186 replies to this topic

[caston]

Would it be possible to simulate something e.g. a virus with a similar amount of energy to what the virus would require to carry out its life cycle outside of the simulator?

I'm talking about something that would be akin to using sub-atomic and quantum particle effects to do something that is halfway between the invitro "test tube" in the lab and the computer based simulation.

Edited by caston, 17 August 2007 - 01:26 PM.

[dimasok]

I'd still like to discuss the feasibility of us emulating something like ourselves in the futureu... From Wiki:

As of 2007, the computational requirements for Molecular dynamics are such that it takes several months of computing time on the world's fastest computers to simulate 1/10th of one second of the folding of a single protein molecule. To simulate an entire galaxy would require more computing power than can presently be envisioned, assuming that no shortcuts are taken when simulating areas that nobody is observing.

Hmm...

[Live Forever]

Leave a message for those who are running our simulation! (and win a real prize in the process):
http://tierneylab.bl...n-a-real-prize/

[Luna]

Leave a message for those who are running our simulation! (and win a real prize in the process):
http://tierneylab.bl...n-a-real-prize/

You seem convinced.

[Live Forever]

Leave a message for those who are running our simulation! (and win a real prize in the process):
http://tierneylab.bl...n-a-real-prize/

You seem convinced.

Convinced about the prize or about the message going to the simulators? I thought it was a funny blog posting.

[Live Forever]

They talk a lot about living in a simulation (among other things) in the latest bloggingheads, for anyone interested:
http://bloggingheads...ideo.php?id=369

[platypus]

BTW, I think the simulation-argument is somewhat ludicrous. The overlords at the end of time would need to simulate the whole multiverse (i.e. each possible outcome of all quantum processes during all time, in other words all possible histories of the universe). Has it even been shown that the computational resources to do that even fit this universe? Can you simulate all possible histories of the universe quickly enough?

Actually, they would only need to simulate all the combined brainpower of all humans who have ever lived.

Where did the simulators get the blueprints for human brains without having to simulate all possible outcomes of the whole evolutionary process on earth? You can reply in the other thread and I'll pick it up from there.

[platypus]

How will the simulators get the blueprints for human brains without having to simulate all possible outcomes of the whole evolutionary process on earth? (this is a response to an argument made elsewhere that "only" human brains need to be simulated, not the whole universe).

[Live Forever]

BTW, I think the simulation-argument is somewhat ludicrous. The overlords at the end of time would need to simulate the whole multiverse (i.e. each possible outcome of all quantum processes during all time, in other words all possible histories of the universe). Has it even been shown that the computational resources to do that even fit this universe? Can you simulate all possible histories of the universe quickly enough?

Actually, they would only need to simulate all the combined brainpower of all humans who have ever lived.

Where did the simulators get the blueprints for human brains without having to simulate all possible outcomes of the whole evolutionary process on earth? You can reply in the other thread and I'll pick it up from there.

I moved it here for ease of answering, but now after I did, I see that you restated it here as well. Oh well, no bother. )

There are several possible answers; Perhaps they are humans (or post-humans) who have scanned their brains in using ultra sensitive equipment (nano-tech?) and reverse engineered it, or perhaps they created an intelligence much the same way people are trying to design AGIs (plenty of discussion on that in the AI forum if you are interested) and decided to place them into our world, or perhaps it is something that I can't think of because they are so much smarter than me so I can't think of it. (or they designed the system to not let me think of it) Really, any way that has been thought of to create an AI of human level intelligence could be applied, and I am sure they have thought of several more that we are as yet unable to think of.

[platypus]

Where did the simulators get the blueprints for human brains without having to simulate all possible outcomes of the whole evolutionary process on earth? You can reply in the other thread and I'll pick it up from there.

There are several possible answers; Perhaps they are humans (or post-humans) who have scanned their brains in using ultra sensitive equipment (nano-tech?) and reverse engineered it,

Ok.

or perhaps they created an intelligence much the same way people are trying to design AGIs (plenty of discussion on that in the AI forum if you are interested) and decided to place them into our world, or perhaps it is something that I can't think of because they are so much smarter than me so I can't think of it. (or they designed the system to not let me think of it) Really, any way that has been thought of to create an AI of human level intelligence could be applied, and I am sure they have thought of several more that we are as yet unable to think of.

But why do humans and animals look like they have been shaped by evolution, if evolution never took place in this simulation? Are the simulators somehow malevolent and are trying to deceive us? If we believe in the latter, we can trust nothing, not even our own reasoning.

Edited by shepard, 23 November 2007 - 05:40 AM.

[Live Forever]

But why do humans and animals look like they have been shaped by evolution, if evolution never took place in this simulation? Are the simulators somehow malevolent and are trying to deceive us? If we believe in the latter, we can trust nothing, not even our own reasoning.

I couldn't venture to speculate on their intentions; That is the point. (as far as the evolution question, they could make it look however they wanted, perhaps modeled after their own world, or perhaps they ran an evolution subroutine; We have the capability to model evolutionary systems with our own computer systems, so it wouldn't even take many advances in computing to accomplish that) You are correct that if it was a simulation, that basic reasoning and logic about the nature of the world would be called into question.

All this is, of course, secondary to if it is actually possible to build such a simulation and the odds of it being the case that we live in one. If Bostrom's logic is to be believed (and I can find no error in it) then if we are able to build these types of simulations, we virtually have to be living in one. (and I can't possibly see how we won't be able to build them in the future)

Edited by shepard, 23 November 2007 - 05:40 AM.

[platypus]

But why do humans and animals look like they have been shaped by evolution, if evolution never took place in this simulation? Are the simulators somehow malevolent and are trying to deceive us? If we believe in the latter, we can trust nothing, not even our own reasoning.

I couldn't venture to speculate on their intentions; That is the point. (as far as the evolution question, they could make it look however they wanted, perhaps modeled after their own world, or perhaps they ran an evolution subroutine; We have the capability to model evolutionary systems with our own computer systems, so it wouldn't even take many advances in computing to accomplish that) You are correct that if it was a simulation, that basic reasoning and logic about the nature of the world would be called into question. .

The other point is that either the simulators are a) deceiving us or b) the biosphere we see really is formed in an evolutionary process, or a complete enough simulation of one. As randomness has a large part in the evolution of the Galaxy, Earth, life, humans and their brains, the simulators would need to be our descendants to have the blueprints for our brains, or they would need to simulate the whole evolutionary process or possiby the whole universe (actually all possble worlds, as quantum mechanics appears to be truly random).

All this is, of course, secondary to if it is actually possible to build such a simulation and the odds of it being the case that we live in one. If Bostrom's logic is to be believed (and I can find no error in it) then if we are able to build these types of simulations, we virtually have to be living in one. (and I can't possibly see how we won't be able to build them in the future)

Boström's argument is rather easy to criticize, see for example these short critiques:

http://henrysturman....simulation.html
http://www.stanford....tromReview.html

I don't think we understand the nature of reality, consciousness or the possibility of "free" will nearly enough to answer whether building a simulation of consciousness is possible.

Edited by shepard, 23 November 2007 - 05:41 AM.

[Live Forever]

But why do humans and animals look like they have been shaped by evolution, if evolution never took place in this simulation? Are the simulators somehow malevolent and are trying to deceive us? If we believe in the latter, we can trust nothing, not even our own reasoning.

I couldn't venture to speculate on their intentions; That is the point. (as far as the evolution question, they could make it look however they wanted, perhaps modeled after their own world, or perhaps they ran an evolution subroutine; We have the capability to model evolutionary systems with our own computer systems, so it wouldn't even take many advances in computing to accomplish that) You are correct that if it was a simulation, that basic reasoning and logic about the nature of the world would be called into question. .

The other point is that either the simulators are a) deceiving us or b) the biosphere we see really is formed in an evolutionary process, or a complete enough simulation of one. As randomness has a large part in the evolution of the Galaxy, Earth, life, humans and their brains, the simulators would need to be our descendants to have the blueprints for our brains, or they would need to simulate the whole evolutionary process or possiby the whole universe (actually all possble worlds, as quantum mechanics appears to be truly random).

I agree, although "deceiving" is kind of misleading since if we were living in a simulation, all of reality would by definition be deceiving us. But, yes, we could be being deceived or an evolutionary process is being simulated. You have to understand though the magnitude of the system we are talking about. If I spend all of my brainpower for my entire life, I could come up with a fairly complex evolutionary system that would fool most people, I think...and that is just one brain. We are talking about simulating all of the brainpower of all of human history. Devote a few thousand (or hundred thousand, or million, or whatever) brainpowers to designing an evolutionary system that seems sufficiently randomized and that wouldn't even increase the size of the system by a minuscule fraction of percent. Hell, the new Spore game can do it in a fun way, and there are modeling programs that we just have in today's limited world that can virtually do what you are talking about. If you could simulate all of the brainpower of human history, simulating an evolutionary process would be a piece of cake, I assure you.

Boström's argument is rather easy to criticize, see for example these short critiques:

http://henrysturman....simulation.html
http://www.stanford....tromReview.html

I don't think we understand the nature of reality, consciousness or the possibility of "free" will nearly enough to answer whether building a simulation of consciousness is possible.

Yes, thank you for putting those links up again. They are of course discussed here as well: http://www.simulism.org/Arguments which was another link mentioned earlier. (just restating it since you were restating those) The problem I have with most of the counter arguments is that they seek to postulate what the possible simulators know or don't know, or their values or non-values, or intentions, or... (etc, etc) Bostrom talks about none of that stuff in his original argument, and although it is fun to speculate on all the possibilities, the initial question to ask is whether it is possible and how likely it is. Anything beyond those two questions is just speculation and impossible to know one way or another. So, again, (sounding like a broken record at this point I know, haha) the question is whether we will at some point be able to run such simulations. In theory it is already very possible (as explained earlier) just using currently understood design principles. (in under a second using less than 1% of the system's resources, etc as described before) If we can do it, then we virtually have to be living in one, period. Now if there is some good reason why we won't be able to create such a system (or would choose not to), then that would be persuasive, but as yet I have not heard one. (and I have read all the arguments on both sides, I assure you)

I would recommend for anyone interested to read the simulation FAQ. It answers some of the more basic questions on a very basic level. I mentioned that link earlier in the thread, but I figured since the thread is getting bumped again, some people might not know what the basic argument is. There are lots more links to different things related to the argument at http://www.simulation-argument.com/

Edited by shepard, 23 November 2007 - 05:42 AM.

[platypus]

The other point is that either the simulators are a) deceiving us or b) the biosphere we see really is formed in an evolutionary process, or a complete enough simulation of one. As randomness has a large part in the evolution of the Galaxy, Earth, life, humans and their brains, the simulators would need to be our descendants to have the blueprints for our brains, or they would need to simulate the whole evolutionary process or possiby the whole universe (actually all possble worlds, as quantum mechanics appears to be truly random).

I agree, although "deceiving" is kind of misleading since if we were living in a simulation, all of reality would by definition be deceiving us

Yes. Such a position is philosophically about as fruitful as postulating that we are "brains-in-a-vat" or "deceived by an evil demon". We could trust nothing, not even our senses or reasoning.

If you could simulate all of the brainpower of human history, simulating an evolutionary process would be a piece of cake, I assure you.

Since evolutionary process is partially truly random when quantum effects are involved (for example cosmic rays scattering off the high atmosphere and damaging the DNA of your ancestors, thereby causing point mutations), the simulator overlords would have to simulate all possible evolutionary outcomes. If quantum effects are involved in the making of consciousness, it might also be necessary to simulate all possible successive brainstates, which does not sound very easy to me.

I don't think we understand the nature of reality, consciousness or the possibility of "free" will nearly enough to answer whether building a simulation of consciousness is possible.

So, again, (sounding like a broken record at this point I know, haha) the question is whether we will at some point be able to run such simulations. In theory it is already very possible (as explained earlier) just using currently understood design principles. (in under a second using less than 1% of the system's resources, etc as described before).

That's speculation, it's possible that consciousness is not substrate-independent but is a property of certain configurations of matter, for example. Also, I don't buy the presently popular freedom-of-will-does-not-exist - rap but see it as an integral part of consciousness. If conscious beings have "free" will to even a small degree, the needed amount of simulation explodes.

BTW, I don't want to sound racist but instead of a group of philosophers I'd like to see a group of physicists debating this issue. Of course, if this is a simulation all physicists are as deluded as everyone else [lol]

If we can do it, then we virtually have to be living in one, period. Now if there is some good reason why we won't be able to create such a system (or would choose not to), then that would be persuasive, but as yet I have not heard one. (and I have read all the arguments on both sides, I assure you).

I doubt that "we" could do it and I'm virtually certain that beings orginally residing on another planet or galaxy cannot do it. Unless you want to say that in the beginning "god" started this simulation...

Edited by shepard, 23 November 2007 - 05:44 AM.

[platypus]

http://perso.orange....rd/pdfrefut.pdf

[Live Forever]

The other point is that either the simulators are a) deceiving us or b) the biosphere we see really is formed in an evolutionary process, or a complete enough simulation of one. As randomness has a large part in the evolution of the Galaxy, Earth, life, humans and their brains, the simulators would need to be our descendants to have the blueprints for our brains, or they would need to simulate the whole evolutionary process or possiby the whole universe (actually all possble worlds, as quantum mechanics appears to be truly random).

I agree, although "deceiving" is kind of misleading since if we were living in a simulation, all of reality would by definition be deceiving us

Yes. Such a position is philosophically about as fruitful as postulating that we are "brains-in-a-vat" or "deceived by an evil demon". We could trust nothing, not even our senses or reasoning.

Good to see we agree on this point.

If you could simulate all of the brainpower of human history, simulating an evolutionary process would be a piece of cake, I assure you.

Since evolutionary process is partially truly random when quantum effects are involved (for example cosmic rays scattering off the high atmosphere and damaging the DNA of your ancestors, thereby causing point mutations), the simulator overlords would have to simulate all possible evolutionary outcomes. If quantum effects are involved in the making of consciousness, it might also be necessary to simulate all possible successive brainstates, which does not sound very easy to me.

No, you would only have to simulate it well enough to fool the minds that you are simulating. (or mislead them, as you stated) Although, like I said, speculation about motives is just a guessing game, I would think the purpose of said simulation would be to study the minds (or use it as a game, or something to do with the minds) not the study of evolutionary processes.

I don't think we understand the nature of reality, consciousness or the possibility of "free" will nearly enough to answer whether building a simulation of consciousness is possible.

So, again, (sounding like a broken record at this point I know, haha) the question is whether we will at some point be able to run such simulations. In theory it is already very possible (as explained earlier) just using currently understood design principles. (in under a second using less than 1% of the system's resources, etc as described before).

That's speculation, it's possible that consciousness is not substrate-independent but is a property of certain configurations of matter, for example. Also, I don't buy the presently popular freedom-of-will-does-not-exist - rap but see it as an integral part of consciousness. If conscious beings have "free" will to even a small degree, the needed amount of simulation explodes.

BTW, I don't want to sound racist but instead of a group of philosophers I'd like to see a group of physicists debating this issue. Of course, if this is a simulation all physicists are as deluded as everyone else [lol]

It is speculation, no doubt. But it is speculation using current design principles. (in other words, no new concepts would need to be created) I am not going to restate the entire design principle for such a computing system, because everyone is just as capable as me of reading it, but I haven't read any serious challenges to the feasibility of constructing such a computer (by physicists or otherwise) in the future. (assuming we don't blow ourselves up or something in the mean time, of course) Of course, like I said, I am open to any such arguments on feasibility and likelihood of such an event occurring, because those are really the only ones that hold water in whether we actually live in a simulation or not.

If we can do it, then we virtually have to be living in one, period. Now if there is some good reason why we won't be able to create such a system (or would choose not to), then that would be persuasive, but as yet I have not heard one. (and I have read all the arguments on both sides, I assure you).

I doubt that "we" could do it and I'm virtually certain that beings orginally residing on another planet or galaxy cannot do it. Unless you want to say that in the beginning "god" started this simulation...

Well, if you doubt we could do it, then I would be very interested in your reasons why. (why it is technically not possible) Like I have said repeatedly, whether or not we are able to build one is the determining factor in how likely it is that we are actually living in one.

Edited by shepard, 23 November 2007 - 05:46 AM.

[Live Forever]

http://perso.orange....rd/pdfrefut.pdf

Yes, a couple of the main points of that are answered in the FAQ, like I said earlier. (and even references that paper) Fairly easy to answer, in fact.

Edited by shepard, 23 November 2007 - 05:44 AM.

[basho]

But why do humans and animals look like they have been shaped by evolution, if evolution never took place in this simulation?

Why do humans and animals look like they have been shaped by evolution in a computer game if evolution never took place in the game?

Boström's argument is rather easy to criticize, see for example these short critiques:

Any argument is easy to criticize if you don't care about the validity of the criticism. For example, from your link one of the counter arguments is:

"And in a simulation we might as well have been given the option to magically move objects from A to B with the flash of a thought. Thus, the fact that no such magic or inconsistency exist is an argument for our world not being a simulation."

That is a pretty clear example of a logical fallacy, a circular argument with the conclusion in its premise.

I doubt that "we" could do it and I'm virtually certain that beings orginally residing on another planet or galaxy cannot do it.

All the evidence indicates otherwise. Arguments that the human mind is non-computational and that there is no possibility that it could run on a universal computational medium simply do not hold up very well.

Edited by shepard, 23 November 2007 - 05:47 AM.

[Luna]

Even if we can do it and even if we can't, we can't know by the arguments we have here in our world if it is indeed a simulation or if it is not.

It's just like God, the only possible way to know if that is the truth is if the simulator will tell us.
Depressing, isn't it?

[basho]

Even if we can do it and even if we can't, we can't know by the arguments we have here in our world if it is indeed a simulation or if it is not... the only possible way to know if that is the truth is if the simulator will tell us.
Depressing, isn't it?

Stephen Hawking has himself argued something similar, that ultimately we can never fully understand the nature of the Universe: "Gödel and the end of physics".

Edited by shepard, 23 November 2007 - 05:48 AM.

[Liquidus]

Well one thing is mostly certain. Either we're obliviously (well most humans) living inside of a simulation created by a post-human character, or we're the initial batch of potential post-humans that we assume we will be. Assuming life-extension becomes efficient in the next half-century, most of us should survive long enough to benefit. If we benefit, there's really no saying just how long, at least individually, we might be able to live (long enough to experience post-humanism).

In that event, there's a very good chance we will create our own simulations given the technologies of the future. We're partially already doing this now in some video games (ie. The Sims). So I guess in any event, consciousness, in the next couple of hundred years, will either be proven to be currently existing inside a simulation, or we will be spawning consciousness inside our own simulations.

As for not understanding the universe, I don't think that I would agree with Hawking. It wasn't too long ago when life seemed absolute to this world, and I'm sure the majority of people naturally assumed that our curiosity could only be limited by what we could observe on this planet. Noting that any planet is relatively vast and full of observation, there was little need back then to go beyond that. As space observation and subsequent exploration became more possible, the ideas of observation expanded. Today, we have the universe opened up to us compared to previously thinking we'd never be able to observe anything beyond earth. Now we're relatively on the brink of exploring another planet, and eventually even colonizing it.

I think it's only really a matter of time before we start dissecting the universe. It might not be next decade, but eventually.

[Luna]

No need for simulations.
If (WHEN) everything goes right and we gain full knowledge and control of physics, we'll become true immortals and shape the world as the best fairy tales.

No one will have death and the world will be a beautiful place for all of us.

I predict we'll gain that knowledge and I hope all of us here can be there when we do.
The only question which will remain is not about the the universe, but why the rules are as they are.

[Live Forever]

Even if we can do it and even if we can't, we can't know by the arguments we have here in our world if it is indeed a simulation or if it is not.

It's just like God, the only possible way to know if that is the truth is if the simulator will tell us.
Depressing, isn't it?

True we can never know for sure, it just increases the odds astronomically high if we can do it. (unless of course they shut off the simulation right before we attempt it to save on computing power)

Edited by shepard, 23 November 2007 - 05:48 AM.

[Liquidus]

Even if we can do it and even if we can't, we can't know by the arguments we have here in our world if it is indeed a simulation or if it is not.

It's just like God, the only possible way to know if that is the truth is if the simulator will tell us.
Depressing, isn't it?

True we can never know for sure, it just increases the odds astronomically high if we can do it. (unless of course they shut off the simulation right before we attempt it to save on computing power)

Flashes of the Matrix as an infinite loop, all over again.

Whilst mentally stimulated with a friend the other evening, we contemplated the Matrix theory in fine detail. I'm studying philosophy this semester, but we're still on the initial parts of it (Aristotle/Plato), if my memory serves me right, Descartes was the modern philosopher who originated the philosophies which lead to the creation of the Matrix story right?

Edited by shepard, 23 November 2007 - 05:48 AM.

[forever freedom]

Thinking about the simulation subject, i came to the conclusion that some of us (in this world) might be already immortal.


We might be beings that were born in the past-singularity-immortality future, that wanted to experience how the past looked like, how evetything looked like before all the technological advances that came after the singularity, the same way that we would like not going to the middle ages or to the ancient times to see what it was like.

Assuming the above, and assuming that we will be able to simulate an entire lifetime in 5 minutes, we could also assume that if someone from the future wanted to live the full experience of the past, they would want to forget about their lives in the future, and believe, while in the simulation, that they really were born now in our times, just like when we are dreaming, when we don't remember the reality, and are only aware of what's going on in the dream.

So maybe if one of you throw yourself in front of a car and die, you may wake up some 200 years ahead in the future


This possibility doesn't exclude the fact that, in the case described above, there may be at most some 1000 real people (considering that the simulations in the future could be like today's MMORPGs, where a lot of people can go online in the same world as long as they start the simulation at the same time, just like an online game room that starts the game as soon as enough people have joined - remember that a lifetime simulation could take 5 minutes long only in the future, so this idea is plausible) in this simulation. All the rest would be just unlucky AI avatars created to populate these simulations, that will have their existence ceased as soon as the people in the future decide to give the simulation we live in an end.

[Luna]

My only arguement against simulation is that it's cruel and a MURDER.
If anyone would make a simulation with living people in it, I think I would jail him for all eternity.

[forever freedom]

My only arguement against simulation is that it's cruel and a MURDER.
If anyone would make a simulation with living people in it, I think I would jail him for all eternity.

By living people you mean AIs, right? That's a very important moral issue. Maybe the laws in the future won't let any simulation that's created to be shut down, avoiding the extermination of all AIs inside of it. Maybe simulations are in fact parallel universes? So they aren't any less real than the reality that created this simulation. With the singularity, i see the possibility of parallel universe creations, so this theory can't be discarded.

Edited by shepard, 23 November 2007 - 05:49 AM.

[Luna]

It's not just shutting them down.
If you create life, and you know you can make them have no disease and death but you let them have those, you are a killer.

[Live Forever]

My only arguement against simulation is that it's cruel and a MURDER.
If anyone would make a simulation with living people in it, I think I would jail him for all eternity.

Trying to envision the morals of beings so much smarter than us is probably a lesson in futility. Do you consider it murder to eat meat (kill a cow or a pig?) or to kill insects that invade your home? The difference in intelligence level between us and anything that could create such a simulation (if it helps, think of them as "post singularity" beings) would be much greater than between a cow and ourselves. They might not think anything more about our lives than we do about killing bacteria or something. (not that they do or don't, just that we can't cast our personal morals on others and expect them to do the same thing as us)

Edited by shepard, 23 November 2007 - 05:49 AM.

[basho]

My only arguement against simulation is that it's cruel and a MURDER.

I'd assume the source code and data are not erased. The simulated entities could be restarted from an earlier save-point, or similar scenarios could be rerun many times. But actually, that's morally troubling too.

Edited by shepard, 23 November 2007 - 05:34 AM.