more simulated the realities, the less probable of any given civilization existing within the real one. Just picture it as balls in a bag. There are 'n' black balls representing the number of simulated realities one may be within, but just 1 white ball representing the real reality. And like you mention, even if somebody was able to reveal and 'disable' the simulation of their reality -- the same logic holds (that the new unveiled reality is likely also false), unless there is some newfound reason to think it might not.
I don't think you can assume simulated universes would be on the same scale as the real universe.
Universe simulations would be very resource intensive and advanced civilizations would have better uses for those resources. You may very well get some highly advanced simulations from the advanced civilizations, but they would likely me on a much, much, much smaller scale than the containing universe.
E.g., if there are a billion simulated universes, they may each be a trillionth the scope of the real universe. So your odds of being in the real universe are 1/(0.001 + 1), or almost 1.
Simulations can't be 100% efficient -- If a system was, I don't think it would be considered a simulation, but just a place in the regular universe -- but even if it were, simulating a solar system would take the full resources of an entire solar system. It's hard to see why an advanced civilization would devote those kinds of resources to creating simulations or if one did, how it would survive for long to maintain it.
1) As computer engineers we understand it's doable. Give it a couple of centuries and our civilization could reach a computing capacity to simulate the entire observable universe
2) A lot of things just don't make sense. Like consciousness for example. We can't define it, and we don't even know it it's real. If you take consciousness out of the equation humans are just laborious software.
3) This is objective of course, but quite a lot of people are experiencing coincidences that seem unreal. That sounds metaphysics and most of us wouldn't touch it with a barge pole but from a simulation perspective makes total sense.
Oh and it makes a lot of people uncomfortable. For me it's one of the best ways to test someone's intellect. You bring it into a conversation and people start feeling unease.
> Give it a couple of centuries and our civilization could reach a computing capacity to simulate the entire observable universe
This is not even remotely true, not even if a perfect Moore's law were to continue for thousands of years. Man, we're hard pressed to even simulate single molecules.
> quite a lot of people are experiencing coincidences that seem unreal
Oh come on x) "Coincidences" are psychological illusions and misatribution biases at work, not the work of an intelligence behind our universe. This is the "I prayed to god and my back pain healed; it's a miracle" type of thinking.
As computer engineers we understand it's doable. Give it a couple of centuries and our civilization could reach a computing capacity to simulate the entire observable universe
This seems at the very least questionable. Why do you think you can simulate the entire universe just using a tiny fraction of the resources of the universe? It seems to me that could only work if the universe were highly compressible and I don't see any obvious evidence for that. And even then I think it could not work because if the universe is in some sense redundant and could be compressed then it actually provides less resources to work with than if it would not have those redundancies.
Take as an analogy for the later two processors executing the same code in lockstep as it is sometimes done for redundancy in safety critical systems. You can mostly get away with only simulating one of the processors but you essentially also only have one processor to work with. And while you can probably simulate the processors on the processors the simulation will probably be at least ten times slower than the actual processor if you only want to use one tenth of the available resources, i.e. clock cycles. And finally memory will be a real problem because you only have as much memory available as you want to simulate.
You don't need to simulate the entire universe, just the tiny bit where interesting things are happening (read surface of earth), places farther away can be simulated with much lower precision as long as they look plausible.
In that sense the universe is indeed highly compressible as it is mostly empty/uninteresting.
The question is how large computing device is needed to simulate a brain, and whether that device can be digital or analog. If the size of such device can be comparable to the size of human body, then we can create a useful simulation.
The part of the universe that is empty also provides nothing you can use for a computer to run a simulation on, i.e. you don't have to simulate it but it doesn't help running the simulation either. And given that we probe the entire visible universe looking at gravitational waves from across the galaxy, spectral lines in the light of galaxies at the other end of the universe, and the decay products of Higgs particles, you still need a quite good simulation or good heuristic approximations for a lot of stuff in order to not produce inconsistencies. Or you have to limit what your simulated humans can do and keep them actively away from discovering your speed optimizations. Faking parts of something as complex as our universe without it becoming obvious would probably not be a small feat.
For all we know, 99,999% of the universe could be just a projection. You don't need to simulate all of it down to the proton level, just the things we can see. And the reason our cosmological theories fail miserably in things like dark matter or dark energy could be because there's really nothing out there.
If 99.999 % of the universe is fake, than you only have to simulate 0.001 % of the universe but you also only have 0.001 % of a universe to turn into a computer and run a simulation on.
I don't know the answer but I guess one could at least put limits on that with information theoretical or thermodynamical methods. My intuition is that you can trade size for execution speed, i.e. you can simulate the entire universe with a small computer if you are willing to wait long enough. On the other hand storing the configuration of the simulated universe would probably take up the entire universe, or at least to me it is not obvious how you could compress the state of the universe in the general case. There are certainly low entropy states that would require less memory than the system provides but they are extremely rare. The only way out I can imagine is storing a low entropy state and spending even more time to calculate parts of higher entropy states on demand but never trying to store those high entropy states. But it is again not obvious to me that this would actually work because it would require rather strong bounds on possible intermediate states.
And I don't think it would be meaningful to say the universe is the simulation. A processor executing some code is not a simulation of a processor executing some code, an electron flying across the room is not a simulation of an electron flying across the room, a dog walking down the sidewalk is not a simulation of a dog walking down the sidewalk. A simulation requires, at least that would be my ad hoc definition, one system recreating the states and behaviors of a different system, they can not be the same system.
What coincidences? I have mine, a lifetimes worth, but there isn’t a framework for saying whether my experiences are somehow special and contrived by a higher intelligence, part of a simulated universe or something else, or just a misunderstood manifestation of some not yet discovered physical law. As far as I know, science doesn’t offer a technique to quantify what we experience as coincidence (beyond the trivial prayer for health example) in any way that evidentially supports the existence of a higher intelligence or design. We’ve replaced superstitions systematically with science, and this trend seems unlikely to change, although a merger of sorts seems an inevitable destination.
Experience is often personal and difficult to express using known human communication techniques, much less quantify scientifically. Heck, we can’t even agree on whether there is a hard problem. These explorations for new approaches to quantum physics simply give hope for further understanding of reality that we may have already considered (simulation theory, for instance) or others we have not. Fun to speculate, though, in the meantime.
As an aside, maybe religious folks are right, maybe MWI is true and heaven and hell exist in the multiverse. What’s possible in the mind can overwhelm to the point of shutdown given the right set of perceptions, and coincidence can certainly drive a person to insanity (or enlightenment) if crafted carefully enough. You can claim aliens contacted you, or that you have perceptions and experiences giving blindingly in-your-face obvious (to you) indication of some kind of contrived, controlled or influenced reality, but you’d better have science to back it up.
Why are you unable to distinguish it? It is not, a priori, impossible. It is entirely possible that there may be ways to determine if we are in a simulation, depending on how the simulation is done. No simulation is perfect, and it is theoretical feasible to find the imperfections.
To look for imperfections in a simulation requires knowing what the real thing actually looks like. If our universe is a simulation, where do you get the knowledge from what a real universe looks like?
That doesn't work. The rules used by the simulator could just be the rules of the real universe, you can not tell the difference. Or to put in programming terms, you are given an object x to inspect, our universe, and you have to decide whether it is an instance of class RealUniverse or an instance of class SimulatedUniverse but you have no knowledge of what any of those classes looks or behaves like.
You seem to be arguing that it isn't always possible to detect you are in a simulation, whereas the rest of us are arguing that it could be possible in some simulations.
As a very crude example, imagine if there were "invisible walls" in our universe like many video games have. Were we to bump into one, or somehow find it, we would then try to figure out a scientific explanation, but if we ended up with a bizarre, labyrinthine theory after trying our hardest, well just maybe we are in a simulation. You could then try to come up with how our simulation works, and then come up with tests to prove those theories.
Now, in the end, if we are in a simulation, that simulation (or one of its "parents") is reality, and so effectively we are all bound by the physics of the base reality, so one could argue that there is no harsh line between our simulations' physics and reality's physics, but that's another thread ;)
Maybe its an acknowledgement of something about the nature of abstraction from observation. Perhaps the true universe is and always be a simulation error away from observation -- under such a hypothesis, science will be an eternal process of finding details in our understanding which are actually accidental results of the simulation capacity of our current understanding, then figuring out how to remove the flaw, and then peaking behind the next curtain ...
It's nothing special about the simulation idea, there are many similar things. Solipsism, only your mind is real and the outside world may or does not exist. Last Thursdayism, the universe came into existence last Thursday in the exact state of last Thursday, before that there was nothing or something very different.
They all share roughly the same idea, there is some reality but by some perfect arrangement of smoke and mirrors you are by definition unable to ever learn something about that reality. There is certainly some value in discussing those ideas on a meta level, finding common characteristics, figuring out limits, but beyond this the individual ideas are, at least in my opinion, mostly a waste of time. What is the point of analyzing something that by definition leaves you stuck?
And as far as I can tell philosophy has long ago mostly accepted this and made it something like an axiom that there is an reality and we can learn about it and doesn't concern itself much anymore with ideas that would, if true, block any possible progress.
I'm not sure if the idea appeals to me or not: I was more saying "I've heard Elon and co talking about this and just assumed it was like they'd seen The Matrix again on Netflix...". But reading this article (late, I guess..) the penny dropped when I saw ideas like the boundary conditions are the maximum information density of particles, and the upper bound to information celerity. Starts to sound like the kind of limitations you'd have if you built a huuuugge simulator.
The article states that axioms they are using for the reconstruction theory end up being abstracted above the underlying reality. It does not suggest the underlying reality is many-worlds, simulation, etc.
> An experiment claims to have invalidated a decades-old criticism against pilot-wave theory, an alternative formulation of quantum mechanics that avoids the most baffling features of the subatomic universe.
After familiarizing with QM/SR concepts some years ago, I came to the conclusion that there is nothing more strange and pointless than the world without these two. If max speed was infinite (or finite but non-relative, which is essentially the same) and there was no probabilistic outcomes, then the world would "happen to the end" instantly and/or with analytically predefined series of events. From any initial state. The entire "consequence" concept would fall apart. There is no point for the universe to exist in classical & atomic-ball electron mode.
Idk why both seem so unnatural to so many people who try to get what it means; it feels like it should be obvious instead - it means evolving existence, the only thing you're able to appear in and think about it. Classical uniform fundamental time is much stranger to me than what emerges from what we have now; we simply shifted our focus to "oh" forgetting how much "wtf" was left behind.
I'm not claiming any deep-enough physics or math knowledge (layman actually), consider it a funny philosophical view.
This doesn't make universe less interesting or more predictable. Consider Mandelbrots set, it is fully analytically predefined, but only way to find it's details is to perform the whole calculation.
Similarly the universe can be predefined, but the only way of finding its state after some amount of time can be performing a calculation fully equivalent to letting the universe to live for that time.
>amount of time can be performing a calculation fully equivalent to letting the universe to live for that time
This one is very interesting since it brings back the question what time really is. Your point involves “metacircular” observer that can only work/simulate as fast as it can, but the definition of “fast” is external to it. Sure, if we simulate predefined-but-calc-required in our universe, then it will take time. But why should we measure it this way? I mean, does calculation process exist at all without something that requires time for it?
[Or we’re really talking about the same fundamental thing that lies behind both our and hypothetically deterministic universe.]
Say it is possible to calculate future of a sufficiently complex physical system like a person on a Turing machine by knowing the current state of the system.
Then, according to this idea, the calculation can not be simplified by using some kind of formula that predicts the final state, and one needs to fully compute all the intermediate states, which means your computation is fully equivalent to that person living for the time for which you want to calculate the final state.
So time here is the internal time for the system of your calculation, and not related to the time you take to calculate. This internal time is nothing more than a counter you use, and You can actually stop your calculation, change something in the initial state and calculate again.
I find this fascinating because it kind of explains the universe being predictable, and people having free will: everything is predefined, but only way to know what the choice of complex system will be is letting that system to make the choice!
If the view that consciousness can be explained without requiring quantum physics is correct, then this is true independently of specifics of simulated/real universe, and is just the way information/computation works.
But then maybe Penrose is right, and consciousness can't be calculated on a Turing machine, and we still have lots of interesting science to discover.
55 comments
[ 2.8 ms ] story [ 125 ms ] threadIt is common to base such thoughts on analogies between our universe and our computers, i.e. our computers have only finite precision while the universe seems continuous, at least to a very good approximation. Therefore if we would detect rounding errors in the universe we would know that the universe is actually a simulation.
But that is at least naive and most likely not true. How would you know that a real universe does not have finite precision? Why would you assume simulation must have finite precision? And you have to answer the same question for any other feature you want to use to make the distinction.
You can’t see the appeal of an unfalsifiable conjecture that everything including us, is an intentional creation of some manner of vast and powerful intelligence? Aliens, Simulators, Bulk Beings, or you know... Gods.
It’s just new window dressing on an ancient dream.
But the question "is simulation possible, and if yes what it can look like" is new, and can have useful answers.
That is pretty interesting, but now how most people formulate the issue. After all, what we could simulate, or what could be simulated in a universe like ours, in principle is pretty interesting. The question of what another reality simulating us would be like isn't, because it could be absolutely anything at all. Questions without the hope of answers are, to me at least, not interesting.
So if by not interesting you mean `this door is closed now, revisit later` i completely agree with you!
The reason is that we can safely assume that we are within one of n + 1 realities. The n simulated realities and the 1 real reality. Unless you attribute some sort of exceptionalism bordering on creationism to our existence, there's every reason to think that n is going to proportional to the number of advanced civilizations in existence and is presumably going to be very very large. And as you increase n the odds of us actually being in the 'real' reality, which are of course just 1/(n+1), approach 0.
Like you mention it's something that doesn't necessarily go anywhere. Even if you "prove" it after death, that life you'd be experiencing at that point faces the same logical issue - at least unless the argument loses consistency at some point. Something most seem reluctant to admit is that it's also comforting as an atheist. I always expected the infinity of time after my existence to be very much akin to the infinity of time before my existence. That's just disappointing. The seemingly high probability that all is not as it seems is certainly emotionally appealing. But it's certainly not just the emotional appeal driving me there - I could obviously have chosen to invoke Pascal's Wager in the decades of time prior to considering this, but it seemed fake -- this, somehow, does not.
It also answers, or at least compellingly kicks the can, of the question of existence. If I write a program to spit out a random number, I presume that there is no entity that suddenly whisks into existence under the false impression that it is choosing a random number - which it then decides upon, before then being whisked out of existence. So why am I, the 'thing' spectating the outcome of the series of chemical and electrical reactions carried out by this body, here? Descartes would have quite the time if the answer to our little question had been yes during his time. It changes the picture so much more than the dreams that drove his ponderances.
---
From a scientific point of view, it's obviously not a great concept as it cannot be falsified. And like you mention, it also cannot really be physically proven without begging the question. On the other hand, it can be treated as a soft science and if a preponderance of evidence does lean towards supporting it then we can at least say, "Hmm.. well we can't really say anything for certain, but that's certainly interesting."
That's a key proposition and you don't support it at all. I find the idea highly questionable.
Or what if, while you are asleep, someone essentially detaches your brain from your body and uses the signals your brain sends to control a simulation of your body in a virtual environment and feeds back the simulation output into your brain as what you see, hear, feel, and smell? How would you detect that you are no longer really in the real world when you wake up?
For example, in Minecraft you can build a redstone interferometer to detect things about the computer you're executing on and over time the statistical behavior will detect the physics of the real system.
A character in Minecraft could build a cosmic ray detector, for instance.
So yes, absolutely -- my expectation is that any intelligent structure in a computational "bubble" will be able to detect that it's in a bubble managed by some outside system.
Is your argument entirely from lack of imagination in how it could be done?
Only because minecraft does a lot of things to save from doing extra work, like unloading chunks or skipping frames if it's under load. If you remove those optimizations (and add some error-correcting ram or a redundant process to double-check calculations to protect from cosmic-ray-induced errors), then it's easy to make a minecraft-like game that's deterministic across platforms and doesn't reveal details about whatever platform it's on.
Take something like Ethereum's blockchain: it's not possible for an executable turing-complete contract to know what type of computer it's executing on, because the VM is strictly specified, and if there were any flaws that accidentally revealed information from the environment, then every nearly every node in the Ethereum p2p swarm would get different results and the blockchain would fork ridiculously and practically halt as all nodes desynchronized from each other until people updated the software to fix that bug.
I'm probably using slightly different terminology than other people -- I haven't played much in a while.
The basic idea is to use two quasi-connected redstone wires hooked up to a piston -- one providing an update signal but no power and one providing power but no update signal.
The circuit is such that the two wires are actually branches off of a single redstone wire, using only redstone -- so you can detect if the left or right branch off of a redstone line updates first (within a single update cycle) based on if the piston extends: it only extends if the line carrying the update signal is updated after the line carrying the power signal. (If they come in the other order, it stays retracted because it updates before there's power available.)
> how would you decide between them being part of the game mechanics and them being shadows of the outer world given that you have no ground truth knowledge of what the game is supposed to be like?
You only have an operational model, just like us -- but it's eventually going to look like an extremely localized computation running their world that occasionally gets bumped by outside sources. That sounds like them discovering that their reality is a subregion hosted in another, richer reality.
The question of "game mechanics" only has meaning to the people outside in their language for discussing their actions -- at the end of the day, there is no "game", there's only what the simulation does as a facet of reality. The people inside will accurately suss out a model that includes a structure that looks like them being run as a simulation.
How they interpret that is likely to be different, but no more or less correct, than how we do.
The point is that reality is little more than what you experience. All that simulating a reality requires is that an individual's experiences are able to be constrained to within the controlled ream. Take a child, 'disable' all senses except sight - and install permanent VR goggles. That child, lacking outside interference, will believe in a reality that is not really real. Should it develop the cognition to wonder it will likely do as we do. Why do large bodies attract one another? We can explain the operation of this through perhaps the curvature of spacetime, but that then again one can ask then why is there spacetime? At some level the answer simply ends up being that because that is the way it is - a inherent facet of our universe.
I dispute the very notion that it's possible to cut someone off from the basic nature of reality, because you must fundamentally have constructed the cage from that reality and thus their interaction with their cage will reveal its mechanics (and hence the mechanics of reality) to them.
You can only make it take more time by suppressing noise, but eventually the "errors" will carry accurate information inside. (And someone inside can accelerate this process by causing noise in certain manners.)
> lacking outside interference
Depending on your interpretation of quantum mechanics, this is either impossible (Bohmian; maybe all of them) or insufficient (Copenhagen, MWI) -- in either case, small "errors" will reveal information about the actual nature of reality.
The Cave itself is only given definition in terms of what we might call the Ultimate (in a Platonic setting); though it's essence is to obscure the Sun (a symbol of the too bright to observe Ultimate), the very ability to do so is fundamentally related to the fact it's also a manifestation of the Ultimate.
Similarly, the Fire and anything used to cast Shadows. By careful observation of the nature of Shadows and Light from the Fire upon the Cave, we observe a manifestation of the Ultimate interacting with itself, because the Light from the Fire is merely a transposition of Truth from the Sun, while the essence of the Cave is revealed in it, and we can see the Ultimate through the lens of the Cave.
It's fundamentally impossible to separate the manifestation of an Essence from an expression of the Ultimate interacting with itself; ergo, any structure capable of self interacting and capturing its "Errors" will find in them the ability to distill a view of the Ultimate.
-- the argument is pretty endemic to any structure capable of expressing existing: what exists cannot be cut off from the nature of existence, because fundamentally it's doing that.
(Also, first time I think those philosophy classes came in handy.)
Then again it might have already happened, and we got rolled back to the last known good image.
The argument that we are most likely in a simulation because there are many of them and only one reality doesn't sit well with me either. Nobody at any level of nested simulations would be justified to believe that he is not simulated but actually lives in the outermost real universe. So why is it more likely that someone else lives in the real universe than us? Someone has to be the exception, why not us? I am unable to formalize it and maybe I am just committing a logical error, but something seems wrong to me if the conclusions everyone draws are, when taken together, obviously wrong, i.e. everyone concludes he is almost surely not in the outermost level but someone certainly is.
There's a logical slip in your math. It is not more likely that someone else lives in the real universe than you. It's equally improbable. And the more simulated the realities, the less probable of any given civilization existing within the real one. Just picture it as balls in a bag. There are 'n' black balls representing the number of simulated realities one may be within, but just 1 white ball representing the real reality. And like you mention, even if somebody was able to reveal and 'disable' the simulation of their reality -- the same logic holds (that the new unveiled reality is likely also false), unless there is some newfound reason to think it might not.
Universe simulations would be very resource intensive and advanced civilizations would have better uses for those resources. You may very well get some highly advanced simulations from the advanced civilizations, but they would likely me on a much, much, much smaller scale than the containing universe.
E.g., if there are a billion simulated universes, they may each be a trillionth the scope of the real universe. So your odds of being in the real universe are 1/(0.001 + 1), or almost 1.
Simulations can't be 100% efficient -- If a system was, I don't think it would be considered a simulation, but just a place in the regular universe -- but even if it were, simulating a solar system would take the full resources of an entire solar system. It's hard to see why an advanced civilization would devote those kinds of resources to creating simulations or if one did, how it would survive for long to maintain it.
1) As computer engineers we understand it's doable. Give it a couple of centuries and our civilization could reach a computing capacity to simulate the entire observable universe
2) A lot of things just don't make sense. Like consciousness for example. We can't define it, and we don't even know it it's real. If you take consciousness out of the equation humans are just laborious software.
3) This is objective of course, but quite a lot of people are experiencing coincidences that seem unreal. That sounds metaphysics and most of us wouldn't touch it with a barge pole but from a simulation perspective makes total sense.
Oh and it makes a lot of people uncomfortable. For me it's one of the best ways to test someone's intellect. You bring it into a conversation and people start feeling unease.
This is not even remotely true, not even if a perfect Moore's law were to continue for thousands of years. Man, we're hard pressed to even simulate single molecules.
> quite a lot of people are experiencing coincidences that seem unreal
Oh come on x) "Coincidences" are psychological illusions and misatribution biases at work, not the work of an intelligence behind our universe. This is the "I prayed to god and my back pain healed; it's a miracle" type of thinking.
This seems at the very least questionable. Why do you think you can simulate the entire universe just using a tiny fraction of the resources of the universe? It seems to me that could only work if the universe were highly compressible and I don't see any obvious evidence for that. And even then I think it could not work because if the universe is in some sense redundant and could be compressed then it actually provides less resources to work with than if it would not have those redundancies.
Take as an analogy for the later two processors executing the same code in lockstep as it is sometimes done for redundancy in safety critical systems. You can mostly get away with only simulating one of the processors but you essentially also only have one processor to work with. And while you can probably simulate the processors on the processors the simulation will probably be at least ten times slower than the actual processor if you only want to use one tenth of the available resources, i.e. clock cycles. And finally memory will be a real problem because you only have as much memory available as you want to simulate.
In that sense the universe is indeed highly compressible as it is mostly empty/uninteresting.
The question is how large computing device is needed to simulate a brain, and whether that device can be digital or analog. If the size of such device can be comparable to the size of human body, then we can create a useful simulation.
And I don't think it would be meaningful to say the universe is the simulation. A processor executing some code is not a simulation of a processor executing some code, an electron flying across the room is not a simulation of an electron flying across the room, a dog walking down the sidewalk is not a simulation of a dog walking down the sidewalk. A simulation requires, at least that would be my ad hoc definition, one system recreating the states and behaviors of a different system, they can not be the same system.
Experience is often personal and difficult to express using known human communication techniques, much less quantify scientifically. Heck, we can’t even agree on whether there is a hard problem. These explorations for new approaches to quantum physics simply give hope for further understanding of reality that we may have already considered (simulation theory, for instance) or others we have not. Fun to speculate, though, in the meantime.
As an aside, maybe religious folks are right, maybe MWI is true and heaven and hell exist in the multiverse. What’s possible in the mind can overwhelm to the point of shutdown given the right set of perceptions, and coincidence can certainly drive a person to insanity (or enlightenment) if crafted carefully enough. You can claim aliens contacted you, or that you have perceptions and experiences giving blindingly in-your-face obvious (to you) indication of some kind of contrived, controlled or influenced reality, but you’d better have science to back it up.
Now, in the end, if we are in a simulation, that simulation (or one of its "parents") is reality, and so effectively we are all bound by the physics of the base reality, so one could argue that there is no harsh line between our simulations' physics and reality's physics, but that's another thread ;)
First, that may be true, but it also may not be true.
Secondly even if it were true, why would that change anyone’s interest or curiosity about it being a possibility?
They all share roughly the same idea, there is some reality but by some perfect arrangement of smoke and mirrors you are by definition unable to ever learn something about that reality. There is certainly some value in discussing those ideas on a meta level, finding common characteristics, figuring out limits, but beyond this the individual ideas are, at least in my opinion, mostly a waste of time. What is the point of analyzing something that by definition leaves you stuck?
And as far as I can tell philosophy has long ago mostly accepted this and made it something like an axiom that there is an reality and we can learn about it and doesn't concern itself much anymore with ideas that would, if true, block any possible progress.
https://www.quantamagazine.org/pilot-wave-theory-gains-exper...
> An experiment claims to have invalidated a decades-old criticism against pilot-wave theory, an alternative formulation of quantum mechanics that avoids the most baffling features of the subatomic universe.
Idk why both seem so unnatural to so many people who try to get what it means; it feels like it should be obvious instead - it means evolving existence, the only thing you're able to appear in and think about it. Classical uniform fundamental time is much stranger to me than what emerges from what we have now; we simply shifted our focus to "oh" forgetting how much "wtf" was left behind.
I'm not claiming any deep-enough physics or math knowledge (layman actually), consider it a funny philosophical view.
This doesn't make universe less interesting or more predictable. Consider Mandelbrots set, it is fully analytically predefined, but only way to find it's details is to perform the whole calculation.
Similarly the universe can be predefined, but the only way of finding its state after some amount of time can be performing a calculation fully equivalent to letting the universe to live for that time.
This one is very interesting since it brings back the question what time really is. Your point involves “metacircular” observer that can only work/simulate as fast as it can, but the definition of “fast” is external to it. Sure, if we simulate predefined-but-calc-required in our universe, then it will take time. But why should we measure it this way? I mean, does calculation process exist at all without something that requires time for it?
[Or we’re really talking about the same fundamental thing that lies behind both our and hypothetically deterministic universe.]
Then, according to this idea, the calculation can not be simplified by using some kind of formula that predicts the final state, and one needs to fully compute all the intermediate states, which means your computation is fully equivalent to that person living for the time for which you want to calculate the final state.
So time here is the internal time for the system of your calculation, and not related to the time you take to calculate. This internal time is nothing more than a counter you use, and You can actually stop your calculation, change something in the initial state and calculate again.
I find this fascinating because it kind of explains the universe being predictable, and people having free will: everything is predefined, but only way to know what the choice of complex system will be is letting that system to make the choice!
If the view that consciousness can be explained without requiring quantum physics is correct, then this is true independently of specifics of simulated/real universe, and is just the way information/computation works.
But then maybe Penrose is right, and consciousness can't be calculated on a Turing machine, and we still have lots of interesting science to discover.
Neutrinos aren't massless
Confirmed higgs
> Problems
Hierarchy problem
Fine tuning
Why is the cosmological constant so "wrong compared to theory"
No evidence for supersymmetry yet at 10^-17cm via LHC
No evidence for WIMPs
How do we combine QM + GR?
Do black holes destroy information? How to solve the BH firewall problem?
> Speculation
Is there supersymmetry at higher energies than the LHC can probe?
Is spacetime doomed? I.e. is space-time a hologram on the boundary or some other emergent phenomenon
Is there a multiverse?
Is decoherence a solution, or does the wave function "collapse", or simply many-worlds?
Is perturbation theory sensible or is there a deeper more elegant way to calculate scattering?
Is time emergent or fundamental?