Is it possible to simulate a universe? ITS ALREADY STARTED.

I have always been intrigued by the idea that we could be in fact living in a simulation. I wont go into great detail regarding the theory as many of you are no doubt familiar with it. But I always wondered, would it ever be possible to simulate an entire universe? The computer power needed to render every atom in a simulated universe would be simply unimaginable. For those who are not here is a link that outlines it well:

www.simulation-argument.com...

Then I read some article where Singularity expert Ray Kurzwiel stated that with the way computer power is increasing exponentially, in a few decades we would have computers we could hold in our hands that would literally be trillions of times more powerful than current ones. So its conceivable that at some point the processing power will start to approach a level where something, crude by reality's standards might be possible. Years in the future to be sure.

Then I happened upon this site: en.spaceengine.org...

It is a free software program, where an entire universe is simulated. I have heard something on the order of 400 QUADRILLION stars planets and galaxies It pulls this off by displaying only what you are looking at at the moment. This is not a plug, Its free. Here is a video:








Our own solar system is rendered. As are a few of the known star systems based on available scientific data. But then there are billions and billions (sorry Mr Sagan) of "procedurally generated" star systems, based on a few basic rules on planet size, composition, gravity, etc. You can land on and explore the planets which are rendered down to a couple meters resolution. There are neutron stars, black holes and nebulae. Sounded great so I downloaded the program.


In a word, its simply STUNNING. Its so vast, you could never begin to see everything in this program in a hundred lifetimes. there are countless solar systems, planets, moons, stars and galaxies. As I said, you could never see all of it in a lifetime.

Its admittedly a far cry from this to simulating reality. BUT, its a beginning, a precursor to programs that will be able to do much more. In a century, a program like this could be overlaid with almost infinitely more detail populated with "sims" simulated life forms, etc. and the resolution could be much much much higher. So I now believe its possible to simulate the universe (albeit in the future)



Here are some screen shots:

Thanks for the post, it looks like a cool program.

S&F Dude!

While i am not a huge fan of the Simverse (can i copyright that word??), i am intrigued by the idea.
This program looks interesting, so i'm d/l it now. Should be fun to play with, at the least.

Cheers.
Scorp.

Excellent thread, thank you!

That program does look amazing, although I suspect my crappy computer wouldn't be able to handle it.

I wouldn't be too surprised to discover that this universe was a simulation. I've played the Sims enough times to realise that everything goes to sh# sooner or later!

Namaste

The main issue I have with the idea of simulation is that whoever the programmer is, cannot be as advanced as they should be to have that capability in the first place. There are plenty of design flaws and just general discrepancies with that idea.

For example, why waste so much time and programming creating hundreds of billions of unique stars in one galaxy alone, when so little of them are capable of life? Many stars would have to have been designed individually. A smart programmer would undoubtedly use a script to duplicate the design of one star into billions of them and simply copy it, but this clearly did not happen. Why? In a programmed simulation I'd expect most stars to be uniform across the board and each one to have a life sustaining planet with it. I'd expect things to be more similar, but even other galaxies are vastly different than ours. I'd expect things to be ideal conditions during most times. Things like seasons, moons effecting tides, etc etc, are all caused by natural phenomena. If they were programmed I'd expect the seasons to continue, but to have a mysterious cause. So far there is nothing really mysterious about how the universe works, other than how it got here, and what things are on the tiniest level. Plate tectonics, electromagnet shields generated from moving cores of planets, millions of individual species that would have to be individually programmed on this planet alone. None of it is the least bit logical.

Why program a spinning molten core to generate an EM field, rather than just programming the EM field itself? In a programmed universe, why is the EM field even necessary? Why has earth been bombarded with tons of space debris over the years, exterminating a large portion of life on earth each time? Wouldn't a programmer just program the earth to never get hit? After all he would know everything that's ever going to happen based on his parameters, right? Why are we threatened with solar flares, gamma ray bursts, and tons of other threatening things that could instantly wipe away all life on earth?

Why program complex evolving DNA and millions of diseases when you can just program a person's appearance directly? Why would you need to create DNA and evolution? If it was all done by a computer programmer, then the programmer is illogical and very bad at it, which doesn't make sense, considering how advanced one would have to be to have the capability to create a big bang or program trillions of stars, galaxies, planets and life forms. Then there is the huge waste of space between galaxies and stars. It's just not an efficient program and nearly every known phenomena in the universe is explainable via natural causes. In computer programs, the graphics are set up, and everything else is artificially generated. Light and daytime in computer games is not caused by physical planets revolving around stars. It is all programmed to do exactly what you want it to do. There aren't external causes for what happens.

Then of course you have to account for somehow programming consciousness / awareness. I have doubts that anyone capable of that, would make such an inefficient universe, and a universe where 99.9% of it is instantly lethal to life. There is a LOT of unnecessary extra programming. Programmed reality is not very likely, due to these reasons, plus many more that I don't have time to break down.

I see it now...

Sim City: Sim Universe

Populate your universe with various life forms, modify the laws of physics, track your Sim Aliens from planet to planet (you will need alot of schools on your planets to develop Sim Universe faster than light tech) start intergalactic wars... then, if you get bored, you can unleash awesome disasters like strange matter or gamma ray bursts...

Can I test the beta?

I'll have a nice greasy pastrami in the restaurant at the end of the universe!

reply to post by openminded2011

I know about this idea open. Procedural generated things are not new.

Here's the problem...

Temporal procedural generation is a different animal. It doesn't just involve mountains. It involves change-over-time. And that change has to make sense to a user.

We have accomplished a (rough) formula for planetary landscapes and (rough) solar systems (planets, moons, stars, etc) and (rough) galaxies and (rough) textures.

But (rough) is far from realistic.

What about creating formulas for life-forms? Again, whole different ball-game. We're FAR from being able to wrap up intelligence in a couple formulas. We do have formulas for movement patterns and maybe a few other things, but that's a far cry from a complete intelligence formula. Fact is, to create the illusion of strong intelligence, we'd presently use scripts, complex AI and lots of storage space.

Furthermore, procedural things cannot be changed by a user unless the changes are stored somewhere OR the changes are themselves procedural in nature. Players are NOT procedural... Any changes a player makes will be stored. and must then be included in procedural formulas. See, a program cannot know what a player will do, so it must store what a player does. And the more a player changes things, the more storage is needed and the slower it's.

What happens if your formulas are not realistic? What incomplete procedural things do is create the ILLUSION of existence and interaction and change. But if you examine them you will see it breakdown because roaming-algorithms are impractical. The whole point to use procedural things is to save space, but this comes at the cost of speed. This means you must sacrifice roaming algorithms that would create accurate interconnections. So if a user examines thigns closely, they will see no interconnections. Instead they will be underwhelmed by a kind of unprocessed chaos.

It's all about... if you stretch the illusion too far, it SNAPS. Then it's not good.

Originally posted by Barcs

The main issue I have with the idea of simulation is that whoever the programmer is, cannot be as advanced as they should be to have that capability in the first place. There are plenty of design flaws and just general discrepancies with that idea.

You don't have to program every single aspect of a simulation. You could just let the computer pick its own parameters using some kind of random number generator. Run the generator once, tweak the outcome a bit, press save, and voila, a universe that is the same wherever and whenever you run it.

You could choose to let the computer pick it's parameters on the go and never have the same universe. Pro is, your program is going to be a lot less massive to download. Con is, you won't have a way back to something you already explored. It would not exist anymore. You could traverse the same route and never see the same again. That's not realistic nor is it fun.

So you might choose to do a bit of both. Have some pre-rendered parameters saved, say, the positions of the stars and a global makeup of them and let the computer pick details on the go wich you let it save so you can trace back your steps.

The real work is going into creating the parameters the computer can choose from. The more parameters, the less likely it gets to end up with things that look the same.

You let the universe put detail only at places someone is looking at it.

Originally posted by jonnywhite
reply to post by openminded2011

 

I know about this idea open. Procedural generated things are not new.

Here's the problem...

Temporal procedural generation is a different animal. It doesn't just involve mountains. It involves change-over-time. And that change has to make sense to a user.

We have accomplished a (rough) formula for planetary landscapes and (rough) solar systems (planets, moons, stars, etc) and (rough) galaxies and (rough) textures.

But (rough) is far from realistic.

What about creating formulas for life-forms? Again, whole different ball-game. We're FAR from being able to wrap up intelligence in a couple formulas. We do have formulas for movement patterns and maybe a few other things, but that's a far cry from a complete intelligence formula. Fact is, to create the illusion of strong intelligence, we'd presently use scripts, complex AI and lots of storage space.

Furthermore, procedural things cannot be changed by a user unless the changes are stored somewhere OR the changes are themselves procedural in nature. Players are NOT procedural... Any changes a player makes will be stored. and must then be included in procedural formulas. See, a program cannot know what a player will do, so it must store what a player does. And the more a player changes things, the more storage is needed and the slower it's.

What happens if your formulas are not realistic? What incomplete procedural things do is create the ILLUSION of existence and interaction and change. But if you examine them you will see it breakdown because roaming-algorithms are impractical. The whole point to use procedural things is to save space, but this comes at the cost of speed. This means you must sacrifice roaming algorithms that would create accurate interconnections. So if a user examines thigns closely, they will see no interconnections. Instead they will be underwhelmed by a kind of unprocessed chaos.

It's all about... if you stretch the illusion too far, it SNAPS. Then it's not good.

edit on 23-5-2013 by jonnywhite because: (no reason given)

Granted by our current standards this is correct. But what will quantum computing be able to do in 1000 years? We went from Pong to photo realistic virtual worlds in games in less than a human lifetime. Try to imagine that curve, over a period of a thousand years, AIDED by sentient artificial intelligence at some point. I dont think its crazy to imagine a simulated universe that is as real to the simulated beings inside of it as ours is to us.

reply to post by D.Wolf


Programming a star generating algorithm would be time consuming and would have to include the initial programming for every single possible parameter of a star. It's not as simple as a random number generator. We're talking stars that utilize nuclear fusion, have individual life cycles, emit energy, etc etc. Since there are millions of unique stars, that would require millions of different programmed conditions or parameters, down to the last molecule. If one specific type of star is ideal for life, then why generate millions of other unique stars? It is a colossal waste of time and space. No smart programmer puts in lines upon lines of extra code that has no real purpose. That would make for a very inefficient program that would be bulky and buggy and run like crap. The universe is like that, except 99.9999% is wasted space, if the purpose is life. Why would you program a star that dies and takes all life on the planet along with it? I'd expect a simulated reality to have stars that never die and for things in reality to simply work out, without having scientific explanations. I'm not saying it's definitely false, because you never really know, but I'd believe a physically created universe over a programmed one just based on my experience in programming.

reply to post by openminded2011

The caveat is the simulated reality would only captivate those "sentients" living inside it. You could have those sentients living inside a text-based version of pacman and they'd think it's real. This is kind of like how when we dream we think it's real, but it's completely subjective.

It comes down to quality versus quantity. You use procedural-programming techniques to generate content/behavior and you get quantity. You use handmade techniques to create content/behavior and you get quality. Quality means more consistent/specific physics/content. Quantity means that things are copy/pasted with weak consideration to the interconnections. This means finding a village with weak connections to things around it. It means generic rules rather than more specific ones.

So it's an inverse relationship between quantity/quality.

I will say though that fractal patterns in nature and even quantum weirdness DOES make me wonder about our universe. I don't know what it means, really. It makes sense to me that common rules across the universe, like gravity, might reproduce certain patterns across many things. Lifeforms, for example, might deal with gravity in a predictable way, thus reproducing a particular pattern. As for quantum weirdness, I don't have the level of education I need to know what it means. I've only read elementary things. I think the principle of locality is possibly on the chopping block because either it or something else is violated. I also know that we cannot know the velocity and position of individual particles. Quantum weirdness means we cannot know our future with certainty. There is only a weak(er) form of determinism. This indicates a possible break in interconnectedness.

It's strange to me. Determinism implies we're slaves to fate because there's only one outcome from a set of circumstances. Non-determinism implies we're slaves to randomness (the unknowable) that forces many outcomes from the same set of circumstances. It seems to me that reality is a balance of some determinism and some unpredictability. So while we can plan our life and increase our survival odds by making good choices, we cannot ensure guaranteed results because of the randomness.

One other thing... Lately my mind has been doddling with the idea that there's no "sudden" emergence of intelligence in our universe. Instead it was always there, at different amounts. Intelligence is a property of a particle or group of particles to react to their environment either partially or completely across intervals of time. The greater the interval of time and the greater the response, the more intelligent the particle or particles appear to be. For example, a rock will sit and wait patiently as another rock is on a collision course. The rock will only respond when it's struck hard by the other rock colliding with it. A human, by contrast, will not be so unprepared. They will see the rock on a collision course before it strikes them and will attempt to move out of the way. Both the rock and the human respond, at least partially, but only the human responds across a larger interval of time. This interval of time is a result of our much more intense sensory/memory system. The key is to understand that they both have some measure of sensory/memory/response capability.

The problem with our definition of life is it's too specific. Our definition causes a divide and creates the impression that life "suddenly" emerges from the primordial ooze. This leads to things like hte philosophy of substance dualism. This really is just anti-science, IMHO.

And in about 45 years time when technology reaches a point where we can uploads our mind into machines is where the big epiphany happens.

We are in closed loop of consciousness once uploaded, the consciousness forgets its uploaded and after millions and millions of copies we come close to the end of the loop again(2013) where we create models of the universe on computers and advance out tech to upload our minds into machines and the loop starts again.

Certain secret societies hold the keys and have this knowledge as can be found all over ATS and the virtual reality we call the internet.