Science Do we live inside a computer simulation?

[Rtyh-12]

OK. I will try to explain it as clearly as I will be able to.

This whole point was not about creating a computer that could think like a human brain. It was about a computer powerful enough to simulate every single particle in a universe, with all the forces acting between them. These particles would then act like "real" particles, which of course would allow them to come together, create galaxies, stars, planets and so on. On some of these planets, life could be born, and some of this life could evolve into intelligent life. These creatures would be simulated, but they would be both alive and concious (I know it's spelt wrong, how do you spell it?). They might or might not be aware that they are living in a simulation. Their brains would contain chemicals/electric wires/other types of information transmitters/pink invisible teapots/whatever. Although this entire world exists only in a simulation, it is still real. The text I'm writing here only exists in a server, but does that make it any less real? If it wasn't, you couldn't read it. The simulated universe is just as real as the universe where the simulants live. Perhaps even that universe is simulated. Who knows??

 

[jedidia]

This whole point was not about creating a computer that could think like a human brain. It was about a computer powerful enough to simulate every single particle in a universe, with all the forces acting between them.

I know exactly what you mean, and my post above was directed towards exactly what you wrote here: To pull it off convincingly, you'd pretty much have to simulate the entire cosmos.

Just that, obviously, you seem to think it takes LESS performance than simulating "only" a brain. The problem is, you're grossly misjudging the nature of computers. "to compute" is, in another word, "to calculate". A Computer is nothing more than an insanely powerfull calculator. It can't do anything else but calculate numbers!

Because of this, everything you want to teach to a computer you have to teach him in mathemathical terms. And he needs to calculate every single bit of it. The functions we use in physics are an abstract way to describe what's happening in reality. Gravity doesn't know that it is a product of mass, and it doesn't care. It simply occurs, because that's the way things are. We found ways to calculate how much mass results in how much gravity and at what distance, but that's "only" for our practical usage. Gravity works just as well without doing any calculations.

Gravity in a computer doesn't. Nothing in a computer programm just happens because that's the way it is. Everything that is occuring in a computer programm has been calculated, which needs time.

If you play a game and look at a character, that character does look as it looks to you because the position of every single vertex in it has been calculated based on your position and viewing direction. You could not, for example, look at it from two sides at the same time. That would need another thread on the GPU to calculate the positions of the vertices from another angle, and would need double as much time. (When you're playing an online game with a friend, you have two GPUs at work, one calculating only his view and one calculating only yours).

So, you can't just programm an object "Proton" and an object "electron", throw them together and expect to get atoms or even molecuses back. Every possible interaction has to be defined by mathematical formulae, and every single step has to be calculated. It's a lot of numbers to crunch. Too many for any possibly imaginable computer, considering we're talking subatomic levels and the whole cosmos here.

I'd suggest that you'd get at least a basic understanding of computer sciences before speculating any further on the topic. The above mentiond "emperors new mind" is a very good start for people like you and me that generally don't have a clue what exactly they're talking about.

The text I'm writing here only exists in a server, but does that make it any less real? If it wasn't, you couldn't read it.

The very fact that I can read it makes it real. A computer has no concept of reality whatsoever, because of the very nature of his purely mathematical logic, he is unable to have any idea of what he is actually doing. You need a way to get a grasp of non-mathematical concepts to develop awareness, because philosophy can't be described mathemathicly. This is why it's so hard to make a good translator: Language is difficult to describe in mathematic terms (not impossible, probably, but difficult). And don't tell me google translator made a lot of progress during the last years. The method of exactly HOW google translator has been improved is pointing away from refined AI towards pure brute force computing relying on a humongous database. Google translator is doing nothing but searching for comparable forms of a sentence on the net and then does a bit of computing to decide which comes up with the highest probability of being aplicable in this case. It doesn't have a clue to what it's actually doing, nor does it have a clue what the stuff it reads actually means. Nor does it care.

Understanding this, you realise that the whole reality of the text you've written is someone who can interprete it and respond to it, which a computer can't (That would be the turing test, by the way). If noone could read what you wrote above, its whole reality would be a few zeroes and ones somewhere on a server indeed. The same aplies to handwriting by the way, which is nothing more than ink sprawled on a paper in funny patterns unless soeone can interpret it. The difference between a stain of ink and a letter is indistinguishable to anyone that doesn't know what a letter is. And a computer doesn't. You can describe to it some forms of letters in mathematical terms, and it can compare those discriptions with what it sees through a scanner and calculate the probability of it aplying to a reference it has in storage. That doesn't mean it can read, because if it happens on a wonderfully ornamented letter that somebody took an hour to draw and any human would recognise instantly if he is familiar with the basic letter, it won't be able to make anything more of it than with the stain of tomato sauce that happened to be on the letter for some reason.

 

[Quick_Nick]

So, you can't just programm an object "Proton" and an object "electron", throw them together and expect to get atoms or even molecuses back. Every possible interaction has to be defined by mathematical formulae, and every single step has to be calculated. It's a lot of numbers to crunch. Too many for any possibly imaginable computer, considering we're talking subatomic levels and the whole cosmos here.

I'd suggest that you'd get at least a basic understanding of computer sciences before speculating any further on the topic. The above mentiond "emperors new mind" is a very good start for people like you and me that generally don't have a clue what exactly they're talking about.



The very fact that I can read it makes it real. A computer has no concept of reality whatsoever, because of the very nature of his purely mathematical logic, he is unable to have any idea of what he is actually doing. You need a way to get a grasp of non-mathematical concepts to develop awareness, because philosophy can't be described mathemathicly. This is why it's so hard to make a good translator: Language is difficult to describe in mathematic terms (not impossible, probably, but difficult). And don't tell me google translator made a lot of progress during the last year. The method of exactly HOW google translator has been improved is pointing away from refined AI towards pure brute force computing relying on a humongous database. Google translator is doing nothing but searching for comparable forms of a sentence on the net and then does a bit of computing to decide which comes up with the highest probability of being aplicable in this case. It doesn't have a clue to what it's actually doing, nor does it have a clue what the stuff it reads actually means. Nor does it care.

Very good points. It seems impractical to simulate physics with conventional computers. So, perhaps the computer is some sort of cellular automaton. The information of everything is stored somewhere(in a computer or in the edge of the universe) and reacts naturally or by a set of rules to the information around it. Then we perceive this 'memory' as a three-dimensional reality, just as in a computer program. (I think I'm restating the holographic idea...I need sleep)

 

[Rtyh-12]

All laws of physics can be defined in mathematical ways. For example, gravity has a very accurate formula. And computers follow Moore's law, so their computing power doubles regularly, so one day they will be powerful enough to compute a universe. Who knows, maybe quantum computing will come and save the day!

 

[jedidia]

And computers follow Moore's law, so their computing power doubles regularly, so one day they will be powerful enough to compute a universe.

Noone ever said that Moores law aplies indefinitaley. We're on a critical point right now, where we can't make processors any smaller and where we are already expieriencing severe trouble with heat. The solution currently is multiple processors, but since we can't make them any smaller, an end of that developement is in sight too. Either we will come up with a completely new concept of something that does the same as a computer but with different means, or the end of processor speed increases is in sight. I'd say it doesn't take a century anymore before the highest practical limit of transistor based computing power is reached.

What with any possible alternative technologies that might advance further, they will operate on completely different (and currently unknown) principles and saying what might or what might be not possible with them is pure speculation.

Plus, even such a supercomputer based on the current technology, would still not be able to have anything we could call self awareness, much less create programms by itself that have such (which is basically what you suggested above).

 

[Rtyh-12]

The following website has some interesting information, although it is quite in-graspable for my little brain:

http://www.simulation-argument.com/simulation.html

 

[Urwumpe]

Just assuming that I could be so flexible, that I could bend my body that much,
that I could stick my head into my anus, then my thought would be:

Why not also stick my arms in it and climb further upwards inside my body
Yes, I know, in reality this is not possible that far.

But let us just ignore the biological arguments. I want to know,
if I could crawl to my head, would I come out of my mouth?
And where would this be then?

Or, if I would be inside my skull, standing behind my eyes and watching through the lens, tell me, what would I see? Am I looking outside me? Am I looking inside me? Does the whole world lie inside me as illusion? Do I exist? Or am I dead? Am I part of this world or am I God?

Some things are just better described in Heavy Metal lyrics. Even if they are badly translated.

 

[Artlav]

All laws of physics can be defined in mathematical ways. For example, gravity has a very accurate formula. And computers follow Moore's law, so their computing power doubles regularly

Let me ask you a simple thing:
Imagine a chess board.

Consider a board like that with 4 cells a side.
On the first cell you put 1 cent coin, on the second - 2 cents, on the third - 4 cents, and keep doubling the sum until all cells are filled.

How much money would be on the board after it? Not much, that's pocket change for some.

Now, let's double the side of the board.
Take the full-sized board, and repeat the coin placement.
How much money would be on the board when you finish?

That's exponential growth.

4 atoms - 4 cells board, 8 atoms - 8 cells board, etc when it comes to quantum level simulation on a regular computer.

Now, how many atoms are there even in a speck of dust, not considering the entire universe, or just least in one galaxy of it?

Don't even try to imagine how fast you'll run out of processing power.


Digital computer - not going to happen.
Some abstract "universe edge" driven computer - how is that different from the universe we have already?

 

[jedidia]

The following website has some interesting information, although it is quite in-graspable for my little brain:

It's exactly the same like what you posted before, just with more words and an uglier layout. It's really nothing more than wild speculation. The problem with the three assumptions is, they are too universal. Let's say we are a simulation, and our creators are the real thing, wouldn't the answer to the three questions be exactly the same? I.e. there's no logical entry point to the whole thing, which makes it indeed a religious matter, and nothing that could be proven or scientifically argued. All we know today is that with the current kind of computers it's not possible, but we don't know if someone comes up with a completely other concept of computation that is far more capable, similiar to our brain.

The problem is, of course, that while such a design would be much better at being "smart", it would most probably suck at math, which means that mathematical functions still have to be supplied by the usual number crunchers (The human brain is actually able of remarkable feats at computing numbers if you change some arrangements by just a little bit, but then it sucks at being smart too. It's basically considered a brain error and is what gave us the movie "rain man"). So for calculating your whole physical reality, you can't really count on anything faster than, say, two orders of magnitude faster than we got today (uneducated guess), which is a lot, but simply not enough.

 

[Rtyh-12]

Yeah, I know. It's the same thing I wrote before. But this is a discussion, where we just assume that something is possible, and then discuss it's consequences. It's just wild speculation, but it's quite interesting. (at least for me, I just want to hear your ideas about that)

And, about the computer power issue, I've got another theory (yes, another wild speculation!:speakcool The laws of physics in the simulated universe (our universe) can be different from those in the "real" universe. So what if they allow computers WAAAYYYY more powerful than those that can exist in our world?

Just my crazy ideas...

 

[Urwumpe]

There are not infinite possibilities how a universe could work though. In every physical law that you experience, there is a deeper sense behind it, that ensures that the universe does not ex..... im..... damn... retroplode. If the electron would have a slightly different mass, the fact that atoms couldn't form would be the least of our worries. A whole chunk of energy conservation in the universe would fail as well.

In a deeper sense, I think it doesn't matter if we are inside a computer or not, because we couldn't tell the difference. Gödels theorem of incompleteness would apply. We can't see the computer around us, we can only interact with the simulation. It is pretty much the same as the question "Am I a human dreaming to be a turtle, or a turtle dreaming to be a human?"

 

[Grover]

its just like the Matrix, and so it doesnt matter whether we are in a simulation or not, because the only way it could make any difference, is if we could get out of the simulation (ALT-F4 :rofl, which would be impossible unless there were people that werent inside it.

THAT at least is possible, but considering that human population is already growing well out of hand, its most likely that they wont want another 6 billion + people adding strain to their planet's resources (clearly we use less when we're inanimate and in a simulation).

but then imagine the infinite theory. if the only "true" civilisation makes one "Matrix" simulation, and those in the simulation evolve into sentient species and make their own simulations (for fun or otherwise). in this sub-simulation, another could be formed, and so on forever. this means that if there is a computer simulating our existence, for the huge amount of simulations that run, the computer would have to be more powerful than is possible to imagine. (also imagine the size of the universe, this would have to be in EVERY simulation, see stephen hawkings space theories for an idea on how big we're talking about)

long story short, asking that is pointless, because it makes ABSOLUTELY no diffrence either way

 

[jedidia]

The laws of physics in the simulated universe (our universe) can be different from those in the "real" universe.

Then it is not a simulation to study evolution, history or behaviour or whatever, in fact it wouldn't serve to study anything. So what point would there be in it? Certainly the three-questions argument you linked to numerous times needs the universe outside and inside the simulation to not invalidate the few traces of logic it has.

 

[Rtyh-12]

Video games.
Studying alternate histories of the universe.
Checking out what conditions life can evolve in.
Explaining students why even small perturbations in the laws of physics can cause big changes.
Creating new forms of life.

Don't know, there could be a lot of reasons.

 

[Grover]

oh we have reasons to do it al right, we have The Sims, we have Orbiter. we simulate everything, you go on holiday, the plane you travel on was tested in simulations before it ever got to the production line.

 

[Wishbone]

The real question is how would a universe be different if one fundamental constant were off by a digit.

 

[Urwumpe]

The real question is how would a universe be different if one fundamental constant were off by a digit.

And does size matter for a universe or could you maybe simulate a tiny universe which has just a few Joule of energy and draw conclusions on the universe around you?

 

[HAL9001]

maybe

 

[Wishbone]

An interesting question. Do you happen to know a decent cosmology book with "answers"?

 

[HAL9001]

even every proof, that this world is real could be simulated