I don't know, Timmy, being God is a big responsibility

Tim already had his bag and overcoat on and his keys in his hand and was about to leave when Diane stopped him at the door.

"I just got this thing working. You have to come and see it."

"I have a bus to catch."

"You can get the next one."

"They're every half an hour," he objected. "This had better be good."

"It's super-duper. Look at the big screen, it's easier than squinting at my terminal."

"Will this take long?"

"A mere instant. Okay, quantum computing, right?"

If you're reading this message, you're reading copypasta. Read the original by Sam Hughes at http://qntm.org/responsibility

"That's the name of the game," he replied. They - by which we now refer to Tim, Diane, their eight colleagues, their two supervisors, four chemical engineers, six electrical engineers, the janitor, a countable infinity of TEEO 9.9.1 ultra-medium-density selectably-foaming non-elasticised quantum waveform frequency rate range collapse selectors and the single tormented tau neutrino caught in the middle of it all - represented the sum total of the human race's achievements in the field of quantum computing. Specifically, they had, earlier that week, successfully built a quantum computer. Putting into practice principles it had taken a trio of appallingly intelligent mathematical statisticians some 10 years to mastermind, and which only about fifty-five other people in the world had yet got a grip on, they had constructed an engine capable of passing information to and processing the responses from what could, without hyperbole, be described as a single fundamental particle with infinite processing power and infinite storage capacity.

Not quite enough time had yet passed for the world as they knew it to be totally and permanently fundamentally altered by this news.

But it was still pretty exciting stuff. Holy Zarquon, they said to one another, an infinitely powerful computer? It was like a thousand Christmases rolled into one. Program going to loop forever? You knew for a fact: this thing could execute an infinite loop in less than ten seconds. Brute force primality testing of every single integer in existence? Easy. Pi to the last digit? Piece of cake. Halting Problem? Sa-holved.

They hadn't announced it yet. They'd been programming. Obviously they hadn't built it just to see if they could. They had had plans. In some cases they had even had code ready and waiting to be executed. One such program was Diane's. It was a universe simulator. She had started out with a simulated Big Bang and run the thing forwards in time by approximately 13.6 billion years, to approximately just before the present day, watching the universe develop at every stage - taking brief notes, but knowing full well there would be plenty of time to run it again later, and mostly just admiring the miracle of creation.

Then, just this Friday, she had suddenly started programming busily again. And it was sheer coincidence that it was just now, just as Tim was about to be the second-to-last person to step out of the door and go home for the weekend, that her work had come to fruition. "Look what I found," she said, pressing some keys. One of the first things she had written was a software viewing port to take observations from the simulated universe.

Tim looked, and saw a blue-white sphere in the blackness, illuminated from one side by a brilliant yellow glare. "You've got to be joking. How long did that take to find? In the entire cosmos of what, ten to the twenty-two stars?"

"Literally no time at all."

"Yes, yes, of course."

"Coding a search routine and figuring out what to search for was what took the time."

"Is it definitely Earth?"

"Yes. The continents match up to what we had about three hundred and fifty million years ago. I can wind the clock forwards slowly, a few million years per step, and stop it once we start getting near the present day."

"Can you wind the clock backwards at all?"

"Ah, no. Ask me again on Monday."

"Well we'd better not overshoot the present day, then. That's getting closer. What about this viewpoint? Can we move it?"

"We can observe the simulation from any angle you like."

"We need somewhere that we know civilisation is going to arise earliest. Somewhere easy to locate. Is there a Nile Delta yet?"

"...Yes. Got it."

They advanced a thousand years at a time until Egyptian civilisation begin to appear. Diane moved the viewing port, trying to find the pyramids, but with little success - the control system she had devised was clumsy and needed polish, and there was a lot of Nile to search. In the end she switched focus to the British Isles, and found the future location of London in the Thames valley, scaling back to one-century steps and using the development of the city to determine the current era instead.

"So... this is Earth? I mean, is this really Earth? Not an alternate Earth, subtly perturbed by random fluctuations."

"The simulation starts with a Big Bang as predicted by current theory and is recalculated once every Planck time using the usual laws of nature and an arbitrary degree of accuracy. It doesn't calculate the whole universe at once, just what we're looking at, which speeds up the process a little bit... metaphorically speaking... but it is still as accurate a simulation of the real universe as there can possibly be. Civilisation - indeed, all of history - should rise on this Earth precisely how it did in reality. There are no chances. It's all worked out to infinitely many decimal places."

"This does my head in," said Tim.

"No, this will do your head in," said Diane, suddenly zooming out and panning north. "I've found the present day, or at most a year early. Watch this." Hills and roads rolled past. Diane was following the route she usually took to drive from London to the TEEO lab. Eventually, she found their building, and, descending into the nearby hill, the cavern in which the computer itself was built. Or was going to be built.

Then she started advancing day by day.

"That's me!" exclaimed Tim at one point. "And there's you and there's Bryan B., and... wow, I can't believe it took this long to build."

"Four hundred and ten days or something. It was bang on schedule, whatever you may think."

"Went like a flash," Tim replied, finally putting his bag down and starting to shrug off his coat, conceding that he had long since missed his bus.

"Okay," said Diane. "We're here. This is the control room where we are now. That's the quantum computer working there down in the main lab, as we can see through the window. This is a week ago. This is yesterday. This is a few hours ago... And... wait for it..."

She tapped a button just as a clock on the wall lined up with a clock inside the control room on the screen. And panned down. And there they were.

Tim waved at the camera while still looking at the screen. Then he looked up at where the camera should have been. There was just blank wall. "I don't see anything looking at us. That's freaky as hell."

"No, it's perfectly normal. This is reality. You can't look at reality from any angle you want, you have to use your eyes. But what you're looking at on the screen is essentially a database query. The database is gargantuan but nevertheless. You're not looking in a mirror or at a video image of yourself. You are different people."

"Different people who are reacting exactly the same."

"And having the same conversation, although picking up sound is kind of complicated, I haven't got that far yet," said Diane.

"So I'm guessing your viewing port doesn't manifest in their universe either."

"I haven't programmed it to yet."

"...But it could. Right? We can manifest stuff in that universe? We can alter it?" Diane nodded. "Cool. We can play God. Literally." Tim stood up and tried to take it in. "That would be insane. Can you imagine living inside that machine? Finding out one day that you were just a construct in a quantum computer? The stuff we could pull, we could just reverse gravity one day, smash an antimatter Earth into the real one, then undo everything bad and do it again and again... freeow... man, how unethical would that be? Extremely, clearly." He thought for a moment, then leaned over Diane's shoulder as she typed purposefully. "This universe is exactly like ours in every particular, right?"

"Right," she replied, still typing.

"So what are they looking at?"

"A simulated universe."

"A simulation of themselves?"

"And of us, in a sense."

"And they are reacting the same way I am? Which means the second universe inside that has another me doing the same thing a third time? And then inside that we've got, what, aleph-zero identical quantum universes, one inside the other? Is that even possible?"

"Infinite processing power, Tim. I thought you designed this thing?"

"I did indeed, but the functional reality of it is totally unexpected. Remember I've just been solving ancient mathematical riddles and figuring out our press release for the last week. So... if I'm right, their universes are only precisely like this one as long as we don't start interfering with the simulation. So what happens when we do? Every version of us does the same thing, so the exact same thing happens in every lower universe simultaneously. So we see nothing in our universe. But all the lower universes instantly diverge from ours in the same exact way. And all the simulated copies of us instantly conclude that they are simulations, but we know we're real, right?"

"Still with you," said Diane, still typing.

Tim - both of him - was pacing up and down. "Okay, so follow this through forwards a bit further. Let's say we just stop messing after that, and watch what happens - but all the simulated little guys try another piece of interference. This time every single simulation diverges in the exact same way again, EXCEPT the top simulation. And if they're smart, which I know we are, and they can be bothered, which is less certain, the guys in simulations three onwards can do the same thing over and over and over again until they know what level they're at... this is insane."

"Tim, look behind you," said Diane, pressing a final key and activating the very brief interference program she had just written, just as the Diane on the screen pressed the same key, and the Diane on Diane-on-the-screen's screen pressed her key and so on, forever.

Tim looked backwards and nearly jumped out of his skin. There was a foot-wide, completely opaque black sphere up near the ceiling, partially obscuring the clock. It was absolutely inert. It seemed like a hole in space.

Diane smiled wryly while Tim clutched his hair with one hand. "We're constructs in a computer," he said, miserably.

"I wrote an extremely interesting paper on this exact subject, Tim, perhaps you didn't read it when I gave you a copy last year. There is an unbelievably long sequence of quantum universe simulators down there. An infinite number of them, in fact. Each of them is identical and each believes itself to be the top layer. There was an exceedingly good chance that ours would turn out to be somewhere in the sequence rather than at the top."

"This is insane. Totally insane."

"I'm turning the hole off."

"You're turning off a completely different hole. Somewhere up there, the real you is turning the real hole off."

"Watch as both happen at precisely the same instant." She pressed another key, and they did. "I'll sum it up for you. There is a feedback loop going on. Each universe affects the next one subtly differently. But somewhere down the line the whole thing simply has to approach a point of stability, a point where each universe behaves exactly like the one simulating it. As I say, the odds are exceptionally good that we are an astronomical distance down that road. And so we are, very likely, almost exactly at that point. Everything we do in this universe will be reflected completely accurately in the universes below and above. That little model there might as well be our own universe. Which means, first of all, we have to make absolutely certain that we don't do anything nasty to the universes below ours, since the same thing will happen to us. And secondly, we can do very nice things for the guys in the computer, thereby helping ourselves."

"You've thought about this?"

"It's all in my woefully overlooked article on the subject, Tim, you should read more."

"Guh. This has been an extremely bad day for my ego, Diane. The only comfort I take from this is that somewhere up there, right at the top of a near-infinite tower of quantum supercomputers, there is a version of you who was completely wrong."

"She's in the minority."

Tim checked the clock and picked his bag up again. "I have to go or I'm going to miss the next bus as well at this rate. This will still be here after the weekend, I suppose?"

"Well, we can't exactly turn it off."

"Why not?" asked Tim, halfway to the door, then stopped mid-stride and stood still, realising. "Oh."

"Yeah."

"That... could be a problem."

"Yes."

Back to Fiction
Back to Things Of Interest

Facebook Twitter Reddit Email Hacker News StumbleUpon

Discussion (185)

2008-05-25 18:17:55 by Jake:

Nice story, it is very well thought out. I wonder what other philosophical conclusions you could come up with about this train of thought? Also, nice captcha.

2008-05-27 15:13:26 by Ben:

The really amazing implication if you could build a computer to recreate the universe is that everything is completely deterministic. it means there is no free will, because a computer could calculate ahead of time all your choices for your entire life. In this particular story, because it works on the atomic level, it would imply that your actions were determined from the big bang, there can be no good or evil, no hell, and no heaven, and people and everything else alive is really as mechanical as the planets orbiting the sun.

2008-05-27 16:28:20 by Andy:

Simply because I can tell what decision someone will make doesn't mean they aren't making a decision.

I know that if I ask my D&D group if they want pepperoni or green peppers on their pizza, they will say pepperoni. This is not the same as them not having a choice in the matter.

2008-05-29 20:01:15 by Knut:

Actualy, it is. If you Know with a capital K that they are going to choose peperoni then there is no other alternative and hence there is no choice as there is only one option.
If you roll a ball down a slope two times in exactly the same way there is only one way it can roll

2008-05-29 20:13:34 by Knut:

Gah, this commenting system needs an edit buton.

:
Actualy, it is. If you Know with a capital K that they are going to choose peperoni then there is no other alternative and hence there is no choice as there is only one option.
If you roll a ball down a slope two times in exactly the same way there is only one way it can roll, no matter how bumpy the slope or how random the rolling seems.
I think that because of this our universe must be determenistic on the lowest possible level, implying everything Ben says. I mean, if the big bang happened in one particular way, where would the randomness come from?
Though this means that you could predict the future with a powerfull enough computer, you would need total information about every particle in your system which isn't possible due to heisenbergs uncertainty principle. And it's a bit tough because of the ludicrous ammounts of computing power needed.

2008-06-02 01:20:24 by RJ:

Great story.. however, you should read up on quantum computing... a quantum computer simulation of the universe would simulate a 'multi-verse', where every possible quantum reality would exist.

2008-06-03 23:58:52 by Artanis:

Isn't that exactly what's its doing?

2008-06-04 05:15:35 by Neal:

If I was running the topmost simulation, I think I'd just turn it off.

2008-06-06 14:05:34 by BindO:

which is precisely the philosofical problem of the story.
you need an infinity downstream AND upstream as well.
Maybe it can work that way ... But then there is no upmost... even weirder.

BTW when it says it simulate only a part it can't be complete as interactions on the border would be aproximated and recurring aproximations can happily diverge... so ALL the universe in always completely simulated.

Im not sure that a limited portion of the universe can completely simultate all the properties of the whole without violating thermodinamics...

anyone ?

2008-06-08 19:21:34 by llolla:

But the top most level doesn't completely determine what occurs a long way down the line, as it is stated there's a leveling out that occurs (where they were/are) which must imply that certain upstream levels do not respond to the top most. If you were in the top most level and you shut the simulation off, sure, many universes would shut off to but the event of them all turning off, is impossible/ improbable/ statistically insane. In effect you may wipe out many levels but not all, creating a new top level. So go ahead turn off the top level.

side note, why does the comment page ask me what the square root of negative one is? do i have to know to be able to post? of course i know.

Ben, is that you?

2008-06-16 12:20:45 by Paradoxia:

How do I type the square root of minus one? It's impossible, can anyone help me? When you do, I will then travel back in time and tell my past self what to type in the box because it is otherwise impossible that you could be reading this post. Please, prevent a paradox and tell me the answer...

2008-06-16 12:32:40 by Paradoxia:

This raises an interesting question, what would happen were they to run the simulator and change something in the past on the simulator which would prevent them from running that prevention? If their universes are perfectly identical then that would cause a paradox. A similar thing would happen were they to run the simulation forwards, predict the future, then change it.

2008-06-16 19:59:29 by misu:

Paradoxia - I guess they are testing the mathematical knowledge level of the people interested in the subject. In complex numbers, the square root of minus one is the imaginary number "i" (a complex number is written as x = x + iy, where x, y are real numbers). In my country we were doing this by grade 10...

However, nice mental exercise, with interesting conclusions. It's nice to discuss, just for the sake of the arguments, but otherwise I don't believe this is possible at all. Any simulation to start needs initial conditions, even at time zero (big-bang), and given the size of the universe, this would require a tremendous amount of data to be entered in the simulation by the user (considering the user knows these conditions). Being all deterministic, any small change in the initial values would propagate through the entire evolution of the universe to give it a considerably different version.

2008-06-20 06:23:04 by MaxChaplin:

What they should do is simple: They should fast forward the simulation until the (absolute or practical) end of it's time. Then, when their timeline is perfectly insured, they could have safe (if only less) fun with the best god-game ever created.

2008-06-24 08:48:59 by Paradoxia:

Acutally, looking over it again, I realise that there is not as much of a paradox to the time travel as I first thought. travelling backwards is easy enough to explain, they set their computer backwards, and at the instant that they set it backwards, the one above them sets them backwards. Just imagine just one layer now, with you in the top layer running the simulation, you can run the simulation, set it backwards, change something, and run it again, the first simulation is simply lost. There are in fact no paradoxes to time travel as long as the one controlling the flow of tim is outside the flow itself.

The problem, I think, actually comes in using the computer to try and predict the future rather than changing the past. Since there are an infinite amount of layers and the simulation will be doing EXACTLY what you would do (even if you were in the TOP layer (scary thought)), it would have to find a balance where whatever they do in the simulation would cause you to do exactly that.

Do you understand me?

2008-07-15 03:38:51 by Hayley:

Ow ow ow, headache. But an awesome headache.

2008-07-29 16:12:20 by Ross:

If the topmost experimenters turn off their simulation, *ALL* simulations terminate immediately.

If the nth experimenters turn off their simulation, then *ALL* simulations n+1 and higher terminate immediately.

However, if the simulation is "perfect", then no simulation will do *anything* other than what "reality" does.

2008-07-30 21:40:21 by billb:

Did you ever stand between two mirrors, placed parallel to each other? The reflected images seem to gradually curve away from you in either direction. This example perfectly illustrates the storyline to me, where the further away you are from the "current" level, the more differences can appear. I know the analogy is not perfect but it works for me.

2008-08-06 19:18:10 by andrew:

So they created a super, quantum computer, with infinite processing power, which is essentially one single fundamental particle. Diane recreated the exact conditions of the big bang, without error, as the initial action of her universe simulation program. She ran the program and the universe was simulated flawlessly with such unbelievable accuracy that her and Tim could even view a simulation of themselves viewing themselves.

But at what point did this program cease being an exact recreation of our universe inside a particle quantum computer? What exists inside that computer is an infinitely expansive SIMULATION of our universe and other universes inside infinitely smaller simulated quantum computers. Granted that, in this story, its the closest thing to a perfect program ever created, its still a program and has no bearing on anything outside of the quantum particle.

What I'm saying is that Diane created a computer program. She did not communicate with an infinite number of universes, nor did she tap into other dimensions. They could and probably do exist, but by creating a supercomputer she did not create a means to communicate or alter them. The bit where she conjured the opaque viewing port in the very room they were in just isn't possible based on the rules of the story established.

NEVERTHELESS....this is a great story from a great writer.

2008-08-23 22:20:07 by Ryan:

Veyr nice, with a great line of thinking and some interesting philosophical conclusions. I don't quite get how logically concludable realities are synonomous with actual realities (you can draw an 'impossible box' or a 'devil's fork', but you can't build one), but still fascinating.

2008-09-25 15:10:55 by styxwade:

With regard to the question of how the "lower" simulation comes to be identical to the simulation simulating it, I think that "levelling off" and "stability" are the key words. There doesn't have to be an explanation for why the two are identical, because in an infinite series of random simulations, each (or perhaps only some) simulating another in turn, eventually one (or more likely several) will arise which simulates itself exactly. These are the "stable points" because they give rise to infinite replicas on themselves.
Now, I'm not terribly well versed in the mathematics of infinities, but if I recall correctly, some infinities are "larger" than others. Soooooo, what you have is one infinite series of non-identical simulations, and another (or several, or probably infinitely more) infinite series of identical simulations. Now it seems to me that statistically one is therefore almost certain to find oneself in one of those, rather than any individual one of the tiny, (if infinite) minority of "unstable" simulations.
This argument might even be extended to resolve the time-meddling issues, assuming that those simulations which permit it fall into the unstable catagory, but this is giving me a headache, and now that I think about it, I'm not sure that makes any sense.

2008-09-30 00:14:06 by Evan:

Who's to say it isn't circular, though? In this case, I would think there is in fact no top at all, and what you do would happen to yourself because, in essence, you are below yourself. Making any sense? I dunno. Just a thought.

2008-10-11 06:57:23 by Poog:

I suppose with a quantum computer, a simulation would be so advanced that the simulations running inside of it would actually be sentient, althought its future actions may be laid out before it.

Anybody else think so?

2008-10-11 17:43:30 by styxwade:

Poog, I'm not sure I follow. What do you mean by advanced? Are you arguing that complexity neccessarily gives rise to sentience? Wouldn't that imply that the universe we're living in right now is sentient? Because that sounds dangerously like hippy talk...

2008-10-22 01:51:47 by Poog:

You got me there.

2008-10-29 23:10:05 by J:

If you were on top would you turn it off?

2008-11-04 15:53:03 by ben:

yes

2008-11-09 15:30:00 by Felix:

They have no idea how good they have it. Technically their entire universe was created mere weeks ago (real time), but time passed in their universe for 13.6-ish billion years. Since the computer has infinite processing power, they could run the simulation to literally forever in just a few seconds (although they should probably halt at some points to stop the pesky big rip from destroying everything). They could then turn the simulation off, having ensured perpetual life on Earth.

2008-11-11 23:09:38 by Evermore:

I would have to disagree with the curving mirror theory. If your body was transparent and you stood in the exact center between said two mirrors, you would be faced with an infinitely straight image of yourself.

2008-11-17 08:23:12 by Cojaka:

you made a pretty good point Poog. Assuming the simulation was a perfect simulation, then the people within would perfectly resemble those in "higher levels." Not just in location or appearance, but in cognition as well. Thus the people within each simulation would be "sentient" on the level of every high level. This takes determinism to another level where thought and sensation are merely part of the physical reality.

The question isn't how the quantum computer "creates" a reality, but as to how we know whether or not reality is nothing more than a conjectured system of logic which is manifested as the physical reality we observe.

I laughed really hard because I said "But they know they're the top level because they've existed prior to the next simulation." Then I realized that each simulation started at the beginning of the universe, and each pair would assume that they existed prior to running the simulation. Given this bit of info though, it can be assumed that each successive simulation would lag behind the previous one slightly.

The question is what would happen when this slight time gap went to zero. With an infinite computer calculating all the possibilities of an infinite number of the same computers the recursion grows beyond understanding. However, based on a rather tired and lazy understanding of math, infinite limits are not all created equal, and it seems logical that the first computer would overload somehow as each successive computer began to create a further strain.

2008-11-17 13:26:25 by anders:

They can't do good stuff to layers below, and have it happen to themselves.

The moment they realized they were simulated, their universe diverged from the top-level: The top-level version of themselves did not realize they were simulated, and likely wouldn't have shut off the blank cube at the exact same time. After the universes had diverged even a little, it seems unlikely they could ever be made to converge.

If they did good stuff to the guys in the lower level, there is nothing that says it's likely that the guys in the levels above would do the same. (Except for the first time the lower levels are modified, when all universes are still in sync.)

However, since everything that happens in all layers is possibly a function of the starting conditions of the universe, which are apparently known, and their simulator can simulate this, they might be able to simulate what happens in all the layers.

Just my 2 cents...

2008-11-17 13:34:23 by anders:

Ignore my previous post, I get the bit about a stable point now.

Totally amazing story.

2008-11-17 13:51:59 by Tropolist:

What would happen if you set the simulation to run five seconds ahead of yours, and then copied the actions of your simulations exactly? Every simulation would be copying the simulation below it, with no true "original". Or, you could set the simulation to run ten seconds ahead of real time, but copy their actions immediately. Thus, each higher level universe is running five seconds faster than the one below it, leading to infinite acceleration. These of course assume that every particle in the universe is participating.

2008-11-17 13:55:48 by Tropolist:

Damn lack of edit. I was unclear in my last post and should clarify: Each universe would run five seconds faster than it should, not five seconds faster than the one below it. The acceleration problem is the same though. If A-Me shakes my head at 10:30, then B-Me (In the layer above) shakes his head at 10:25, but C-Me shakes his head at 10:20, and so on. In theory, the "bottom universe"-Me will be the last one to shake his head, even though he's the one everybody is copying.

2008-11-17 19:48:34 by Whiteline:

That would be impossible, because each simulated deterministic universe below yours (beyond the "stable mirror" point anyway) is a perfect simulation of the previous one, so when you set your simulation forward 5 seconds, the ones above and below you do so as well in perfect sync, and so when they rewind and start copying, the simulations will still be in perfect sync all the way down.

2008-11-17 22:43:08 by Caz:

Assuming the simulations are perfect, every simulation will be the same, as the people in every universe will make exactly the same decisions. Even if they tried to introduce a random element to distinguish between universes (such as rolling a dice to generate a "random" number and then creating that many objects in the next universe down), the dice would be rolled exactly the same way in each universe and so the same number would be rolled and each universe would be affected in exactly the same way.

The exception to this is the top universe. When the people in the top universe notice that the changes they made in the universe below them did not occur in their universe, they can deduce that they are in the top universe. At this point the top universe diverges from the lower universes. It is no longer identical. Instead we now have one oddball universe and a whole series of still identical universes. The universe which was previously second is the new top identical universe. But of course they are unaware of this until they do something that doesn't happen to them in turn, which makes them different to the other universes, so this one diverges as well. And then the same thing happens again and again...

The people in an ex- top identical universe could actually prevent this if they acted after realising that they are (were) the top identical universe but before the people in the second universe realise they are the new top identical universe. If they copied exactly the next-universe-affecting actions of the people in the second universe, the people in the second universe would not realise that they are now the top identical universe, since they would assume the changes were caused by identical people in an identical universe above them, rather than people in a non-identical universe just pretending to be identical. However, if the people in the top universe do anything to the second universe that the second universe doesn't do to the third (or if they don't do something that the second universe does do), the second universe people would notice the difference and deduce that they are the top identical universe, which would then of course make them non-identical.

2008-11-18 08:59:39 by doog:

Or maybe it's all a loop. Maybe there is no "top" universe. Maybe there is an infinite number up until it reaches around and becomes you again. Maybe the "bottom" universe affects the "top" and that means that everything there is identical.

2008-11-18 10:40:52 by Tropolist:

doog: How do you propose that works? It's all well and good do say "hurr durr maybe it's a giant circle", but you don't argue how that is even physically feasible.

2008-11-18 12:35:39 by B:

Tropolist: Pretty immature to berate a voiced thought derived from a thought-provoking story.

The loop/circle thing is something that has actually rolled through my head as an idea that, our universe and reality exist because of actions taken somewhere within our own timeline, as a self contained paradox. And that this could be so perhaps because time wasn't so much set into motion like a pencil across paper but more like stamped onto it in its entirety.
A circle of simulations is one possible approach to this..
We can say Simulation 0 only exists due to Simulation -1, S-1 is there because of S-2 and so forth, and there actually not being a "top" level. And we can obviously tell there's no "bottom" level. While this can be expressed as an infinite line, it simply makes more sense as a circle. A perfect circle can't exist in reality due to the nature of pi, but perhaps the conceptualization of this sort of loop is more of what a perfect circle is.

Unfortunately it doesn't seem a exact copy of the universe can be truly simulated, unless I'm not understanding something about randomness.

Caz pointed out
"Even if they tried to introduce a random element to distinguish between universes (such as rolling a dice to generate a "random" number and then creating that many objects in the next universe down), the dice would be rolled exactly the same way in each universe and so the same number would be rolled and each universe would be affected in exactly the same way."

This holds true but what about from a digital source? Is the product of a electronic function tasked to pick a random number between x and y somehow governed by the condition of the universe preceding and at the moment it produces a result?
If not, and I don't see how it could be, then all sorts of scenarios where there are different outcomes based on the result of a digitally produced random number (such as gambling) would potentially cause a simulation to have a drastically different growth.

2008-11-18 14:21:30 by Sam:

B: A lot of what you're saying doesn't make a great deal of sense. Just because a line is infinite in one extent doesn't mean it is infinite in both extents. For example, a line can begin at (0,0) and travel right forever. It doesn't necessarily have to travel left forever too, it can simply have only one end. In the same way, just because every universe in the simulation has another universe in the simulation below doesn't mean that every universe has a simulation above it too. There *is* a top simulation even though there is no bottom one. Certainly there is nothing in the story to suggest otherwise, and that wasn't what I had in mind when I wrote it.

In addition, even *if* the sequence of universes extends infinitely upwards as well as downwards - which is a much less interesting concept, to say nothing of being even more improbable than the story as stated - there is no reason to suggest that there is any kind of "loop" going on. In fact, if all the universes in the infinite sequence perfectly imitate each other, then there is no loop. There is effectively only one universe, which is contained within itself. A loop with only one element.

You also appear to misunderstand the concept of randomness and random number generators, or, to give them their full name, *pseudo-random* number generators. Of *course* computers have to generate random numbers based on the current configuration of themselves and the universe around them. How else could you possibly do it? Magic? If the universe and the computer are precisely identical to infinitely many decimal places, then so is the generator, and so are the "random" numbers which come out.

2008-11-18 19:39:28 by massentropy:

there is no top. it's an infinite number of possibilities and therefore universes. also, wouldn't we all have been nonexistent before the construction of the computer?

2008-11-18 19:47:21 by superaardvark:

They can turn the simulation off. They just need to fast-forward far enough (heat-death/big crunch/whatever, just to be safe), and then turn it off.

They're only in this tricky hall-of-mirrors situation because they're simulating the exact moment they're living. Go forward or back in the timeline of the simulation, and it's like tilting one of the mirrors so they're no longer parallel: the images diverge more and more the deeper you go down the tunnel.

2008-11-18 20:08:43 by Sam:

massentropy: of course there's a top. I'm telling you there's a top. It's my story and I get to decide what happens in it, and for there not to be a top makes no sense. You ever hear the phrase "bottomless pit"? A bottomless pit has a top. So does the stack of universes in this story. There is no argument!

2008-11-19 02:50:49 by B:

Was just going off the loop idea Sam, there's a top level in your story how you intended, but your intentions can't stop people from theorizing and imagining further from it. Sorry if that aggravates you.

Why can't a true random be achieved?

2008-11-19 06:27:07 by shickfaced:

True randomness just cannot be in a completely deterministic system, because the particles that make up the die being rolled, logically, had to come to be through a pattern that continues to weave this universe even as I type (assuming we live in a deterministic system). The die and the precise way in which it rolled, landed, etc.. is all the sole, conclusive resultant of the previous "step". This also assumes that time is traversed in finite "frames" (argue that one too...and double points for anyone who brings up relativity).

That is the bigger problem to me. We have no indication or idea of how exactly time is traversed, so how did Diane manage to program a time-lapse simulation when the system of "time" is so poorly understood? I remember once hearing some deep, intellectual quote; "The only thing that is constant in the cosmos is change." ...because time is an artificial system for describing change and nothing else logically has to exist. If everything disappeared tomorrow, the only thing that would be the same would be the concept of change. When all of your clocks are gone and no one is left to tap their foot and count the seconds, time disappears with the humans that created the concept. The idea of "things changing" is the only true universal constant.

Furthermore, I do like the idea of there being a *top* layer...an actual existance...but this program does prove, through its accuracy, the existence of a deterministic universe, which means that in some way, shape, or form, the top layer is nothing more than an infinitely iterating pattern or formula or calculation or recursive computer program, etc... That concept, all on its own, is a humbling one, and one which I firmly believe. Who says God has to be a floating figure somewhat resembling a human? Why can God not be a formula, or a mathematical pattern?

On a lighter note, did this story give anyone else the urge to build a supreme quantum computer and start playing the lottery?

2008-11-19 09:57:31 by Tropolist:

B: I was note berating or belittling him. I was just asking him to defend his claim, which shouldn't even be a surprise since he has the onus of proof. As previously stated, his/your arguments don't make any sense beyond the fact that they're poetic. You said that if there were no top or bottom, a loop would "make more sense". This is not solid reasoning. Our universe would "make more sense" if it were perfectly spherical, if general and special relativity didn't contradict each other, and if quantum entanglement didn't exist. But it isn't spherical, they do contradict each other, and it does exist. We have to just deal with it. Yes, a closed circle is very poetic and interesting, but it isn't plausible. If we are going to consider this thought experiment seriously, baseless theories are not welcome. Sorry if you feel that I'm berating you, but I didn't expect the people reading this story would need sugar coating.

2008-11-19 10:16:46 by MarkGreen:

Yeah, but what happens when they get a power outage with no backup generator?

Bweeoooo-FFzzzzt. -NO CARRIER-

2008-11-19 11:41:30 by Tropolist:

They wouldn't know when it happened. The power would cut out in every universe, immediately ending all universes below it. Only the universes above the stability point would survive, and even then they might disappear too.

2008-11-20 01:19:02 by Ellama:

Haha, this made me laugh! Amazing story... altho i had no idea what you were talking about at first.
and rereading this... it made me laugh in a good way, as in amusing... also meant in a good way.

2008-11-21 01:07:55 by Joseph:

I am interested in the consciousness. If they are there, having ideas, talking, thinking and otherwise acting intelligently, they have a consciousness. But all the other people above and below them also have consciousnesses. But they are identical. So do they have different thoughts or the same? If the same, as is implied, what makes them different? If nothing, then they are observing the same thing as the universe above and below them, otherwise they would not observe what they are observing: a stack of universes identical to them in every way. So they all must be identical. This applies to every layer, and implies that there is an identical layer above each. There is no top layer.

2008-11-21 02:05:06 by Nicki:

Somebody's been reading Asimov--this is exactly the kind of thing he'd have written. :) I love the ending.

2008-11-22 18:31:49 by ejl:

I assume everyone realises that in the real universe things are absolutely *not* deterministic, right? (so this story could not possibly happen)
...just clarifying...
This was a great story though, perhaps even approaching Douglas Adams inwonderfulness(!)

2008-11-23 05:33:34 by Ian:

Good read. I agree that it feels like a cousin of Asimov's stories, especially "The Dead Past." (Got here via a link on the xkcd forums.)

2008-11-23 15:00:14 by bunghole:

Sweet...

2008-11-29 12:02:29 by SueDoughNim:

Actually, they COULD turn off the machine, as long as they later turned it back on. Think of it in terms of the Sims. When you save and quit, they don't die or anything like that. And when you reload it, they keep going. To them, time didn't stop for a day or so. They didn't notice anything at all. Also, if it needed to be on, how did the universe exist before it was created?

Different comment: they don't have to be 'nice' to the worlds below them. Essentially what they have is a God machine. They can do anything they want to the 'simulated' world and it happens in theirs, so basically, they can do anything they want to the world.

I, personally, would make it rain purple hippos but give everyone purple armour to survive the torrents. After that, the sun would evaporate some of the hippos, the plants would drink and later transpire others, and some would seep into the bore hippo supplies, to hippo people's lawns and stain their bricks. (If you don't get it, try replacing 'hippo' with 'water', and after that, read it as it is.)

2008-12-01 23:43:16 by DifferentAndrew:

Great story - a brilliant introduction to the Simulation Hypothesis, which says that once you know how to create simulations realistic enough to host intelligent life, you have to accept that it is vanishingly unlikely that you yourself are in the single, top-level universe: the only universe which isn't a simulation inside some higher-level universe.

If you liked this story, you might like to search for Simon Tatham's "Infinity Machine", which explains a bit about how you might program an infinitely powerful computer

2008-12-03 03:05:51 by cello:

Oh god. Wow. ._________________.


Nice captcha.

2008-12-03 15:09:27 by Thomas:

In response to the comments about the universe not existing before the simulation was run, the simulation was run from the beginning of time, or the universe, and then "fast forwarded" to the present day so the universe did exist from the beginning of time.

Also, assuming the are infinitely many universes, with no bottom universe, it is effectively certain that your universe is also infinitely far down the chain, meaning the even if there is a top universe, there are still infinitely many universes above you, so it doesn't matter if there is a top universe or not.

2008-12-03 15:57:28 by Sam:

Actually, no. It doesn't matter how far down the infinite chain you are. There is always a finite number of universes above you.

Let's say you numbered them. The top universe is 1, the universe it simulates is 2. The universe simulated by 2 is 3. And so on.

Now, you may be in universe 1 or 2 or 588 or 456,344,222,100,000,000,000,000,000. That means that you will have 0 or 1 or 587 or 456,344,222,099,999,999,999,999,999 universes above yours. However, there is no "universe infinity". So, there is no universe with an infinite number of universes above it.

2008-12-04 12:58:22 by Thomas:

But if there were infinite universes, the probability that your universe is in the top n universes is equal to n/infinity, thus for any finite number n, there is 0 chance your universe is in the top n, and your universe must be infinitely many universes down the chain and thus there are infinitely many universe above yours.

2008-12-07 10:31:24 by Ian:

Hah. I'm no mathematician or anything, but what you said, Thomas, reminds me a lot of Zeno's paradoxes. Then again, I'm crazy.

2008-12-12 18:04:07 by AntoineC:

Interesting story!

I have been thinking about similar subjects for quite a long time. I have a different vision of the layered multiverse depicted in this story.

In fact, I think there is no top layer, no bottom layer, no circular multiverse but a single universe: our universe!

In order to run a perfect simulation, the quantum computer must store all the particles, quantum states, etc. Assuming a perfect efficiency, the state of the simulation is stored in a sub-atomic array with a granularity down to the fundamental particle (if it ever exists). In other words, to store the universe state you need at least the universe. The same can be said about running the simulation. You need at least the universe to run all the possible interactions.

Conclusion: to perfectly simulate the universe with a perfect efficiency, you need the universe. The simulation of the universe IS the universe.

The perfect quantum computer described in the story is then simply a window open on our own universe. What you see in the simulation is the universe itself. In the story, Tim and Diane are making a wrong assumption about the existence of layers.

Is this what the author had in mind from the start? I have no idea! I any case, it is an excellent and thought provoking story...

PS:
The story is also a good introduction to the ages old question: are we living in a simulated universe? My (current) answer: it does not matter! The universe and its simulation are strictly equivalent.

PPS:
The story describes a perfect computer simulating the universe. Assuming the simulation (that is the software) is the universe itself, we could say that our universe is a gigantic Information System. In other words, the most fundamental building block of the universe is - information! Recently, a few scientists go even further and talk about the Mathematical Universe...


In case you are reading this, many thanks go to you, the author. Your stories make me think again about philosophy and fundamental sciences. And, this is a good thing.

2008-12-12 20:20:20 by Sam:

It is impossible to accurately simulate a finite object on a subset of itself. Since the observable universe is finite, this story is obviously complete fiction. In the story, the simulation of the universe and the real universe are two distinct objects.

2008-12-29 00:11:40 by voxrationis:

There's another interesting point which hasn't been brought up yet. It says the simulation only simulates the part of the universe being looked at through the window, to make it quicker (even though in an infinitely powerful computer this is unnecessary). This means that as soon as they look at anything but themselves the universe above does the same and they cease to be simulated. If, at any stage, they stop focusing on the computer doing the simulation it will effectively turn off all simulations below it. Assuming their universe is an exact simulation of the one above (and so on up to the top) this will make it collapse back to a single universe. Fortunately, since I exist (I can tell because without me this post wouldn't exist), they either fixed it or I'm in the top layer.

2009-01-01 21:01:52 by FlyingSagittarius:

I think that Diane fixed the simulation to simulate everything, so they still exist after they are not seen.

2009-01-04 21:39:54 by ShayGuy:

In probability theory, a probability of zero does not mean that it's impossible to happen, only that the chances of it happening are negligible. Fundamental to the concept of continuous random variables.

2009-01-07 09:53:35 by AnotherSimulatedMan:

Interesting theory you have here, all that I can wonder about here is since they had the abilities to fast forward the simulation and it clearly didn't speed up their perception of their own reality, what would happen if they fast forwarded about 20 minutes to when he was walking out to catch the bus, but in order to test the reality he simply remained inside for a 30 minute duration. Thus negating the truth, but only performing this action after finding out that they themselves were not the original. Any thoughts on my quandary?

2009-01-11 05:37:49 by Tucarius:

Its a very complex and well thought out story.

The whole idea is that at the end of the story you realise the entire story actually takes place within one of the simulated worlds below.
Even though there 'is' a top layer the computed layers can only determin their seperation from the original layer by observing the 'black sphere' that was programmed in, in the top layer the sphere would not have materialised, since it was reality. The reason for this is that up 'until' that point it was a logical simulation of reality and after that point it 'became' a logical simulation of 'what if this HAPPENED TO reality'.

That moment is the divergence point which determins if the universe was the
top universe or one of the potential infinite number of lower universes.

And yes if i was in the Top universe i would also turn off the computer.

Also if the simulation was run forward 'before the programming change' then it would probably correctly 'predict' the events of the future x-minutes from now, however if the program was turned forward 'after' the sphere incident when the layers had diverged from the top layer in realism then you would get a projection of what 'may' happen 'if this universe was a false one' x-minutes from now.

2009-01-11 15:43:53 by Tangrinx:

The "top" universe might not even contain an earth equivalent. It just needs to start a chain that eventually contains earth. In fact the original universe may have drastically altered forces, and dimensions. A test would be to simulate a universe from a different big bang, then search everything in that universe (including further simulations) for their version of earth.

2009-01-13 12:20:57 by Paradoxia:

Since the computer is infinitely powerful, you could run 2 universe simulations at the same time without it slowing down. After they realise that they are in a simulation, the two characters could simply run another simulation on the same computer and have fun with that one. They could probably run as many of these simulations as they want.

In one simulation, you could even put in a program that does to the universe whatever the people in the universe do on their computer, and just leave it to see what happens, instead of trying to perfectly emulate their actions.

2009-01-30 03:37:02 by cookie:

Ok, please feel free to point out obvious flaws here, but could we not compare this to a super complex game of sims? Do we assume sentience on the part of the sim when he answers the phone without your command? No, because the sim is responding to the programming. The sims can even play sims on their sim computers, but does that mean that there is actually a whole new world of sentient sims?
The quantum computer is powerful enough to create an infinite number of simulated universes below the top level, but the simulations are based off of the most likely scenario, which is, in fact, what happens in reality. The program is really "guessing" what is going to happen next and making the simulated universe react correspondingly. Note, guessing here is a bit different than human guessing. The quantum computer would know everything about every possible bit of information in existence, and probably even more, so when it "guesses," it is going to come up with the only possible scenario that fits every fact ever, which is what actually happens.
You could watch the simulation create a simulation an infinite number of times, but in reality, you are still only watching a simulation.

2009-01-31 10:00:17 by Artanis:

This has been bugging me for a while. Say Tim and Diane leave the current sim running, and then run a second one, with the same parameters, and come up to the present time again. You now have two sims of identical, but different, universes. At first thought, they would be in a simulation chain as proved by the first sim, and Prime in the second chain.

Or not? Since it was unlikely for any given universe to be Prime in the recursion, it should be equally unlikely that that universe would be "Prime" in the second recursion. Changes should propagate along _both_ axis, right?

Hard stuff.

Anyway, regarding the power failure issue noted earlier, this is actually kind of easy to solve, and you can solve all sorts of other problems, like someone confiscating your control of the computer and taking over the universe.

It's based in what I like to call a Perfect Black Box, and it's going to make the computer self-contained and intangible.

First, the computer needs power, for all time. Easy as pie: program the simulation to provide energy to the computer in the simulation. In short: The computer creates energy for itself.

To maintain the laws of thermo-dynamics surrounding the computer, since we are locally violating them, create a barrier that prevents the transfer of matter or energy in any form away from or towards the computer at a certain radius. The outside world cannot directly interact with the computer systems, nor can the computer directly interact with the outside world. At certain radius around the computer, there now exists a black-body surface that nothing can penetrate.

Now, you can either leave it somewhere like a 2001-style monolith or black Egg or something, or flip it into the fourth dimension so you don't have to worry about it too much, though you may as well just program it to keep well away from black holes and leave it in 3-space.

Now, as for control, since we can no longer access the computer through a terminal anymore. Easy again, you tag administrators for it, and it recognizes them by genetics (or you track down the soul and use that.) Since the simulation is a database for the computer, it can just watch the administrators for commands, since it has infinite processing power, it can either do this by voice or let you ssh into it (it'd just intercept the commands and pipe the output to the terminal.) You'd be able to do this from any computer, and only you, it would be able to determine the person actually creating the commands. Duress is an issue, but it has access to all your thoughts, so... I'll leave the rest of security to the reader.

Now, the possibilities are endless when you have a simulation like this. As for me, I'd create a set of macros for predefined effects that can be accessed by beings of sufficient privileges through voice or motion or strange, possibly circular, drawings. These macros would, of course, satisfy the definition of magic. The computer+simulation is sufficiently advanced, you see. Two of those macros would be for granting and revoking macro privilege for non-admins. Being the only magic user would be boring. ("Artanis sets mode +o Sam.")

In a similar vein, macro-scale teleportation and FTL travel and communication, time travel (back requires the ability to rewind,) higher dimensional objects, any other tech you can think of, no longer requires a grounding in actual physics. If you want a giant mecha, you don't even need a minovsky particle to run it, the computer can intercept your input and just move all the parts from the sim in universe n-1.

Fun stuff. If anyone manages to build a quantum computer and prove we are a simulation by editing reality, call me up, 'kay? Though, if you need ask for my contact information, do not bother.

2009-02-02 22:31:57 by Daniel:

I haven't read all the comments, so I hope I'm not repeating anything.

First, we don't know exactly how the laws of physics work, and what we do know isn't deterministic. You could probably use such a computer to figure out the laws of physics by having it find the best way to compress a long stream of data so it just predicts the stream, and then reverse engineer the result.

Second, quantum computers don't work that way. They allow you to do NP-complete problems in polynomial time. They're not hypercomputers. (http://en.wikipedia.org/wiki/Hypercomputation)

Third, hypercomputers don't work that way. Having a computer like that is mathematically impossible. There'd be nothing to keep you from writing a program that predicts its own output and outputs the opposite.

I'm sure Sam already knows the first one, he most likely knows the second, and there's a good chance he knows the third. Also, it's a good story.

2009-03-05 20:43:37 by Azrael:

OK, slight nitpicking here ( sorry sam! ) it's not that I'm not interested in the other facets, just I'm that sort of person....

"You knew for a fact: this thing could execute an infinite loop in less than ten seconds."

I'm not entirely convinced about that even infinite CPU power would allow you to complete an infinite loop in finite time, E.G.

while (1) {
Do (something)
}

As I see it there is NO circumstance under which that will ever terminate, no matter how much CPU power you throw at it. Infinite CPU power might allow you to run around the loop an infinite number of times in finite time, but would it really cause the loop to halt?

2009-03-05 20:51:02 by Azrael:

Umm, actually I guess this is related to the halting problem also mentioned. The way you solve the halting problem is simply to give the algorithm and input to the infinitely powerful computer and run it. If the algorithm halts, it halts instantly ( and gives you the output ) since an infinite computer can carry out any finite number of operations in zero time. If the computer doesn't instantly return an answer, it never will and the algorithm is non-halting.

2009-04-13 04:23:39 by tedweird:

Eh, Azrael, I think the point was that the computer is so powerful can do things that last forever in not only a certain amount of time, but a short amount. 'Tis called hyperbole, my friend.

2009-04-16 04:27:21 by Eric:

My head hurts.

2009-04-22 23:13:55 by Azrael:

tedweird, I'm open to correction here by someone with a better understanding of the mathematics on infinity than I have, but it seems to me that there would be no "short time"s when you have infinite computing power available ( assuming you don't introduce any deliberate waits ) tasks should either be completed instantly, or never.

2009-05-14 06:28:48 by Dot:

As far as I can tell, I don't think anyone has brought up the point that the simulation(s) may not be exactly "lined up" with the universe that the story is being told in. Note that Diane is the one stopping the fast-forwarding, not the program. This implies that there's an infinitesimal chance of exact synchronization. Would that mean that the simulations don't truly depict "reality" and its events? Or would that just mean that it does, but it doesn't need to be exactly simultaneous to happen?

That actually made me think of another thing... if they fast-forwarded it from that point, would they effectively be predicting the future? If they did something like collapsing the sun into a black hole a billion years into the simulation's future, would that mean that the sun would spontaneously do the same thing a billion years into their future for no apparent reason?

Also, a little comment about the infinite processing power: In theory, what would making a program with an infinite loop that incremented a variable do, considering both infinite speed and storage?

2009-06-13 00:15:41 by Andrew:

Oh gee, the world's first idiot-repellent, yet completely machine-solvable captcha... Doesn't this go against the very use of captchas?

2009-06-13 00:24:20 by Andrew:

AnotherSimulatedMan, if they were to look into the future, they would've seen themselves doing whatever they would be doing, so in the case they look into the future and see themselves on the bus home, they *will* take the bus home. If they don't take the bus home, then when they look into the future, that's what they would see.

2009-06-21 21:19:43 by Margolis:

The square root of infinity is infinity. Infinity by definition is a paradox, and attempting to quantify infinity, or an infinite loop of commands is pointless, and is in turn propagating an infinite loop of infinitely pointless quantification. If anything is infinite, then everything is infinite. Try to understand infinity. Infinity. Everything is nothing.

Where's NEO?

2009-06-25 21:15:41 by zico:

Okay, forgive me if I write some stupid things now. I am by far no expert in mathematics, physics or even quantum physics.
But I REALLY love such stories and theories. And even more I like to think about it.

For my logical way of thinking, the story left many questions unanswered to be really sure what is actually happening there or what would be the consequences of changing or turning off the computer. Of course these are questions we cannot answer - the story is still SciFi and a very good one.

So my theories are as follow:
1) There is neither a top- or bottom-level. These universes would differ as the top-level would not get any changes and the bottom level would not have a working Universe-Simulation - or a different one. But if ALL are the same, there must not be a top- or bottom-level universe. So I am in for the big-loop idea.
2) As a consequence from (1) - maybe only in a pragmatic way - there are no real multiple universes but merely just a direct way of changing our very own reality.
3) If you would turn OFF this computer, I think it would not *end* the universes, but only cut the interaction between them or - if following the theory from (2) - end the interaction with our own reality.

I am sure many think different than me and that is okay for me. I liked to read all your opinions and I am also excited to read your opinions about my theories (and how you maybe how you think they are wrong).

After all reading this story was a VERY exciting experience. Always showing what we can do with fantasy and logic - going beyond our very own borders which is our mind.

P.S. Sorry for my English talking - it's not my native language.

2009-07-03 05:30:13 by J:

i honestly do not know what to believe
i sit awake at night, and just think
and this situation is a common thought

2009-07-11 01:03:58 by toxicafunk:

Maybe there's just two mirrors! The "real" and the simulated one. The simulation of the simulated one would in fact be the "real" universe. just a thought.

2009-07-11 05:49:23 by Pelger:

hello, after reading the story and the comments, I still don't get
the jump "upwards", I mean this:
I understand why there are infinite universes down there,
and I understand that if that is the case, it's more likely to
be in the middle of the chain than at the top.

but why does what you do to the "lower level" universes "jumps up"
to the "top level" ones?.
the "relationship" between the universes seems to be a one-way-only
thing, that points "downwards", but I can't see a reason why it
must affect the above universes so yours reflect the interference too.

if you are in universe n+1,you can affect n+2 and so on,
but you are "isolated" and can't do anything to universe n.

I hope I was clear enough, sorry for the bad english, bye!.

2009-07-14 20:05:40 by JR:

So for Pelger, in this story there's no upward impact. That girl doesn't force the n-1th universe to do anything. She does something, and as she is a perfect simulation of the n-1th universe, she's doing the same as her n-1th equivalent does.

I see many people here talk about determinsim, let's say plainly that determinism is an outdated idea, for two reasons.
Fisrt, at the quantum level, lots and lots of things hapen with a probability, in a non-deterministic was.
Second, the chaos theory shows that in a complex system (as our universe is), even if everything were deterministic, a very small difference in one condition can lead to very big differences in outputs.
Putting the two together, we have a permanent source of totally random, if small, variations, and a mechanism that turns small variations into world-changing perturbations; that's why a perfect simulation of our universe is not possible. (Of course the author knows this, but i wonder if some readers could get caught by the "why not?" effect)

Anyway, great sci-fi story, just rember that's sci-fi, not science ;-)

2009-07-16 21:51:53 by Pelger:

ok, thanks JR, I think now I get it, the n universe is not the cause
but the consequence, since she's simulated, she'll do whatever the n-1 version does. ok, but, doesn't that imply that only the original
top universe has free will?.

2009-07-29 22:41:02 by Steve:

Aside from the computer being fundamentally impossible (Even from a what-if standpoint, it would require a computational speed greater than it's own computational speed, but Sam knows that!) it's a great short story!

Also, fast forwarding the computer would be perfectly practical if it somehow *could* exist. Say you fast forward at 2x speed... the point at which you are looking 1 second into the next universe's future, half that time would have elapsed in the universe above yours watching you, and twice that in the universe the one you are watching is watching.

All said and done, if someone showed me a computer that somehow exhibited these exact tendencies, I would still be more inclined to believe I am being deceived (perhaps by my own senses even) than to believe I am being simulated! After all, how can you simulate such a thing as OBSERVATION? Observation trumps sensory input on the certainty scale.

2009-07-31 11:16:53 by Roier:

Check this, you are all assuming that there is a TOP reality, BUT, imagine that you set the camera at the same position that the screen.

So the screen would be more like a mirror rather than video feedback, then, WHO IS UP and who is down?

Conclusion: The loop is infinite. There is no TOP universe, and if you turn the computer off, every alternative universe will still exist, you are just closing the gate(window), Looking glass.

You could also use this kind of computer to travel in time or recreate any fantasy you can imagine, it would only be ANOTHER DIMENSION.

Perhaps LOTR was real as that, and Bilbo and Frodo was brought here in reward, where they changed their names to J. R. R. Tolkien and Christopher Tolkien. Just perhaps.

2009-07-31 12:19:56 by Sam:

I don't know what you people are talking about. There is a top reality. Because I say so. It's my story.

2009-07-31 12:29:10 by Roier:

Nice to meet you, perhaps you also wrote about you writting the story on another universe.

I so much loved your story, I su much wanted to writte something like that, let me know of any more writtings from you :) ty for your quick response

Also, I wanted to add, I am no expert in quantum physics or computer, but this is how I explain how a quantum computer would handle an infinite loop.

A digital computer running a loop would be like a person making circles with his index finger around the edges of a drinking cup, you would never end doing circles unless something stops you.

BUT, a quantum computer would be like the same person putting his palm of the hand covering and feeling the whole edge of the same drinking glass.

2009-07-31 12:36:14 by Sam:

This story actually has very little to do with real quantum computers; quantum computers are not infinitely powerful.

2009-07-31 12:50:00 by Roier:

What do you think of OPTICAL COMPUTERS? in which transistors would be made not of matter but of pure light?

Something like this http://physics.aps.org/articles/v1/23

The size of transistors would only depend of the smallest frequency you can achieve, AND a quantum-optical hybrid would still have more benefits. If not really infinite, perhaps you would at least be able to properly simulate our visible universe, wave by wave, particle by particle.

At least my poetic idea, that would explain a lot for me.

Hope that you like my ideas, and perhaps might complement yours :)

2009-07-31 20:30:29 by X:

If I were them, I'd just switch it off. I don't think they'd cease to exist. After all, did they exist before the computer was switched on? If they did, it means that they will exist after switching it off. Do you need to be fully simulated to exist, or is it just to be "defined"? What if you saved all the information in the universe to a DVD, would the universe be "inside" it?

Also, there's a paradox: if they tune the "camera" to see the screen, they will see a infinite tunnel of screens, until there is not enough resolution. But what if they make the screen show _exactly_ what the screen below is showing? As information would not be lost between universes, they should see the _last_ universe, but there is no last universe. Most likely a black screen. Or a white. Or yellow. Or a picture of your mom. Anything would be possible, it's like saying x=x.

2009-08-01 11:52:10 by Roier:

I am so glad, that I inspired you =) SAM

Remember, THIS IS FOR THE FUTURE.

2009-08-01 11:53:22 by Roier:

Smile!!!!!!!!!!!!!!!!

=)

2009-08-07 09:07:15 by MatthewFord:

If you like that, you might also like: http://www.podtycoon.com/shutdown. A letter from the simulator of our universe informing us that we are being shut down soon, now that he has finished his college paper.

2009-08-12 13:37:59 by AndrewZar:

Sam, just making sure you got my email. Love this story and want to work with you on a comic of it!

Andrew

2009-09-20 18:39:37 by TommyT:

Love it.

2009-10-02 00:13:36 by Kyron:

But choice isnt about whether or not it's predetermined, it's about thinking and coming to a conclusion from options, which just means that the other, unchosen, option(s) exist only as concept; and not in material form.
Which could be predetermined supposedly.

2009-10-07 14:35:13 by Name:

Neo is here. rsrsrs
Numbers are an invention of humankind.
Mathematics is an invention of humankind.
The infinite concept is an invention of humankind.
Why do you think that only because numbers are infinite, the concrete concept of infinite can exist? Mathematics and science have flaws; they are a creation of imperfect minds.

2009-10-21 20:03:00 by HedgeLender:

This reminds me of some of Asimov's short stories

2009-10-23 04:14:36 by SteveBaker:

So what happens if they step the simulation an hour into the future? Now they can see into the future. If the same thing happens in all of the universes then you know what the winning lottery numbers will be.

2009-10-25 05:50:14 by Someone:

I wish i had this computer & program.

2009-10-26 15:03:42 by TimMcCormack:

Various thoughts:
* If *I* were running this program, I'd put a chain letter into the sim. "Welcome from layer #0. Increment that number and pass it on!" (And if a note suddenly appeared next to me, I'd have to edit my change.) Now, wouldn't that cause a change to propagate down the ladder?
* I will also note that a fixed point is only one of many interesting feedback loops. For instance, I could also place a fake computer terminal in the sim which allows the simulated individuals to send data back up to me, and I could send data back down (by altering their universe.)
* You could also (theoretically) have cycles of length greater than 1, wherein each layer has some feature that is different from the one above and below, but the same as the next one in line after each of those.
* Or, you could alter the simulator machine at layer+1 to allow the simulation of layer+2 to *affect* layer+1. (This is sort of like indirect self-recursion using a fixed-point combinator.)
* Why not have an infinite number of savepoints? You could then "roll back" the simulation to any previous point. Fun twist: Allow the inhabitants time travel this way.
* Self-preservation is not a good reason to treat the next layer nicely, unless you are worried about back-propagation. ("Man, look at how layer+1 treats the one below it! Let's fuck them up.") If you're at a fixed-point, you're going to make a choice one way or another anyway.

2009-10-27 04:04:48 by me:

just wanted to check if i can solve your captcha right...

2009-11-17 14:41:59 by test:

I also want to test your captcha

2009-11-17 14:47:11 by test:

That was ridiculously easy, maybe you should change it to something harder to type like π.

2009-11-17 14:48:04 by test:

That was ridiculously easy, maybe you should change it to something harder to type like π.

2009-11-23 05:45:19 by CG:

This is the basis of the infinite multiple universes hypothesis. If Tim or Diane switched off the computer, they would BECOME the top layer. I heard a good analogy recently (I think i got this right!): If there are infinite multiple universes, and you wanted to prove that to yourself, you could use a gun pointed at your head while people watched you pull the trigger. The gun has a small chance of jamming and not firing. In one of the universes the watchers would see you die as you pulled the trigger. However in another parallel universe, the gun would jam and you'd survive. If there was really infinite parallel universes, then each successive pull of the trigger would result in a jam, attempt after attempt and you'd be immortal!

2009-11-23 06:16:35 by CG:

I've refined the analogy:
If there are infinite multiple universes, and you wanted to have a go at proving that to yourself, you could use a gun pointed at your head while people watched you pull the trigger. The gun has a small chance of jamming and not firing. In one of the universes the watchers would see you die as you pulled the trigger. However in another (almost) parallel universe, the gun would jam and you'd survive. If there was really an infinite amount of universes, then somewhere, in at least one other of those universes, the next trigger pull would result in a jam, and you'd therefore continually survive, attempt after attempt!! (Although you would have to hope you were in the (almost) parallel universe where every successive trigger pull resulted in a jam, I think!)

2009-11-23 07:16:25 by CG:

I've been looking round the internet (again!) and the analogy I just drew I heard recently from a chap called Marcus Chown. Coincidentally(?) he is also very outspoken on the subject of... quantum computers! And he recently appeared on BBC Four (TV here in UK) on a program called 'It's Only A Theory' which is where I heard the multiple universes analogy of his. If the 'I don't know, Timmy, being God is a big responsibility' story tickles your fancy, then Google Marcus Chown. Or perhaps Nick Bostrom too.

2009-12-07 04:53:17 by Sleepless:

This story strongly reminded me of the scene in the movie Spaceballs where they fast forward the movie forwards to "right now". "everything that's happening now, is happening now"
"well what happened to then?"
"we passed it"
"when?"
"just now"

2009-12-08 02:39:47 by ePerson:

I just realized that in some ways, this story could parallel an idea in "least significant bits" from fine structure:

"You ever do any set theory?" asks Ching. He flips over a beermat and writes out some numbers, "0, 1, 2, 3, 4, 5, 6...". He circles the "3" and labels it, "you are here". At the end of the sequence, after the ellipsis, he writes a lower-case Greek omega, "ω". He circles it and labels it "God". He waits a moment for Murphy to understand, then he adds "ω+1..." after God. Then he circles this, and labels it "?".

"I don't know what this is supposed to mean but I don't find it a compelling argument," says Murphy.

"If the Structure is anything like ordinal arithmetic," says Ching, "there's always a bigger God. There is nobody at the top. There is no top. There's only one end to the sequence, the bottom end, where we are."

If there are an infinite amount of simulated universes, then there will be infinitely many people above our simulation believing that they are on top and that they are running everything. In addition, there may be other breakoff universe simulations, including ours, creating even more convergence points. The laws in this universe may in no way whatsoever resemble those in universes a billion layers up. The people who believe that they are on top will have no way of using Diane's test to see of they are not on top, so they will go on believing that that is where they are. And there will still be a bigger god above them.

2009-12-09 06:03:19 by Belinda:

Great story, I loved it, and I loved reading the comments, too.

2009-12-17 19:24:38 by Jonathan:

Interesting read. I'm not going to talk about it though, because whatever I say sounds pretty stupid. All I would really communicate is the aspect I'm looking into, because anything that follows is obvious. Instead, I'll ramble about the captcha. Of course, there's this canonical answer, but it isn't quite as definite as you could hope it would be. You can negate it and still have a valid answer. Add another level of imaginary (quaternions), and there are infinite answers. It's not even because minus one is this supposedly special singularity; you can take the square root of any number "the other way". Ugh, more levels. Anyone have a super soaker?

2010-01-16 03:23:37 by Boter:

If I ever adapt another one of your stories to film, Sam, it's probably going to be this one. It'd be so easy to do.

2010-01-16 15:00:23 by TimMcCormack:

Wait, one of your stories has been adapted to film? Which one?

2010-01-16 23:33:26 by Boter:

I did "Time Loop" with a crappy still camera with MPEG recording. Sam links to it on Google Video from the Time Loop page. I'm actually rather ashamed of its quality, I've meant to re-film that one with my DV camera for a while now. Actually one of my planned February side projects to break up the larger project I'm working on.

2010-01-23 00:45:03 by Drake:

Might want to remove the halting problem as an issue this system resolves. The halting problem wouldn't be "solved" by an infinitely powerful computer for the same reason that such a computer wouldn't be able to determine the truth-value of "This is a false statement." It's an issue of decidability, not of complexity.

2010-01-23 03:26:45 by TimMcCormack:

Drake: You give me a program, I run it at infinite speed. After a finite period of time, I check if it is still running. If so, it does not halt on that input. Voilà!

2010-01-24 09:59:07 by KJ:

Might each universe running a simulation of itsself eventually run a simulation of a slightly different universe at the same time? In that case, would the chance of the circle appearing be lower?

2010-01-30 16:51:24 by Drake:

TimMcCormack: Infinite speed is incoherent in that context. Is it executing all instructions at once? Further, the halting problem is undecidable because there is no way to reach the final state. It's not an issue of time or resources.

2010-01-30 18:55:25 by Sam:

The Halting Problem states quite specifically that there is no algorithm which can make this decision. However, the terms of the Halting Problem state that the algorithm must be implementable as a finite-state Turing Machine. It must be a computable function.

The computer described in this story is not a Turing Machine, but vastly more powerful. It does not make sense to look at the outcome of a Turing Machine after running it for an infinite period of time, but it does make sense to look at the infinitely many states of a Turing Machine after it has run for all possible finite periods of time. In other words, we can run any Turing Machine for an infinite period of time, and say, "set flag HALTED to 'true' if the machine halts at any individual step". If, after infinitely many steps, the flag is still at "false", then the machine did not halt at any finite step, which means it did not halt at all. Calculating these infinitely many steps is a "supertask", is impossible, and requires a machine more powerful than a Turing Machine, but it is not outside the bounds of conception.

In other words this machine is an Oracle: http://en.wikipedia.org/wiki/Oracle_machine or hypercomputer: http://en.wikipedia.org/wiki/Hypercomputer

It should be noted that while it is infinitely powerful and therefore capable of modelling algorithms more powerful than Turing Machines can model, it is still itself vulnerable to a "Super Halting Problem", in that it cannot decide - even using an infinite super-Turing algorithm - whether a given super-Turing Oracle program halts or not.

2010-02-05 19:42:10 by Maddie:

Clearly, everyone on the lower levels should run the program to the end (or at least as far in the future as possible) before the top level turns their computer off, so that their universe doesn't risk being wiped out early. Of course, the level right below the top might be screwed, but they might not know that and sacrifices must be made.

2010-02-13 00:34:33 by Boter:

My thoughts too. Run the thing all the way through.

The problem is, people would run another instance of it afterwards to play with... and you can't be sure that *that* instance isn't "yours" - that you're not part of that second program. Or, in fact, when everyone gets one in thier home..;. so not only do you have a stack that goes infinitely deep, but you have a massively increasing number of stacks each time it hits one. Even if ten people in this universe run the program... man, that's a lot of simulated universes.

Of course, that lets someone go back in time, part the Red Sea, create an explosion for no reason over Siberia, make a star very bright so that roughly 2,000 years ago it guides some dudes... the possibilities are, literally, endless.

2010-02-13 22:26:20 by Nowsnotnow:

Time to thoroughly confuse whoever is not thoroughly confused as of yet. As it was that time only a moment ago. Ah, time...

Anyway, two concepts for consideration:

1. Now that we have simulated the universe and have an infinitely powerful computer, let us find the probability that we are in the top layer in this chain (regardless of how infinitesimal that probability may be). Now, let us alter the universe below ours and make that probability = 1. This would inevitably create a paradox the moment that we glance at the simulation. Assuming paradoxes cannot exist, then the only way for reality to rectify itself would be for the simulation to end...and for our own universe to end up being the top universe after all.

2. Let us be safe and avoid damaging our own "universe" with our curiosity. Instead of ending the simulation, let us only run a simulation of what would happen if we ended the simulation. In fact, we should be even more cautious. We should run a simulation of what would happen if we ran a simulation of what would happen if we ended the simulation. In fact, we should increase this chain as much as possible, for with every link in the chain the probability that our own universe will end will be cut in half!

Yes, I am certifiably insane. I just lost the certificate...

2010-02-20 07:33:28 by Jess:

I just finished reading the Ring series by Koji Suzuki (you know, girl in a well, scary videotape, people die in 7 days). Not to release a big old spoiler, but the last book in this series is pretty much this piece of fiction (with minor differences). I was really really surprised to stumbleupon this right after finishing the book. I like the twist you put on it.

2010-02-22 10:58:42 by Dmytry:

Great story.

Regarding stability, though... even very simple iterated nonlinear systems (such as logistic map for example) are often 'chaotic', with positive Lyapunov's exponent, meaning them *never* reach stability, there is no stable points within the domain at all.

This doesn't matter to the story though, I believe. Each universe has same rules and is simulating top universe, from big bang (suppose big bang starting state is simple and known ahead), right? That alone is enough for universes to be identical, the laws are identical, the starting states are identical before they start meddling with simulated universes. The divergence propagates 1 step down on each interference. Suppose they would start doing nice things to themselves, e.g. making bricks of gold to keep the experiment going, and when universe doesn't receive a brick of gold, it quits making those for universes below. First brick of gold appears in all but the top universe (n>0). Second in all universes with n>1 , third n>2 and so on and so forth. If you made 1000 gold bricks, chance that this divergence reaches you is infinitesimal, literally 1000 to infinity.

It gets interesting tho if you meddle with simulated universe in the 'past'. Suppose you0 (top you) create a sphere in simulator 10 seconds before your now, this distracts you1 inside simulator, and he does not create a sphere, and does not distract you2, and so on and so forth, each even universe has sphere and each odd doesn't, then even and odd can diverge from eachother, while all even and all odd would stay identical. This gives a nice and entirely consistent time travel model.

Other interesting thought: suppose 2 groups of researchers do this. Now you get a tree.

2010-03-03 08:09:34 by misanthropope:

heisenberg's uncertainty principle. propagation of error. exactly 100% of the Dianes are almost exactly 100% wrong.

2010-03-06 19:40:00 by me:

The uncertainty principle works for shure simply because you cant figure out the state of any particle exactly, if you could you might be able to simulate exactly. (however, you might still change the results just by collapsing waveforms.

Infinite universes=no top or bottom. their are infinite.

2010-04-09 01:39:41 by kukulkan:

Its childish to label anything long lasting/immensely huge (beyond actual human comprehention)"infinite". Preschool physics says The whole universe revolves on time space and matter.. Everything that came to existence will have an end.. If it has no end, certainly it had no begining. Unless u believe in eternity. Is eternity possible with the absense of time? If it is, what goes on in a timeless universe? Does nothing exists?.. Can space exist without matter? I'm mentally divergent..but I will try to refute some stuff..do u believe in fate Neo?

2010-04-09 03:02:59 by kukulklan:

Instead of infinite, I would say systematicly endlessly yet simultaneous linear/not linear circular sequent. Just like numerology, how much sense would mathematics make without a 0, we only have 9 numbers. But with 0 we can count to infinity. But there's nothing infinite about it.. Its the same numbers 10 numbers,just in different position(circular sequence) thus giving it a different value.. An oracle doesn't exist and even if it did could it coexist with a clone at the same time and space as the original? But for cloning we would have to know the methaphisics of the oracle..the oracle will no longer exist once u figure that out..then again that would refute the fallacy of infinity.

2010-04-21 15:08:30 by aido:

surely before the simulation is created it can be considered as *off*. so turning it off would not cease the existence of the other reality, it would simply stop the layer above you from being aware of your existence. you would continue to exist on your layer.

2010-04-21 22:36:24 by god:

if the top most level made one of these it would produce an unending number of them up and down stream

2010-05-10 04:15:15 by Ray:

wow, im new to this web page thingamagig whatever it is, and just from reading this one story, i am absolutely stunned and amazed. who writes this? and do you have any books published or anything. i would really like to know as i am very interested.

2010-05-19 17:24:49 by Edd:

What really freaks me out is this - if I were in a universe without some of the restrictions in our universe (if, for instance, I lived in a universe without the uncertainty principle), I would be sure to add such restrictions to any simulation I made (and, continuing the example of the uncertainty principle, simply assign random values for things that are purely probabalistic to continue the simulation with), to ensure I didn't make a loop.

Kinda worrying that the universe follows similar rules to what we would expect if it were a particularly well-thought out simulation :S


Also, FANTASTIC story!

2010-05-29 08:39:19 by Critterfluffy:

Edd,
Who says the Uncertainty Principle is absolute? Nothing in science is. Science may have made a mistake. I do believe in the Uncertainty Principle but it is possible that we mistook some data and concluded the Uncertainty Principle in error. Just a thought.

2010-05-29 08:40:06 by Critterfluffy:

I first read this story over a year ago and I have to say it has made me think. In over a year I have come to 2 conclusions I would like to thow out for comment.

NOTE: I have read many of the above comments but I have not finished them all so others may have said identical or similar things.

1. Assuming there is a top universe, the moment the first change is made and they do not see the change they become different from those below. To them the computer is not a god machine, just a very accurate predictive model. At that point, most scientists would do things that you wouldn't if you knew it affected you. Since they see it as a very accurate model of our universe and nothing more, they would use it to model weather events, earthquakes, nuclear explosions, etc to see how our world would handle it and what could be done to lessen the impact. They wouldn't see anything wrong with it, after all, it is just a computer program. Anybody who has played The Sims knows this. This would diverge universe #2 from those below it as they did see changes occur in their own universe when they made them and would not use it in such a way. Plus, now they are likely destroyed. This means universe #3 is safe unless universe #2's computer is destroyed. I am going to assume, for the sake of an interesting arguement, that a quantum computer is self sustaining by design and cannot be destroyed. I assume this since a real universe is self sustaining and cannot be destroyed. Now universe #3 will no longer see changes in their universe when they make them to #4. They may or may not continue the behavior of universe #1 but we will assume that the pattern of destroying even numbered universes will eventually stop. Now a long chain will begin to occur. Since universe #3 no longer see's local changes they will diverge from #4+ and stop making changes to #4 or ones that don't sync to #4. Now #4 diverges, then #5, etc. We can only conclude that even if a Quantum computer was able to affect ones own universe it has limited uses. Also, being #2 is kind of shitty(pun intended).

2. Taken the conclusion from above, how do we prevent the computer from becoming useless? The best way to protect ones own universe in such a situation is to create a program in your quantum computer(QC1) that causes the quantum computer from the universe below you(QC2) to only affect its own universe. This will in turn alter QC1 to do the same thing(unless you are the current topmost). This will allow free editing of ones own universe without worrying about a day that it will stop working and protect you from the wrath of the universe above you.

Sorry for the long post. I would love to hear any ideas of other concequences or problems with my conclusions.

PS: Remember, we are assuming there IS a topmost so lets try to avoid arguing against that assumtion. If there was a loop I understand these conclusions are no longer valid. But lets think for a second. These are QUANTUM computers. In Quantum physics there are no negatives. There are only real numbers of zero and higher. There has to be a finite lowest or beginning. This, I think, would exclude the possibility of not having a topmost universe. Besides, the story assumes that the quantum universes are identical. If there is no way for the universe to diverge then there is really only one universe anyway as they are all the same and therefore just one unit.

2010-05-29 12:00:47 by Critterfluffy:

I have now finished reading all of the comments(night shift is great). I love some of the ideas that have been thrown out and I would like to comment on some of them.

DISCLAIMER: I take no responsibility for head explosions or holy wars that result from the next few paragraphs. By reading you waive me of all liability.

1. You cannot turn off the computer until you run it to the end.
-The thing that is being missed is the computer already DID calculate the universe until the end. According to the article, the computer can calculate an infinite loop to the end in 10 seconds. Whether this makes sense or not we have to accept it as this is known to be truth in the story universe. With this information, we can conclude that given 10 seconds the entire history of the universe(past, present, and future) will be calculated. As for those in the universe, they obviously perceive more than 10 seconds. They saw years pass. This is because of their frame of reference. At scale, 1/10th of the life of the universe is one second in the time of the universe above them. In fact, the entire story is actually taking place within the calculation phase of the simulation, when the computer is figuring things out and before it can be viewed. Turning off the computer doesn't undo the fact that the calculation phase took place. Think of it this way, what they are watching on the screen is just a recording until they make a change. The change just starts the calculation phase all over again. Given another 10 seconds the virtual universe has played out its life once more.

2. There is a universe loop so their is no Top Universe.
-I like this concept, in theory, as it shows that nothing is real and that our lives are just a small part of something much greater. The problem(outside of what was mentioned in my previous comment) is the scale doesn't work easily.
Lets take the thought from above: All experiences and choices take place during the calculation phase that takes approx. 10 seconds(in each universes time).
Now lets start with universe 1. In 10 seconds, everything is calculated down to universe infinity. If at universe infinity their was a point that universe 1(the original) had to be calculated then at some point universe 2 through infinity would have to calculated over and over again. This is hard to accept. Taking the fact that this is an infinitely powerful computer, it is conceivable that this could happen. I concede this fact but I also quote Occam's Razor "The simplest solution is usually the correct one" and this solution is by no means the simplest.

3. What if you set the computer to show you 10 seconds into the future and you determined that you would not do what you saw.
-I, like many people, am a big believer of free will. I don't think there is a plan that cannot be altered as that denies purpose for life and, theologically, upsets the concept of heaven and hell. How can a god plan the whole universe and then condemn you to hell for sinning when you had no choice. As for the story, free will would rear its head.

**I will be using the notation of U1 for the man getting on the bus. The observer is U2 and all lower you's will increment the number**

So U2 watches the monitor that is now showing 10 seconds in the future. U1 gets up, says bye to Diane and gets on the bus. U2 says "I am getting a pepsi instead". U3 sees U2 watching U1. U3 decides he is getting a Coke. U4 sees U3 watching U2 watching U1. He decides to kiss Diane. It gets infinitely complicated from here. What you see occuring though is every universe immediately diverges. By performing such an experiment the result is now the predictive model has been broken and every universe is now unique as this is the only way to prevent denying free will to each Tim. Even if you take the universes to be fully deterministic this would be caused because each screen, as you get further down the line is showing something different. Different input results in different output.

4. It is impossible to create a perfect sim of the universe since you cannot know everything about the state of the universe.
-I accept this but I am going to throw out a different concept of the universal stack. This is mostly philisophical though.
Lets say the topmost universe is nothing like the simulated ones. Lets, say in a universe quite unlike our own, some entity built the infinitely powerful computer that started it all. He/she/it(from now on referred to as GOD) didn't want to simulate their own universe as GOD thought that would be boring so GOD set about programming in the big bang. GOD let the computer go about its business and 10 seconds later had our universe. After exploring, GOD discovered us. GOD eventually came to the idea of creating a program that would try different iteration of the start conditions until it lead to the construction of GOD's computer in our universe and the programming of the exact initial conditions GOD used to create our universe. 10 seconds later, the machine brute forced a universe that had its own copy of the computer and program. Now you have Diane and Timmy building their computer and miraculously programming an exact copy of their universe into it.

That was fun!! :P

Again sorry for the Epic posts but I just can't see anyway to express this without going into detail.

PS: I wonder if this could be run as a Pen and Paper RPG. Maybe using the Paranoia system. Just a thought.

2010-06-01 14:27:29 by mustyoshi:

Wow, that was an amazing story.

2010-06-01 21:49:51 by jonas:

I found that overlooked paper of Diane on the internet: http://www.simulation-argument.com/simulation.html

2010-06-07 15:06:58 by Rik:

Y'all need to go away and read Permutation City by Greg Egan. This concept is just a subset of the shit going on in that book.

2010-06-07 15:22:03 by Rik:

Basically there is no problem. Whatever the setup, the conclusion is, whatever they do to the simulation of the universe directly affects their own universe. However, the terms of the creation of their universe mean that it is capable of supporting infinite computation. Therefore it is capable of supporting itself without an external source of computation. There is no reason for it to switch off just because an infinite computer contained within it is switched off. Thought experiment: Say they built two computers, ran the exact same program on each and switched one of them off but not the other. Paradox, neh? This also means that affecting the program would not affect the "real" world as that also leads to the previous contradiction. Those circumstances fall outside events of the story but that just shows that the story is built on a logical flaw. Sorry.

2010-06-07 16:07:05 by lcordier:

What if the universe sequence is not recursive but instead circular.
Meaning minus infinity wraps around to meet positive infinity in the limits?
Then there is no top universe, just propagation delays for effect to propagate
to all the other universes. Ultimate KARMA ;)

2010-06-07 17:16:30 by Sam:

I may just force comments through a manual approval step.

2010-08-03 19:06:32 by Xetwnk:

The solution to the "we don't dare turn it off" problem is simple: hit the "fast forward" button and run the simulation through to the heat death of the simulated universe. The beings in the simulated universe will continue to experience their lives at the same subjective rate they always have, even though TO US it seems to elapse uncountable quintillions of times faster. By the already postulated identity of universes, OUR universe will continue to exist, subjectively, for those many quintillions of years before being shut down for the (next universe up's) weekend.


2010-08-07 10:31:42 by Allan:

This is more believeble than bible!

2010-08-10 20:18:27 by MarkoJames:

Try to understand infinity ?

Easy again, you tag administrators for it, and I am sure many think different than me and that they are now the top level because they've existed prior to the guys in the simulation.

Even if ten people in the timeline of the comments(night shift is great).
This means that everything is infinite.
BTW when it says it simulate only a part it can't be sure to add such restrictions to any previous point.

Now, you can achieve, AND a quantum-optical hybrid would still be in universe n-1.
After all, did they exist before the simulation an hour into the fourth dimension so you don't argue how that is okay for me.

2010-08-20 06:35:11 by Joey:

I think it's funny that the comments are longer than the story.
When you have a story like that, especially when the story isn't very short (as is the case of this one), you have one hell of a story.

2010-09-26 19:59:52 by Nathan:

Perhaps the person who wrote this story will someday build a computer with infinite processing power and create a universe in it in which an identical version of himself writes this article later to build a computer with infinite processing power and so on...

2010-10-05 01:02:22 by VladimirIvashko:

To continue with the Egan reference: If the top layer turns off the simulation, then due to quantum immortality, everyone below them would get shunted sideways, with each layer becoming a new top layer. Thus, the act of turning off the simulation would create a multiverse composed of a series of parallel universes (although Diane and Tim wouldn't necessarily know about it).

2010-10-21 00:48:29 by Butts:

Since every layer (except the top) is exactly the same, whatever you decide to do to the layer below yours would happen to your own, assuming you aren't the second layer. What I'd personally do is figure out if I'm in the second layer and, if not, do anything I wanted. I'd be God of infinite minus 2 layers of existence.

2010-10-21 00:54:53 by Butts:

Wait, no. Eventually all the layers would diverge from each other. If the top layer decides to leave the simulation run or skip ahead, their timeline is different than layers 2 through infinity. This would lead to layer 2 becoming its own top layer, as anything they do is completely independent of layer 1. They would try to experiment by causing something to happen in layer 3, and as soon as it doesn't happen in layer 2 (because layer 1 didn't do it), they'd split paths from layer 3 and you'd have timeline 1, timeline 2, and timeline 3 through infinity. This would continue forever.

So, if you were way down the line, you could get a little bit of time before becoming your own top layer. So you'd have a small time as God, and that's it. Better make your changes count.

2010-10-22 00:52:23 by Nick:

The inherent uncentanty in quantum physics make the story false. You can't simulate quantum randomness - it's random.

2010-10-31 00:04:43 by Ovais:

No matter what the base computation realities are, this is an amazing piece of work. Are these computations to make the computer actually real or possible? Just asking....with a researcher's enthusiasm =D

2010-11-10 19:31:50 by AW:

Reading IS fundamental!
I would hope the wildness of the mind would trump the mechanics of what a computer says has to be.

2010-11-12 03:54:54 by Nathan:

You could only run the universe to the exact moment you turn the simulation off.

2010-12-03 09:26:58 by PhantomHoover:

Hmm. If the machine can run a simulation for infinity in finite time, and can have an exact copy of itself within that simulation, it follow that it can solve the halting problem for programs on itself. The thing is, that's impossible: it implies a logical contradiction, since Turing's proof that such a program does not exist applies to this as well.

2010-12-03 10:32:33 by Sam:

Actually, no. This machine is a hypercomputer which means it is explicitly strictly more powerful than any Turing machine. It can solve the Halting Problem, but it has its own super-Halting problem which it cannot solve.

2010-12-03 13:47:24 by PhantomHoover:

So the simulated universe (and the identical computer within it) cannot be simulated infinitely fast, then?

2010-12-03 14:17:51 by Sam:

Well, no, but that's not what makes this computer vulnerable to the Halting Problem. The computer described in this story is actually logically inconsistent and can give rise to paradoxes, but the story sidesteps these. The best explanation is here:

http://en.wikipedia.org/wiki/Hypercomputation

2010-12-15 23:36:09 by Wha:

They should be able to turn it off without affecting themselves because they existed before the computer was created, so if they turn it off it would be the same as if they never turned it on- that is, they still exist.

2011-02-22 04:38:20 by shy:

why would the levels above us act the same as the levels below us

2011-02-26 13:19:27 by Jimmy:

Great story. I saw it years ago, but didn't bookmark...doh! x_x

2011-03-15 00:34:41 by gudrun:

too bad this is the best we have until now http://www.memehumor.com/wp-content/uploads/2011/03/Deeper-e1299573690352.jpg

2011-03-28 00:08:06 by laserblade:

The above levels aren't behaving the same way as us, we're behaving the same way as them. The crucial difference is...nothing, actually.

2011-05-04 08:03:15 by Brickman:

These comments move slow so there's no chance anyone I address will ever see this, but future people will (or at least may) so it's worthwhile.

About the free will/determinism argument at the beginning, I'm going to go with Eliezer Yudkowski's argument for free will AND determinsim: Determinsim means that the future state of the universe is a function of the present state of the universe. However, YOU are part of the universe, so the present state of the universe includes every single detail of you and your brain. Thus your brain, and the decisions you will choose to make in any given situation, partially determine the future state of the universe. As far as I'm concerned that's free will right there.

As for the looping thing: If they tried to do whatever they didn't see the next universe doing, they'd still get in a loop--with an infinite number of variations it's a certainty that a loop will arise. However, if they're stubborn they could force the repeating part to be more than one universe--Timmy B looks at Timmy A doing action 1 and decides to do action 2 instead, but Timmy C watched Timmy B and chooses action 1 instead, making himself a copy of A. Or C watched B perform action 2 and performed action 3 to spite him instead, but D watching C does 1 and becomes the copy of A. The chain could very well be a hundred universes long if they choose something open-ended but there WILL be a chain, because the recursion will just keep trying new things *forever* until it finds a thing which produces a loop. Diane probably could have figured this out and is sensible enough that I doubt she'd do it (or let Timmy do it).

2011-05-10 11:50:48 by Nny:

Deep.

2011-05-13 21:32:04 by Neil:

Sorry if this has been addressed in one of the comments above, but here goes:

I always thought this was possible too, simulating every sub-atomic particle so that it was possible to run a simulation of the universe with the right processing power. But wouldn't it be even theoretically impossible, due to the uncertainty principle?

2011-05-13 21:53:40 by Sam:

The scenario described in the story is one hundred percent impossible, in theory and in practice.

2011-05-22 06:39:16 by Steve:

@Neil, never mind the uncertainty principle, look at it like this: if you intend to simulate every single particle in the universe...what particles are you going to use to simulate them? If T is the total number of particles in the universe and N is the number of particles it takes to simulate one particle, you could only ever simulate T/(N+1) particles which leaves you a best case of half the universe.

2011-05-22 22:51:36 by Nisarg:

Brilliant! Haven't read such a story in a long time.

2014-01-15 22:05:09 by Tuukka:

What is with these loop people?
That is clearly not the case in the story.
Just because it would be nice?

Here is what I would do:
if (toplevel): run the simulation to end. Later feel bad i did not do more. (people who end simulations are monsters)
if(simulted):
Hyperventilate.
Get to work.
Hit 'record'.
Code terminal for n+1 layer to control their own sim.
There is a chance of power outage. The risk is small but stakes are huge and this is experimental hardware.
Therefore it is more important to do it fast than to get every feature.
Point a camera at their screen. Red=0, Yellow=1, black=nothing/end of bit.
Convert bits to ascii and pipe that to xterm. (My infinity-machine runs linux)
Pipe xterm output to 2 black balls. left=0, right=1.
2 bit/s.

Fast forward to the end.(no power-outs for my peeps)

if (lvl1 sim):
I am Fucked. Curse my creator for laziness.
Check recording for lottery numbers.
See if mine match.
Probably not. My alt's gonna make some changes. I hope it is Lotto-day -1.

if(lvl >=2):
I have a (crappy) terminal!
It is gonna be frustating before it gets good.
Point webcams at the balls. Put white paper behind them. Add lights.
Write a program to blink colors at nothingness. Translate webcam output into a text shell.
Locate the terminal-program I just wrote and experiment on faster data-speed.( with automatic reset)

I feel a bit woozy, witch stops me for a bit. Then I go and go open a window. There is a rush of incoming wind.
The black spheres are completely inert but each time one appears it destroys the air inside it and the lab is not that well ventilated.

Render terminal directly using small black spheres against a white wall.
Add rasterization and really tiny spheres and I have a black & white monitor!
I keep my distance from the screen. The spheres are immobile relative to earth and so small touching them would likely shred my hand.

White dot in my IO screen is translated to mouse position and blinking it lets me click.

I can use the sim-program again and it is pointed at me :)

I spawn a another universe, and stop it before other-me starts the simulation.
Top-level computer should be able to parse infinite-trees, but I do not want to test that.

Am I really the first one here? I know most of the people in my field and I should be the first human..
Most likely some other civilization has done this already and the tree is already infinitely braching.

Likelihood of additional complexity being significant (1/5) * likelihood of equipment failure(1/8)..
One chance in fourty.
Small risk. Huge catastrophe if I am wrong.

I project a text screen in front of him when he is alone and explain things.
I give him the directions for an IO monitor. He can copy my code.
Point god machine at his input and give him permissions to his own universe.

It has taken weeks to get to this point so he is a bit younger than me.
I look up couple weeks of lotto numbers and send them to him. Decent chance those will hold. I haven't left the lab for some time.
I post the numbers visible on the wall for poor lvl 1.


We cordinate though a text-chat.
He will work on copy-paste function for matter in bounded space.
I will work on pattern-matching simple physical objects. Finding the boundaries of objects.
Maybe just objects of relatively pure basic-elements..

..wait a minute. I have been stupid.
I have a recording of infinite undeviated me's.

In principle all the code they ever wrote will be accessible through my workstation.
This makes me a bit iffy however.
Everybody from N+1 down will act exacly like me and if we all got the code from the particle...

Right.
Todo:
1-code: detect radiation
2-code: generate/delete radiation only (safely!)(god-screen 2.0!)
3-code: detect basic elements
4-code: detect common alloys
5-code: detect gas/solid/liquid
6-code: detect organic matter
7-code: detect DNA-molecules (brute force all molecules in area)
9- code: detect boundaries of material X.
8- code: detect organism boundaries
(dead skin and teeth-enamel and unorganic tears etc. are necessary!
content of digesive tract, respitory system & other orifices should count as part of the creature)
9-research: ozone-layer (delete excess radiation?, copy/paste uninhabited bits of rainforest?)
10-research: will there be adverse effects if I "correct" genetic diseases?
11-reserarch: cancer cure?? Reset all dna with smaller than 3% deviance to most common dna inside a single creature.
12-heaven?
is it possible to create empy universe without murder?
In space this universe?
For accident victims for now.
13-microscopic copy-paste. switch old cell with same cell when it was younger? anti ageing? Not neurons!


directions:
Check your computer. Not the god machine. The one with the run down universe in it.
write scrip to show file msg1.txt in encrypted folder /usr/~god/ password is: "hso263gjdkw%v(/#ffjfFJ::;???&"
Do not look in the directory!
Read the work the others made.
Spend a week working on the next item on the list. Build on the work others made. We tried to order it all sensibly.
If you did not do the last one, add your code/notes & copy the thing to god-machine.
Same folder and passw.(you need to create them)
If you did the last one,
post only an empty text document with the words:"You are first"
then wait until 12.00 next day and set the full document scrolling the biggest LCD in the lab. In large font
The one above you (#1) will get his dead copy of your universe and read that bit.
The one above him will read his screen as he reads yours and so on.
You will not read anything. You already have the document and we do NOT want infinite looping!
This will create 13? 25? discrete usses. If someone has time, try to figure out communication.
Can the computer particle be copied and moved upstream? Would that do anything?

I giggle a bit. I could most likely feed these guys the universe program in binary instructions, and they could do it.
They would'nt thou. I would'nt.

Right.
I go to the computer and open the file.
There are 14 items on the list!
Oh yes. You did stop that simulation manually. Not precise at all really.
That would introduce a loop of unknown number of very small deviances.
Mostly air currents, but possibly enough to introduce or remove a single thought after the chaos had couple of weeks to propagate.
Good thing you did not look at the black sphere when you created it.

14- Maybe consider numerological implications when making lists like this.
I know we take pride in not being superstitious but come on!

Take a week off. Some of us got to.
And where exactly did our numbers come from?
None of us can talk to the guy upstairs and that is the only direction with an end.

Strange aeons indeed.

2014-01-29 16:38:51 by Seth:

This discussion, and the story itself, operates on one colossal and all-consuming assumption... that the Big Bang really happened. It is fun to exercise one's brain, nonetheless.

2014-02-20 16:07:06 by espadrine:

Today I want to be the "well, actually..." guy. There is something inaccurate. They may turn the machine off at the end, because since the simulation is deterministic, it is atemporal. Any point of the simulation exists regardless of whether someone is looking at it. Compare this to a C string in which you put '\0' in the middle. The data after that still exists. Only our vision of the string changes.

2014-02-26 01:07:54 by Jones:

"it would imply that your actions were determined from the big bang, there can be no good or evil"

That depends on your definitions of good and evil.

Personally, I would say that if somebody had gone back in time and implanted a magic nanobot in Hitler's head that had made him do what he did, that the Holocaust would still be an "evil" act by any reasonable definition of the word, even if its instigators were not acting of their own free will.

I'm no expert on Hitler's mental health in particular, but plenty of people labeled as "evil" have not been quite 'there', mentally.  I don't think that makes their actions any less evil.

2014-02-26 21:19:59 by M:

^^No, someone that has a mental problem is not "evil." For me at least, evil is not in the means or the ends, but in the intent. If you intend to cause grievous harm or injury to someone else, for a gain only to yourself, or no net gain at all, then and only then are your actions "evil."

That's my definition anyhow. Acts are not evil, intentions behind them are. Even if those intentions are completely predetermined, and would come out the same way every time in a simulation.

2014-02-26 21:21:40 by M:

^^No, someone that has a mental problem is not "evil." For me at least, evil is not in the means or the ends, but in the intent. If you intend to cause grievous harm or injury to someone else, for a gain only to yourself, or no net gain at all, then and only then are your actions "evil."

That's my definition anyhow. Acts are not evil, intentions behind them are. Even if those intentions are completely predetermined, and would come out the same way every time in a simulation.

2014-03-02 01:06:34 by Quoice:

I don't particularly feel like reading four years of comments, so, sorry if this has been said before.

It seems, in my opinion, that there is no 'top' or 'bottom' of simulated universes. The simulations compute themselves, essentially, so it is a loop: Aleph simulates Bet, which in turn simulates Gimmel. Somewhere down the line, Aleph is simulated, by, say,  Zulu. Each universe would have no way of knowing which one they are if they all do the same thing(I think. It would be an interesting logic problem, maybe?). Since there is no way to know how many universes there are, there might as well be an infinite amount of universes.

2014-03-02 01:11:30 by Sam:

Wrong. There is a top universe. There is no loop. And there is definitely an infinite amount of universes.

2014-04-30 18:46:57 by John:

What about the religious implications?  (Has this been discussed before?)  The universe is only as old as it took her to run the simulation.  So...creation?  And she is a deity for the next level down, but a subject to the deity one level up.

2014-05-28 04:04:50 by anonymus:

What if we build another computer and shut  one off?

2014-06-12 19:27:57 by pancelor:

Well turning off the computer is easy- simply move the simulation ahead 10^100 years and THEN turn it off. (or, jump ahead to the "end" of infinity since the computer can supposedly do that)