I'd prefer slow to "in the best case, everything blows up at once - in the worst, there's a memory error that if exploited, turns your world into a dystopia".
“There is a theory which states that if ever anyone discovers exactly what the Universe is for and why it is here, it will instantly disappear and be replaced by something even more bizarre and inexplicable.
There is another theory which states that this has already happened.”
The Java simulation would be just as fast as the C one in the median case. The two problems with it are that 1) the big bang would have taken forever to start up and 2) the universe would eventually skip a beat for a stop-the-world GC cycle.
If the world is simulated then "real" time simulation speed does not have any influence on our experiences inside simulation. For all we know the next-up world can have different dimensionality from ours or no at all.
What if the simulation running our universe isn't air gapped and we were able to break out of the sandbox running us and inject ourselves into the parent universe?
Aghhh not this again. This is pseudoscience/creationism marketed to scientists. It's impossible to flasify this, since any evidence for or against simulation can be a part of the simulation.
So why bother discussing any theory that cannot outright be disproven, is what you're saying? Your comment could have been the reaction to any number of major scientific realizations throughout history. I'm not saying we are living in a simulation, but why rage against any discussion of it?
> This is pseudoscience/creationism marketed to scientists.
It's not science at all, pseudo- or otherwise. It's philosophy. And I see no link to creationism. If we're in a simulation, it's not turtles all the way down. Somebody's running the simulation. The same ontological questions we grapple with now simply apply to that person instead of to us.
Heh - no reason for that to be true. The folks running the simulation could be very unlike us - maybe N-dimensional beings, or they exist across multiple universes, or heck maybe in a 2-dimension world. I can simulate more dimensions on my computer than I exist in! Why couldn't they?
Anyway their ontological questions might not be anything like ours. Perhaps they ARE god, and don't need to wonder about that.
The creationism link, I suspect, is that some people associate unanswerable questions like "why is there something rather than nothing" with the preliminaries of some arguments for the existence of a creator.
Just like creationism, simulation is not an explanation for anything. If we are a simulation, there must be something that performs the simulation. What is that? A simulation too? In the end, there must be something "real" that is not simulated. Why that cannot be our own universe?
Basically, if we really were brains in vats, then "we are brains in vats" couldn't ever be true, because the referents of our discourse wouldn't have any causal connection to real brains and real vats, and so it wouldn't even be possible, if one were a brain in a vat, to think the thought or propound the proposition "I am a brain in a vat", and so, it's necessarily false that we are brains in vats.
This is a fascinating argument, but not because it tells us much about whether we are brains in vats. Instead, it tells us something deep and interesting about linguistic meaning, and the mysterious relationship between statements and propositions.
Putnam is not actually saying that we aren't brains in vats. Rather, he is saying that we cannot form accurate and meaningful statements to that effect if we are.
The argument, as I understand it, turns on a statement's ability to refer. Putnam's point, as you nicely summarized it, is that in order to accurately describe the world (I make a point here of not talking about whether the statement is true) a statement has to be causally related to the things it describes in the right way. This seems plausible.
But this is different (on must theories) from saying that the abstract proposition that might be represented by such a statement cannot be true. It could be! The puzzle is just whether you could form a linguistic statement to properly express that proposition due to the reference problems Putnam points out.
Putnam's own example demonstrates the point. A bunch of ants cannot draw an image that represents Winston Churchill in the grass, because they cannot know what Winston Churchill is or (probably) what he looks like. Any similarity between an ant-made image, and a real image of Winston Churchill has to be merely coincidental. Ants cannot close the causal link that we normally require in order to say that an image "represents" something. But this doesn't mean that there is no such thing as an image of Winston Churchill. Or that an ant-made imagine cannot resemble Winston Churchill. It just means that ants can't form a representation of him!
Likewise, our putative inability to form an accurate statement to the effect that we are brains in a vat does not mean that we aren't brains in a vat. An omniscient being (perhaps the one who put us in the vats!) might, for example, read our silly HN posts about brains in vats and say, "gee, its incredible that this series of symbols actually matches the symbols that you would use to form a true statement statement about the world, if only the speaker were in the necessary causal relation to my vats in order to truly say such a thing." Likewise, we say about the ants "gee, its amazing that they managed to make an imagine resembling Winston Churchill even though they don't know what he looks like and therefore can't truly be said to have 'represented' him."
(There is a parallel line of argument that you might run about our ability to know something, since knowledge, it is often said, has to be related to the external world in a certain way in order to be more than a mere belief.)
From the first episode of The Magicians, two students reason about being a brain in a vat:
Quentin: "Am I hallucinating?"
Elliot: "If you were, would asking me help?"
So, I either I don't understand Putnam's argument or it's plainly wrong.
I recognize that when we say "We are brains in the vat" it seems, at the surface, self-referentially false. Because if we are brains in a vat, we wouldn't have the experience of being brains in a vat--- we'd have the experience of walking around and being people. But I can make a statement like: "Despite my experiences, I think that I am a sentient being that is contained inside a vat."
His argument that the word "vat" refers to different things in "vat-English" and normal English doesn't matter. The concept held behind vat & brain are fundamental enough that they would be self-generating among a community of sentient beings. If we say: "I am a sentient meatcomputer inside a vase filled with liquid food" we may be saying something different, but we are still expressing the core idea.
Even though the simulated world might be vastly different than the host world, language allows abstractions such that we can make the statement.
Putnam goes on to entertain the idea of a twin earth where beeches and elmes are switched, or a world where water has a different chemical forumula than H2O. That is a better starting point, but it doesn't help. I can simply generalize my statement: "I am a sentient being experiencing a hallucination, and the brain which executes my thoughts are self-contained in an environment I cannot perceive, but which nurtures my brain while networking stimulus with it."
I'd like to add a thought experiment to this. Imagine you have two vastly different communities with vastly different chemistry, and possibly operating with more than 3space and 1time dimensions. What would happen if, by the act of an Angel, both communities were given the rules for chess?
The names of the pieces don't especially matter--- if one community never had a bishop, we can simply pick a different word. Or call it piece #4.
With enough games played and thought put into it, the communities would discover tactics. They would compare and contrast the Queen's Pawn vs King's pawn. They would figure out 3-move checkmate. They would understand why en passant was a good rule to be added. And despite have different words and names for all of these concepts, these concepts still refer to the same things. Even if one world played chess with stone pieces, and the other world used leashed rabbits with hats.
A great example of this is when I went to China and play Go (or Weiqi, as they call it). I spoke 10 words of mandarin, and they spoke zero english. I was able to play a couple of games without having any shared language about the game. In the second game, an older chinese guy would sometimes yell at me and move the stone I had just played--- essentially saying: "no, that's not the best move, HERE it is." I didn't always understand why it was the best move, but often it was plainly clear that I was incorrectly read the truth of the board.
So. The act of thought supersedes what we immediately refer to in language, by it's ability to abstract. boom!
In your thought experiment, you are presupposing many common factors; we can understand this better perhaps because you are positing "3+ dimensions" so we can think "the dimensions we see plus a couple more" which makes it seem like the other community is a superset of ours.
Likewise, in your former, generalized statement, it works if you continue to think of "our universe". The entire notion of anything remotely resembling a brain, or anything like what we understand to be cause and effect (like, nurturing something to keep it functional, or even the very concept of "functional"...) existing in the host environment (whatever that means! :) ) makes the abstract statement devoid of meaning.
Just to be clear, your penultimate paragraph does not rule out there being no knowledge distinct from belief, even if some people believe that their beliefs are related to an external world in in a certain way that somehow validates them as real. For example, many humans have created representations of gods and believe them to be realistic, but we are under no logical compulsion to agree.
Yeah, no. We realized that trying to language-lawyer reality like that doesn't really work not too long (well, actually, entirely too long) after Zeno proposed his little paradox.
Instead of rejecting this argument out of hand, as "language-lawyering," allow me to suggest a different approach. It sounds like you and I agree that Putnam's argument doesn't actually tell us that we are not brains in vats. I think Putnam would agree with us as well. So what's the real point? Any discourse about whether we are brains in vats is hopelessly defective. Even if we accidentally issued the right string of symbols, or made the right sequence of noises, our statements would be objectively meaningless, because they are not causally related to the phenomena they would otherwise describe.
What's the use of this kind of thought? Here's a candidate: any philosophical debate that can be made the target of this pattern of argument is an argument that is not worth having. It cannot lead to real knowledge even if the disputants happen to arrive at what would be the right answer. This argument takes as a premise that the conclusion "I am a brain in a vat" is not causally related to the actual brain or the actual vat. This would seem to suggest, among other things, that there can be no concrete evidence for this claim--and Putnam is telling us that if we indulge in such a debate anyway we are not merely unlikely to hit upon the right answer, but that we are doomed to talk nonsense.
Well... what kind of evidence could demonstrate a universe isn't simulated?
Continuousness, perhaps, but I don't think it is possible to prove that a universe isn't discrete unless you can measure with infinite precision. Even if it was possible, a simulation could certainly be continuous with the proper hardware. Similarly, a real universe certainly could be non-continuous, so I don't think the question is even relevant either way.
I think you may be missing the point of that specific statement. It's conceivable that we might discover that we are living in a simulation. On the other hand, trying to prove conclusively that we're not is futile.
As someone with a conservative upbringing, I can't say I've heard that argument. In my experience the most common explanation is that the dinosaurs died in Noah's flood along with the rest of the animals. The dinosaurs that survived on the ark were unable to adapt to the new environment and went extinct shortly after. While I don't doubt the "God put dinosaur bones here to test our faith" argument has been used by Christians, I can't imagine it's very prevalent.
I don't think I ever heard this growing up as a Mormon, at least not as something considered as a serious belief. My Aunt quipped once how silly it was that my Granpda (also a Mormon) believed that 'the Jews planted dinosaur bones to discredit Christianity' - he's dead now and I don't know if he actually believed that, but it's possible that he did.
My experience growing up as a Mormon is that questions of how exactly God created the world are mostly left up to science. I haven't seen any hints of evolutionary science being considered 'taboo' with any seriousness in quite a long time.
The Omphalos hypothesis holds that God made the universe approximately 6000 years ago, but made it look like it was many billions of years older. I don't think it has ever been the majority view among creationists, but some have advocated it: https://en.wikipedia.org/wiki/Omphalos_hypothesis
The interesting thing is, as it pertains to this general story, that it's not really a God-made-it-look-old (as in to trick us) issue, but that's a side effect.
When I think about the simulation hypothesis, I picture Skyrim. You boot it up, load the maps, etc, and you see a picture perfect (on the PC anyway) world that looks millions of years old. Mountains are weathered. There are pre-existing historical races that died out thousands of years ago. Portions of the terrain are covered in deep snows. There's rain; it might even be raining randomly when you spawn from a save.
Bethseda isn't lying when they make the place old. It's that way because it has to be. Who knows, our reality for God could be the same thing?
Picture the Wind Fish from Link's Awakening. Whole island with people, places and attack chickens. It was a simulation in the mind of the Wind Fish.
As someone with a conservative upbringing, I can't say I've heard that argument.
If only I'd received a nickel for every time I've seen that online. I might buy myself a latte.
(Downvote? If one is arguing for the evidence based life, then it's a bit weird to engage in non-factual hyperbole. I doubt I've encountered that more than about 30-40 times, because by the time one has seen that stupid non-argument and all the other stupid arguments about 30-40 times, one is done with the places online where people think it's clever to post such nonsense.)
I recall an episode of The Simpsons where Flanders says something to that effect; I assumed that it was a parody of creationists rather than an actual, common belief.
It was taught in a private Christian school my ex-girlfriend's daughter went to. Saw it on paper even, they weren't ashamed. Also seemed to be very against evolution and considered that to be the devil whispering in your ear.
These things have changed in the last decades. I was raised as an evangelical Christian. During my childhood we were never taught to directly disagree with science, with the idea that science and religion were separate things. Nowadays they make a point of telling that science is wrong.
I lost some respect for Chalmers when I heard his TED talk that suggested consciousness was a fundamental force. He has some good ideas about consciousness but also a bit too mystical for me to agree with.
It would have been interesting to hear Wolfram's take on the debate. The debate revolved around "why would future intelligent beings want to simulate lesser intelligent beings from the past?". But the broader picture is that given enough time, "intelligence", which really means "computation", is an emergent phenomena of the universe at large. Given enough time, the universe will simulate itself.
We perform computations in the same sense that complex weather patterns do - there's nothing intrinsically special about our ability to compute. So it's naive to think that we're the ones being simulated - the universe may be simulated, and we are but one emergent property of that universe.
On a similar basis I take some exception to this statement, regarding individual immortality:
> If we’re programs in the computer, then as long as I have a computer that’s not damaged, I can always re-run the program.
You would quite possibly need to dump the entire universe state and re-run it from that point to get anything predictably "you" back out of it - probably not what most people have in mind when they think of a brain in a vat.
A universe can't really "simulate itself". There is inherent overhead to simulation, so a perfect simulation of a physical system would have to be larger and slower than the system it's simulating.
Any simulation worth running would be an approximation of the real thing, with many corners cut. Because of butterfly effects, any approximative simulation will diverge from the original, and the more corners you cut, the faster that will happen (so ancestor simulations would require near perfect accuracy, so they would probably be prohibitively expensive).
If we are a simulation, our parent universe would be several orders of magnitude more complex than this one. It's not clear to me why such a simulation would be made (what purpose it would serve to the parent universe).
That's not necessarily true in discrete systems. For example we can cache results of common interactions for reuse. See HashLife and relatives, for example: a vastly more efficient way to compute game of Life without considering one cell at a time, but giving identical results.
I'm aware, but I think the efficiency of these methods is rather circumstantial: they work well because of the way our computers are organized. What you are comparing are two simulation methods, with are both very far removed from what they simulate.
Instead, imagine implementing HashLife inside a Game of Life universe. It is possible, because the system is Turing complete, but the overhead would be horrific: probably thousands of cells for each cell you try to simulate. You would have to duplicate the cache all over the place to minimize roundtrip costs, each cell would be a whole machinery to allow signals to be routed there, plus logic to handle synchronization issues between each "unit", and so on.
It's a very neat idea that works very well with our computer systems, but if you were to implement it using the resources of the simulated system, it would be a catastrophe. And similarly, I don't think you could build a physical "cache" that can simulate physical interactions faster than they happen, except in a few contrived cases.
> It's not clear to me why such a simulation would be made (what purpose it would serve to the parent universe).
Imagine the simulating species is running something akin to a weather simulation, and you just happened to evolve - along with the weather patterns of interest - in the simulation.
That is, it's not about you, it's about the phenomenon the simulators want to study. Maybe it's galactic evolution?
If we could prove we were running in a simulation, perhaps we could guess the purpose of the simulation, and interfere with the results in a manner that might catch the attention of the simulators?
"This is the third time in a row that the weather simulation has produced rain only on prime numbered days."
Better hope the simulators don't fix that problem by re-imaging the universe.
I had this same thought a while ago. If our universe is just the Nth universe in a chain of simulated universes, and the number_of_bits(N) < number_of_bits(N-1), then each simulation gets more and more "compressed" and approximate the deeper you go.
Unlikely. The universe has too much gratuitous detail at the subatomic level. Nothing below the electron level is really necessary. At the macro scale, the universe is way too big and mostly empty. If this is a simulation, it's inefficient by tens of orders of magnitude.
If the simulation oversimplifies on the boring parts, we probably would have picked up on that by now. Although it does make you think about the Copenhagen Interpretation.
Unlikely. The universe has too much gratuitous detail at the subatomic level. Nothing below the electron level is really necessary. At the macro scale, the universe is way too big and mostly empty. If this is a simulation, it's inefficient by tens of orders of magnitude.
Wouldn't that depend on what you're trying to simulate?
Which makes a certain degree of sense, as would the idea that exceptionally fast speed takes up all computing power and therefore creates a tradeoff with computing time. A plot of that would probably look like deflecting a vector to a different angle--which is what the relationship of traversal of space and traversal of time actually does do under high speed.
Unlikely. The universe has too much gratuitous detail at the subatomic level. Nothing below the electron level is really necessary.
If I were writing a simulation of a universe, I'd save myself a ton of cycles by only "rendering" interactions that are observed by the "players" of the simulation. Much like one possible interpretation of quantum physics.
I don't know if that would work as well as it sounds. I mean, I'm not observing what's going on in China at the moment, but it will certainly impact my life in some way, if only through the butterfly effect. In the worst case, my actions one minute into the future may depend on something as far as one light-minute from me. That's larger than the whole planet.
And there are more problems. One of them is computers: in a sense, every computer created by one of the entities you simulate becomes a new "player". You have to fully simulate these computers all the time, lest you want the players to notice computers only work when they look at them.
I mean, ultimately, it seems to me that the only way to efficiently simulate a universe without arbitrary restriction on motion or glaring artefacts is to build a 1:1 model of the thing: store the data in 3D so that things that are nearby in simulated space are nearby in physical space too, and process the data locally.
Wouldn't it be amazing if some poor grad student in the future had to explain to his supervisor that his simulation of the past failed because the people in the simulation figured out that they were being simulated? That would cause so many moral issues: do you kill your experiment, or do you let it run until big freeze?
There's certainly been something changed, if not added- the idea is slowly getting more pessimistic. Plato thought we could get out of the "cave" simply through sustained rational thought. Brains-in-vats thinkers at least said we were still true beings in the "real" world, even if deluded ones. But simulationists think we're entirely beings of the "lower realm."
There is a rather interesting science fiction movie, called "The Thirteenth Floor", that explores this concept, in fact if I remember correctly it talks about nested simulated worlds.
This makes perfect sense to me, if the theory that we are living in a simulated universe is correct, the exact same logic would apply to the beings that are running the programs to simulate us and so on, ad infinitum.
There were a couple different movies that explored this theme around the time The Matrix came out. eXistenZ is another one that I enjoyed a lot, where video games become completely immersive, but there's still a few tropes like vapid NPCs and generic place names that kind of give it away.
There was a late night TV show. 80s Twilight Zone or something of similar vintage.
Height of the Cold War, on the brink of Nuclear War which after a series of cock-ups arrives and MAD results. Camera pulls back to reveal "Game Over", then further to show two alien teens at an arcade machine. A short exchange and another coin in the slot to start at pre-life. Show ends.
> “I was driven to error-correcting codes—they’re what make browsers work. So why were they in the equations I was studying about quarks and electrons and supersymmetry? This brought me to the stark realization that I could no longer say people like Max are crazy.”
Is he loosely associating mathematical constants with arbitrary error codes that we made up in computing?
He is not. An error code is a way to refer to an error. Error-correcting code is something else entirely; it is a way to encode information in which small errors in the transmission of the information can be detected and possibly corrected.
For a simple example: I want to send you 5 bits of data, and you receive 01101. You assume (but cannot prove) that you got the right data; maybe a bit was flipped somewhere and I actually sent you 11101!
So you and I agree that we'll include a parity bit in our messages. If I'm sending you an odd number of 1-bits, the parity bit will be 1. If I'm sending you an even number of 1-bits, the parity bit will be 0.
So now, I send you five bits of data, plus one parity bit. You get the message 011011. You see that you have fives bits, 01101, and a 1 for the parity bit, which means that you should be expecting an odd number of 1-bits. That is what you have, so you can assume that as long as no more than 1 bit can be flipped in transit, you have the data with no malformations.
This concept can be extended to detect more errors (what if two bits get flipped) and to detect where the errors take place, which helps to correct them. That's why it's called error-correcting code.
Hell if I know, man. If I did, I'd have included it in my explanation! But there was a pretty fundamental misunderstanding of what was being said, and I felt like it was worth explaining what I knew so that at the very least, incredulity was being expressed for the correct reason.
If we were in a simulation and the simulation would fail to realize its purpose if a sufficiently large number of beings in the simulation became aware of the simulation and changed their behavior, how smart would it be to try to get more and more people interested in the issue and look for "smoking gun" evidence? ;=-)
iirc, 'Surface Detail' explored this in depth from one society's use of a 'virtual hell' afterlife to instill behavioral control in 'the real'. My favorite Culture character's view on the subject was expressed in 'Matter', the previous novel in the series, by ex SC agent Xide Hyrlis. In Hyrlis's view, the characters were experiencing the base level of reality based on an a moral argument because "... only reality produced ultimately by matter in the raw can be so unthinkingly cruel." and that to be otherwise "... god or programmer, the charge would be the same - that of near infinite sadistic cruelty."
The simulation hypothesis is, sadly, perhaps the only self-consistent solution to the Fermi Paradox, one that doesn't require very unlikely propositions.
All of the other suggestions require all intelligent life without any exceptions in our past light cone to behave in a consistent way, i.e. fail to build obvious megastructures, fail to come calling, etc, so as to leave everything with the appearance of being natural. The Fermi Paradox is more the Wilderness Paradox in nature. Given that a tiny faction within any sufficiently advanced civilization could on their own generate self-replicated probes that converts the observable universe to computronium, and that some portion of humanity will do that if not stopped, and that we appear in no way to have exceptional origins or be in an unusual region, that really suggests there is something fundamental that we don't yet understand.
There is no such thing as a perfect computer that runs eternally or that can simulate something with an infinite degree of precision.
If we lived in a computer simulation, we would have noticed unusual meta-physical glitches as discrepancies within the 'order of things'.
For example, if the simulator used floating points, we would actually notice that numbers end up loosing precision after a certain amount of decimal points. We would probably have tons of theories about that and wonder why the metaphysical rules that are involved in counting suddenly change at specific precisions. etc.
no glitches = no simulation
There are many more problems with this premise but to me, these types of questions are a waste of intellectual energy so I'll leave it at that I guess, have fun.
But the simulation doesn't have to be perfect. It doesn't have to be complete. Maybe only you and your friends are actually being simulated, and everything else is just a partial stub. Maybe only the relevant parts are rendered.
I actually read Nick Bostrum's original paper and it's a pretty interesting read. Basically it looks at the long tail end of things. When you predict things waaay into the future, you tend to gain accuracy for macro predictions.
For example: our sun will most likely run out of energy and die out one day. The plate border in East Africa will most likely break apart at some point causing that continent to split.
The big one: humanity will one day go extinct. ... or...we'll expand to Mars and other parts of the solar system and thrive for millions of years. There are other options of course (we get blown back to the stone age, some humans survive and technology has to restart from a few thousand years ago...repeatedly...until we either leave this planet or go extinct).
If our machine understanding continues to grow, we should be able to eventually make machines that can simulate parts of the world we're in. So either we've already done that (some other we), and we (you and I we) are the products of their simulation ... or we'll go extinct one day.
...or simulating this world, perfectly or imperfectly, isn't possibly at all. ... and we'll go extinct one day.
Maybe the glitches are too small for us to detect. Perhaps lesser than the Planck constant. Maybe most of them have been debugged away.
It is impossible to infer the capabilities of the "host" machine, being inside the simulation.
My mind can't also infer why beings advanced enough to simulate our universe would stick with something as primitive as IEEE floating point. Surely their machines would be fine with arbitrary precision?
It is well known how to represent arbitrary-precision decimal numbers in a computer, so I don't get your point about precision of counting. Furthermore, the fact that the simulating computer stores the state of our atoms in some kind of floating point doesn't mean that when we think about the concept of numbers, we are limited to floating point. The abstraction capability of the human mind means that we can think about numbers that are larger, or more precise, than any physical magnitude in the universe. It is easy to write a number way larger than the number of elementary particles in the universe, and we don't make the universe overflow when we do that.
In fact, we know that there are limits in our world to how small a physical magnitude can be - Planck's time, Planck's length, etc. This seems quite coherent with a computer with finite memory simulating our world.
If the guys creating the simulation have enough resources and skill (superhuman beings from an external universe wouldn't necessarily make bugs when coding) I don't see why we would notice any glitch. Even more so when the part of the universe that we perceive is relatively tiny, and anyway the set of things that we can and cannot notice may be determined by the program itself.
If we are simulated, it'd be interesting to know how optimized the simulation is.
Maybe the simulation is reminiscent of The Matrix in that reality is essentially an elegant artificial abstraction that acts as glue tying together the perceptions of each simulation participant.
Perhaps a proper, low-level simulation isn't computationally prohibitive, but simulating the rest of our universe is. In that case, the cosmos itself may be an illusion despite an otherwise interference-free simulation.
Or, perhaps the entire cosmos is simulated just as we are—including the hundreds of billions of galaxies present in the observable universe. This would heavily imply that computational resources are not finite.
I'd argue that regardless of whether we're in a simulation or not, the sheer vastness of the observable universe heavily suggests that space may in fact be infinite—which by extension creates the potential for simulations to exist at grand—or even infinite—scales.
I suspect that the proof against would be along the lines of:
'The theoretical minimum computer for simulating any 1 particle at realtime is at least X particles big. If it would be 1 particle it would not be a simulation. If it would take less it would not be able to capture all possible states. Eventually, every realistic simulation of a universe will either be orders of magnitude less complex than reality or will be orders of magnitude slower than realtime. In both cases realistic simulations of ones universe own past will never be possible and hardly worthwhile and thus we probably don't live in one.'
This also says that computing power cannot grow indefinetely.
This argument breaks down if we consider that we don't know anything about the "host" universe.
It is as if Minecraft beings reasoned that it would be impossible to simulate them, as the redstone circuits required would be much larger than the observable universe. Which is fine, except we have nanoscale transistors.
Minecraft is many orders of magnitude less complex than our universe. We could be in a simulation, but then the parent universe would be orders of magnitude more complex and we would not be a significant simulation of a civilization's past as the original argument goes.
Of course we might be entertainment, but to explain this universe you would then have to explain the way more complex universe in which we run. We cannot proof this is not so, but now it seems pretty similar to the god hypothesis.
That assumes our version of "particle" is the same as whatever the theoretical simulation exists in. Maybe ours are orders of magnitude less complex somehow.
The other thing is that the way to represent an infinite fabric is by only calculating what you need (similar to functional languages with infinite series). Waveform collapse upon observation/interaction would seem to have some striking commonalities with a lazy-calculation approach at first glance.
I'm not a physics genius. But to me it seems like the enthusiasm for this theory is based on some very old cultural assumptions about man vs. nature.
For example:
> “If I were a character in a computer game, I would also discover eventually that the rules seemed completely rigid and mathematical,” said Max Tegmark, a cosmologist at the Massachusetts Institute of Technology (MIT). “That just reflects the computer code in which it was written.”
Computers are not exempt from nature, of course. The reason a computer works is because the physical laws of our universe can be described mathematically. If quantum theory didn't work mathematically, then neither would a computer.
So Tegmark is probably assigning the correlation the wrong way. A computer can only exist in a universe with mathematical physical laws; therefore the detection of mathematical laws doesn't tell us whether we're in a computer ourselves.
Likewise:
> Furthermore, ideas from information theory keep showing up in physics. “In my research I found this very strange thing,” said James Gates, a theoretical physicist at the University of Maryland. “I was driven to error-correcting codes—they’re what make browsers work. So why were they in the equations I was studying about quarks and electrons and supersymmetry?
Information is also not exempt from nature; it is a physical thing that is subject to the same laws of the universe as anything else. And what we call information theory arose from our study of real physical systems, like telegraphs and telephones.
Now, maybe information theory will turn out to be a better mathematical framework for describing the laws of the universe. But that does not mean that it is any more likely that we're living in a simulation. One mathematical framework is not more "natural" than another. And there are plenty of examples of where a math framework developed in one context became useful in another.
We use computers, and I think we still see ourselves as somehow distinct from the rest of nature. This lends itself to a somewhat supernatural way of thinking, like "I use a computer to manage information, so therefore if everything is information, then there must be a super-me running a super-computer."
But taking scientific theory at face value, a computer running SimCity is just as natural as a cat, or a palm tree swaying in the breeze. If our theory says that everything is information, then so are we, and so are our computers--so there is no need for anything else to explain that.
206 comments
[ 2.0 ms ] story [ 268 ms ] threadWhy do we assume that it's written in C, as opposed to - say - Erlang? ;)
There is another theory which states that this has already happened.”
- Douglas Adams
A "slower" language would just mean that this universe's "ticks" would take longer. Cars would be just fine.
https://www.xkcd.com/505/
It's not science at all, pseudo- or otherwise. It's philosophy. And I see no link to creationism. If we're in a simulation, it's not turtles all the way down. Somebody's running the simulation. The same ontological questions we grapple with now simply apply to that person instead of to us.
and they are simulated, and so are they, and ..., until you reach turtles. Super smart turtles.
If I'm going to believe something silly it may as well involve super smart turtles.
Are they really turtles, though? Or are we subtley being tricked to think of them as turtles when in reality they are dolphins?
By turtles.
Anyway their ontological questions might not be anything like ours. Perhaps they ARE god, and don't need to wonder about that.
If they were, how would they know?
How do you know this for sure?
Also, this does not mirror creationalism. Here is your argument rephrased:
>any evidence for or against God can be a creation by God.
Anyway a universe simulation doesn't sound any more stupid (to me) than any other reason for the universe being here.
Basically, if we really were brains in vats, then "we are brains in vats" couldn't ever be true, because the referents of our discourse wouldn't have any causal connection to real brains and real vats, and so it wouldn't even be possible, if one were a brain in a vat, to think the thought or propound the proposition "I am a brain in a vat", and so, it's necessarily false that we are brains in vats.
ieas.unideb.hu/admin/file_2908.pdf
Putnam is not actually saying that we aren't brains in vats. Rather, he is saying that we cannot form accurate and meaningful statements to that effect if we are.
The argument, as I understand it, turns on a statement's ability to refer. Putnam's point, as you nicely summarized it, is that in order to accurately describe the world (I make a point here of not talking about whether the statement is true) a statement has to be causally related to the things it describes in the right way. This seems plausible.
But this is different (on must theories) from saying that the abstract proposition that might be represented by such a statement cannot be true. It could be! The puzzle is just whether you could form a linguistic statement to properly express that proposition due to the reference problems Putnam points out.
Putnam's own example demonstrates the point. A bunch of ants cannot draw an image that represents Winston Churchill in the grass, because they cannot know what Winston Churchill is or (probably) what he looks like. Any similarity between an ant-made image, and a real image of Winston Churchill has to be merely coincidental. Ants cannot close the causal link that we normally require in order to say that an image "represents" something. But this doesn't mean that there is no such thing as an image of Winston Churchill. Or that an ant-made imagine cannot resemble Winston Churchill. It just means that ants can't form a representation of him!
Likewise, our putative inability to form an accurate statement to the effect that we are brains in a vat does not mean that we aren't brains in a vat. An omniscient being (perhaps the one who put us in the vats!) might, for example, read our silly HN posts about brains in vats and say, "gee, its incredible that this series of symbols actually matches the symbols that you would use to form a true statement statement about the world, if only the speaker were in the necessary causal relation to my vats in order to truly say such a thing." Likewise, we say about the ants "gee, its amazing that they managed to make an imagine resembling Winston Churchill even though they don't know what he looks like and therefore can't truly be said to have 'represented' him."
(There is a parallel line of argument that you might run about our ability to know something, since knowledge, it is often said, has to be related to the external world in a certain way in order to be more than a mere belief.)
In any case, classic Putnam.
Since no one has any trouble uttering or understanding such a statement, I question the value of this viewpoint.
Quentin: "Am I hallucinating?"
Elliot: "If you were, would asking me help?"
So, I either I don't understand Putnam's argument or it's plainly wrong.
I recognize that when we say "We are brains in the vat" it seems, at the surface, self-referentially false. Because if we are brains in a vat, we wouldn't have the experience of being brains in a vat--- we'd have the experience of walking around and being people. But I can make a statement like: "Despite my experiences, I think that I am a sentient being that is contained inside a vat."
His argument that the word "vat" refers to different things in "vat-English" and normal English doesn't matter. The concept held behind vat & brain are fundamental enough that they would be self-generating among a community of sentient beings. If we say: "I am a sentient meatcomputer inside a vase filled with liquid food" we may be saying something different, but we are still expressing the core idea.
Even though the simulated world might be vastly different than the host world, language allows abstractions such that we can make the statement.
Putnam goes on to entertain the idea of a twin earth where beeches and elmes are switched, or a world where water has a different chemical forumula than H2O. That is a better starting point, but it doesn't help. I can simply generalize my statement: "I am a sentient being experiencing a hallucination, and the brain which executes my thoughts are self-contained in an environment I cannot perceive, but which nurtures my brain while networking stimulus with it."
I'd like to add a thought experiment to this. Imagine you have two vastly different communities with vastly different chemistry, and possibly operating with more than 3space and 1time dimensions. What would happen if, by the act of an Angel, both communities were given the rules for chess?
The names of the pieces don't especially matter--- if one community never had a bishop, we can simply pick a different word. Or call it piece #4.
With enough games played and thought put into it, the communities would discover tactics. They would compare and contrast the Queen's Pawn vs King's pawn. They would figure out 3-move checkmate. They would understand why en passant was a good rule to be added. And despite have different words and names for all of these concepts, these concepts still refer to the same things. Even if one world played chess with stone pieces, and the other world used leashed rabbits with hats.
A great example of this is when I went to China and play Go (or Weiqi, as they call it). I spoke 10 words of mandarin, and they spoke zero english. I was able to play a couple of games without having any shared language about the game. In the second game, an older chinese guy would sometimes yell at me and move the stone I had just played--- essentially saying: "no, that's not the best move, HERE it is." I didn't always understand why it was the best move, but often it was plainly clear that I was incorrectly read the truth of the board.
So. The act of thought supersedes what we immediately refer to in language, by it's ability to abstract. boom!
Likewise, in your former, generalized statement, it works if you continue to think of "our universe". The entire notion of anything remotely resembling a brain, or anything like what we understand to be cause and effect (like, nurturing something to keep it functional, or even the very concept of "functional"...) existing in the host environment (whatever that means! :) ) makes the abstract statement devoid of meaning.
What's the use of this kind of thought? Here's a candidate: any philosophical debate that can be made the target of this pattern of argument is an argument that is not worth having. It cannot lead to real knowledge even if the disputants happen to arrive at what would be the right answer. This argument takes as a premise that the conclusion "I am a brain in a vat" is not causally related to the actual brain or the actual vat. This would seem to suggest, among other things, that there can be no concrete evidence for this claim--and Putnam is telling us that if we indulge in such a debate anyway we are not merely unlikely to hit upon the right answer, but that we are doomed to talk nonsense.
This is where the argument starts to sound like creationist claims that god put the bones of dinosaurs on Earth just to test our faith.
Continuousness, perhaps, but I don't think it is possible to prove that a universe isn't discrete unless you can measure with infinite precision. Even if it was possible, a simulation could certainly be continuous with the proper hardware. Similarly, a real universe certainly could be non-continuous, so I don't think the question is even relevant either way.
My experience growing up as a Mormon is that questions of how exactly God created the world are mostly left up to science. I haven't seen any hints of evolutionary science being considered 'taboo' with any seriousness in quite a long time.
When I think about the simulation hypothesis, I picture Skyrim. You boot it up, load the maps, etc, and you see a picture perfect (on the PC anyway) world that looks millions of years old. Mountains are weathered. There are pre-existing historical races that died out thousands of years ago. Portions of the terrain are covered in deep snows. There's rain; it might even be raining randomly when you spawn from a save.
Bethseda isn't lying when they make the place old. It's that way because it has to be. Who knows, our reality for God could be the same thing?
Picture the Wind Fish from Link's Awakening. Whole island with people, places and attack chickens. It was a simulation in the mind of the Wind Fish.
If only I'd received a nickel for every time I've seen that online. I might buy myself a latte.
(Downvote? If one is arguing for the evidence based life, then it's a bit weird to engage in non-factual hyperbole. I doubt I've encountered that more than about 30-40 times, because by the time one has seen that stupid non-argument and all the other stupid arguments about 30-40 times, one is done with the places online where people think it's clever to post such nonsense.)
We perform computations in the same sense that complex weather patterns do - there's nothing intrinsically special about our ability to compute. So it's naive to think that we're the ones being simulated - the universe may be simulated, and we are but one emergent property of that universe.
> If we’re programs in the computer, then as long as I have a computer that’s not damaged, I can always re-run the program.
You would quite possibly need to dump the entire universe state and re-run it from that point to get anything predictably "you" back out of it - probably not what most people have in mind when they think of a brain in a vat.
Any simulation worth running would be an approximation of the real thing, with many corners cut. Because of butterfly effects, any approximative simulation will diverge from the original, and the more corners you cut, the faster that will happen (so ancestor simulations would require near perfect accuracy, so they would probably be prohibitively expensive).
If we are a simulation, our parent universe would be several orders of magnitude more complex than this one. It's not clear to me why such a simulation would be made (what purpose it would serve to the parent universe).
(For the uninitiated: https://en.wikipedia.org/wiki/Turtles_all_the_way_down )
Instead, imagine implementing HashLife inside a Game of Life universe. It is possible, because the system is Turing complete, but the overhead would be horrific: probably thousands of cells for each cell you try to simulate. You would have to duplicate the cache all over the place to minimize roundtrip costs, each cell would be a whole machinery to allow signals to be routed there, plus logic to handle synchronization issues between each "unit", and so on.
It's a very neat idea that works very well with our computer systems, but if you were to implement it using the resources of the simulated system, it would be a catastrophe. And similarly, I don't think you could build a physical "cache" that can simulate physical interactions faster than they happen, except in a few contrived cases.
Imagine the simulating species is running something akin to a weather simulation, and you just happened to evolve - along with the weather patterns of interest - in the simulation.
That is, it's not about you, it's about the phenomenon the simulators want to study. Maybe it's galactic evolution?
It's sad to think that we're just a side effect and being left unnoticed.
In that case, are we special (i.e. unexpected piece of code questioning about its existence)?
If we could prove we were running in a simulation, perhaps we could guess the purpose of the simulation, and interfere with the results in a manner that might catch the attention of the simulators?
"This is the third time in a row that the weather simulation has produced rain only on prime numbered days."
Better hope the simulators don't fix that problem by re-imaging the universe.
This is true of simulations in our universe. There's no way to know how external universes might work.
http://schwitzsplinters.blogspot.com/2009/01/dust-hypothesis...
http://gregegan.customer.netspace.net.au/PERMUTATION/FAQ/FAQ...
If the simulation oversimplifies on the boring parts, we probably would have picked up on that by now. Although it does make you think about the Copenhagen Interpretation.
Wouldn't that depend on what you're trying to simulate?
If I were writing a simulation of a universe, I'd save myself a ton of cycles by only "rendering" interactions that are observed by the "players" of the simulation. Much like one possible interpretation of quantum physics.
And there are more problems. One of them is computers: in a sense, every computer created by one of the entities you simulate becomes a new "player". You have to fully simulate these computers all the time, lest you want the players to notice computers only work when they look at them.
I mean, ultimately, it seems to me that the only way to efficiently simulate a universe without arbitrary restriction on motion or glaring artefacts is to build a 1:1 model of the thing: store the data in 3D so that things that are nearby in simulated space are nearby in physical space too, and process the data locally.
http://www.smbc-comics.com/index.php?id=3931
https://en.wikipedia.org/wiki/Allegory_of_the_Cave
See also, https://en.wikipedia.org/wiki/Brain_in_a_vat
There's a long and rich history of philosophers restating this possibility.
Including Accelerando by our very own https://news.ycombinator.com/user?id=cstross
This makes perfect sense to me, if the theory that we are living in a simulated universe is correct, the exact same logic would apply to the beings that are running the programs to simulate us and so on, ad infinitum.
Height of the Cold War, on the brink of Nuclear War which after a series of cock-ups arrives and MAD results. Camera pulls back to reveal "Game Over", then further to show two alien teens at an arcade machine. A short exchange and another coin in the slot to start at pre-life. Show ends.
Was refreshingly well done.
Is he loosely associating mathematical constants with arbitrary error codes that we made up in computing?
For a simple example: I want to send you 5 bits of data, and you receive 01101. You assume (but cannot prove) that you got the right data; maybe a bit was flipped somewhere and I actually sent you 11101!
So you and I agree that we'll include a parity bit in our messages. If I'm sending you an odd number of 1-bits, the parity bit will be 1. If I'm sending you an even number of 1-bits, the parity bit will be 0.
So now, I send you five bits of data, plus one parity bit. You get the message 011011. You see that you have fives bits, 01101, and a 1 for the parity bit, which means that you should be expecting an odd number of 1-bits. That is what you have, so you can assume that as long as no more than 1 bit can be flipped in transit, you have the data with no malformations.
This concept can be extended to detect more errors (what if two bits get flipped) and to detect where the errors take place, which helps to correct them. That's why it's called error-correcting code.
https://www.exratione.com/2015/05/the-cosmological-noocene/
All of the other suggestions require all intelligent life without any exceptions in our past light cone to behave in a consistent way, i.e. fail to build obvious megastructures, fail to come calling, etc, so as to leave everything with the appearance of being natural. The Fermi Paradox is more the Wilderness Paradox in nature. Given that a tiny faction within any sufficiently advanced civilization could on their own generate self-replicated probes that converts the observable universe to computronium, and that some portion of humanity will do that if not stopped, and that we appear in no way to have exceptional origins or be in an unusual region, that really suggests there is something fundamental that we don't yet understand.
Can I prove that? No. Is it false even though I can't prove it? Yes.
Not everything that cannot be disproven should be taken seriously.
If we lived in a computer simulation, we would have noticed unusual meta-physical glitches as discrepancies within the 'order of things'.
For example, if the simulator used floating points, we would actually notice that numbers end up loosing precision after a certain amount of decimal points. We would probably have tons of theories about that and wonder why the metaphysical rules that are involved in counting suddenly change at specific precisions. etc.
no glitches = no simulation
There are many more problems with this premise but to me, these types of questions are a waste of intellectual energy so I'll leave it at that I guess, have fun.
I actually read Nick Bostrum's original paper and it's a pretty interesting read. Basically it looks at the long tail end of things. When you predict things waaay into the future, you tend to gain accuracy for macro predictions.
For example: our sun will most likely run out of energy and die out one day. The plate border in East Africa will most likely break apart at some point causing that continent to split.
The big one: humanity will one day go extinct. ... or...we'll expand to Mars and other parts of the solar system and thrive for millions of years. There are other options of course (we get blown back to the stone age, some humans survive and technology has to restart from a few thousand years ago...repeatedly...until we either leave this planet or go extinct).
If our machine understanding continues to grow, we should be able to eventually make machines that can simulate parts of the world we're in. So either we've already done that (some other we), and we (you and I we) are the products of their simulation ... or we'll go extinct one day.
...or simulating this world, perfectly or imperfectly, isn't possibly at all. ... and we'll go extinct one day.
Why waste so much resources, when it can be just You being simulated?
It is impossible to infer the capabilities of the "host" machine, being inside the simulation.
My mind can't also infer why beings advanced enough to simulate our universe would stick with something as primitive as IEEE floating point. Surely their machines would be fine with arbitrary precision?
In fact, we know that there are limits in our world to how small a physical magnitude can be - Planck's time, Planck's length, etc. This seems quite coherent with a computer with finite memory simulating our world.
If the guys creating the simulation have enough resources and skill (superhuman beings from an external universe wouldn't necessarily make bugs when coding) I don't see why we would notice any glitch. Even more so when the part of the universe that we perceive is relatively tiny, and anyway the set of things that we can and cannot notice may be determined by the program itself.
Like... dark matter?
Maybe the simulation is reminiscent of The Matrix in that reality is essentially an elegant artificial abstraction that acts as glue tying together the perceptions of each simulation participant.
Perhaps a proper, low-level simulation isn't computationally prohibitive, but simulating the rest of our universe is. In that case, the cosmos itself may be an illusion despite an otherwise interference-free simulation.
Or, perhaps the entire cosmos is simulated just as we are—including the hundreds of billions of galaxies present in the observable universe. This would heavily imply that computational resources are not finite.
I'd argue that regardless of whether we're in a simulation or not, the sheer vastness of the observable universe heavily suggests that space may in fact be infinite—which by extension creates the potential for simulations to exist at grand—or even infinite—scales.
'The theoretical minimum computer for simulating any 1 particle at realtime is at least X particles big. If it would be 1 particle it would not be a simulation. If it would take less it would not be able to capture all possible states. Eventually, every realistic simulation of a universe will either be orders of magnitude less complex than reality or will be orders of magnitude slower than realtime. In both cases realistic simulations of ones universe own past will never be possible and hardly worthwhile and thus we probably don't live in one.'
This also says that computing power cannot grow indefinetely.
It is as if Minecraft beings reasoned that it would be impossible to simulate them, as the redstone circuits required would be much larger than the observable universe. Which is fine, except we have nanoscale transistors.
Of course we might be entertainment, but to explain this universe you would then have to explain the way more complex universe in which we run. We cannot proof this is not so, but now it seems pretty similar to the god hypothesis.
The other thing is that the way to represent an infinite fabric is by only calculating what you need (similar to functional languages with infinite series). Waveform collapse upon observation/interaction would seem to have some striking commonalities with a lazy-calculation approach at first glance.
For example:
> “If I were a character in a computer game, I would also discover eventually that the rules seemed completely rigid and mathematical,” said Max Tegmark, a cosmologist at the Massachusetts Institute of Technology (MIT). “That just reflects the computer code in which it was written.”
Computers are not exempt from nature, of course. The reason a computer works is because the physical laws of our universe can be described mathematically. If quantum theory didn't work mathematically, then neither would a computer.
So Tegmark is probably assigning the correlation the wrong way. A computer can only exist in a universe with mathematical physical laws; therefore the detection of mathematical laws doesn't tell us whether we're in a computer ourselves.
Likewise:
> Furthermore, ideas from information theory keep showing up in physics. “In my research I found this very strange thing,” said James Gates, a theoretical physicist at the University of Maryland. “I was driven to error-correcting codes—they’re what make browsers work. So why were they in the equations I was studying about quarks and electrons and supersymmetry?
Information is also not exempt from nature; it is a physical thing that is subject to the same laws of the universe as anything else. And what we call information theory arose from our study of real physical systems, like telegraphs and telephones.
Now, maybe information theory will turn out to be a better mathematical framework for describing the laws of the universe. But that does not mean that it is any more likely that we're living in a simulation. One mathematical framework is not more "natural" than another. And there are plenty of examples of where a math framework developed in one context became useful in another.
We use computers, and I think we still see ourselves as somehow distinct from the rest of nature. This lends itself to a somewhat supernatural way of thinking, like "I use a computer to manage information, so therefore if everything is information, then there must be a super-me running a super-computer."
But taking scientific theory at face value, a computer running SimCity is just as natural as a cat, or a palm tree swaying in the breeze. If our theory says that everything is information, then so are we, and so are our computers--so there is no need for anything else to explain that.