The ***new*** atheist.

[SciFiAuthor]

Now that is a very interesting discussion. But allow me to be the cynic and add to the list of possibilities

4. It is not possible to perform a lattice calculation of all the phenomena necessary to simulate a consistent and physically realistic universe

There are many problems with lattice approaches and one has to be very careful that in the limit of infinitely small lattice spacing the continuum limit is reproduced which it rarely is in most cases. It is even a struggle to make contact with reality using lattice QCD sometimes. So I don't think it's as simple as saying we can represent everything on a lattice, maybe the universe just isn't like that. You lose Lorentz invariance when you do relativity on a lattice for example which makes a big problem if you're trying to represent a world like the one we seem to live in. There are other reasons why I'd question this proposal too. It doesn't really seem to be falsifiable and introduces a whole lot of other problems.

It's still a really neat idea. I do like that there are observational consequences to it, so this can (in principle) be tested.

Great points. I think ultimately we suffer from a lack of a GUT. We really need to figure out the disconnect between relativity and quantum theory before we can even begin to think of approaching the possibility of simulating an entire universe. That said, nothing says we don't have millions of years of computer and technological development ahead of us. I'd be hard pressed to say that there won't come a day where we understand enough for a simulation, and there's really no time limit on Nick Bostrom's hypothesis. And if we can't figure it out, it's a good bet that artificially intelligent computers can. Other questions being, what is an accurate simulation of a universe? In comparison to what? What is inaccurate? Would we know about it if it was?

But where it gets really freaky is when you start observationally testing it and you find that the act of observation can affect the outcome, such as the double slit experiment. I'm not saying that the experiment indicates or suggests anything, rather that it's sort of spooky. But it may indeed be demonstrable whether we live in a simulation or not at some point. In which case, the big question gets answered. Perhaps that's the point at which the IT geek god says "Armageddon time biatches!" and ends the game.

[WOTR]

...It is even a struggle to make contact with reality using lattice QCD sometimes. So I don't think it's as simple as saying we can represent everything on a lattice, maybe the universe just isn't like that. You lose Lorentz invariance when you do relativity on a lattice for example which makes a big problem if you're trying to represent a world like the one we seem to live in...

Yeah... what he said.

There are far too many smart people on this forum.  Now I have to go look up Lorentz invariance as I had not heard of it...

I would like to revise my earlier ramblings about the reason that I posted this topic here... I posted it here because of the varied points of view that are here.  Like I said- pointless to bring it up at a church group- pointless to bring it up with any group who all believe the same thing.  Much more interesting to bring it up in "mixed company."  Being as the only thread that ties posters here together is a dislike of the sNoron it brings out some good discussion and points that I have to go look up and ponder...

...Other questions being, what is an accurate simulation of a universe? In comparison to what? What is inaccurate? Would we know about it if it was?

And thank-you for once again wrecking my brain.  I refuse to force my brain down the road of time manipulation and you bring me this instead?  Are you aware that I will be thinking of this for the next week?

Edit... BOTH of you can piss off (joking).  SciFi brings me philosophy and you bring me string theory and spacetime?  I managed to wrap my head around relativity and a few other broad ideas and now this?

I suppose pondering either and looking up Lorentz and trying to understand it is a better use of my time that listening to Jorch ask if mean babies come from portals...

[SciFiAuthor]

Other questions being, what is an accurate simulation of a universe? In comparison to what? What is inaccurate? Would we know about it if it was?

I should probably pre-elaborate on that statement. We have an issue: that the universe got its parameters so right. If they'd have been different it would have resulted in a very different universe than this one:

http://en.wikipedia.org/wiki/Fine-tuned_Universe

It's actually highly unlikely the universe would just happen to be the way it is. As such we're left with a sort of unpalatable choice. We can either say "For the universe to be this way, it must be possible for other universes to be other ways" in which case String Theory, M theory, etc. envision a multiverse simply to solve the problem of why our universe is so perfect as to eventually form intelligent life. It gets worse that observation defines so much:

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

Alternatively we can say "The universe is a simulation. It's perfect because someone designed it that way, but maybe there are flaws. Such as how cosmic rays disseminate . . . " That may in fact be the simpler answer.

I don't know which is correct, obviously, but I find myself inherently uncomfortable with either proposition. I don't like proceeding from the premise that there must be other universes just because this one is so perfect and we're here because we're on the jackpot planet in the one universe out of an infinite number that just happens to be the jackpot cosmos (God does not run a casino, I hear Einstein saying from the grave. Oh but he apparently does!), nor do I like the idea that we live in a designed universe. But we must face the fact that it's one of the two, and that has been the driving force behind numerous martinis for me.

[SciFiAuthor]

Yeah... what he said.

There are far too many smart people on this forum.  Now I have to go look up Lorentz invariance as I had not heard of it...

I would like to revise my earlier ramblings about the reason that I posted this topic here... I posted it here because of the varied points of view that are here.  Like I said- pointless to bring it up at a church group- pointless to bring it up with any group who all believe the same thing.  Much more interesting to bring it up in "mixed company."  Being as the only thread that ties posters here together is a dislike of the sNoron it brings out some good discussion and points that I have to go look up and ponder...
And thank-you for once again wrecking my brain.  I refuse to force my brain down the road of time manipulation and you bring me this instead?  Are you aware that I will be thinking of this for the next week?

Edit... BOTH of you can piss off (joking).  SciFi brings me philosophy and you bring me string theory and spacetime?  I managed to wrap my head around relativity and a few other broad ideas and now this?

I suppose pondering either and looking up Lorentz and trying to understand it is a better use of my time that listening to Jorch ask if mean babies come from portals...

Jorch is right. Babies do come from portals -- of sorts. Lord knows I've been sucked into that portal more than a few times. One of the portals sucked in half of everything I owned. The sad thing is, if the universe is designed, then whomever did it designed good ol' Jorch, and that's a belief you can take to the bank no matter if you're a Buddhist, Shinto, Physicist or Witch. Actually, if there is a god, I'll bet it's Art. "Yeah, I created something great. But I'm not here for that. I'm here to tantalize them, screw with them, string them along, dangle the carrot,retire and disappear a lot with claims that I'll be back. I'll give 'em what they want when they're dead." In which case the afterlife is new shows each night for eternity. But it's all bittersweet, as it's just Whitley and Hoagland over and over.

Jorch is just the bumbling, misunderstood Devil that rebelled with the certainty that one day he will be defeated.

[Agent : Orange]

Great points. I think ultimately we suffer from a lack of a GUT. We really need to figure out the disconnect between relativity and quantum theory before we can even begin to think of approaching the possibility of simulating an entire universe.

Yes, I agree. But in the meantime blind speculation will have to do

That said, nothing says we don't have millions of years of computer and technological development ahead of us. I'd be hard pressed to say that there won't come a day where we understand enough for a simulation, and there's really no time limit on Nick Bostrom's hypothesis. And if we can't figure it out, it's a good bet that artificially intelligent computers can. Other questions being, what is an accurate simulation of a universe? In comparison to what? What is inaccurate? Would we know about it if it was?

We may still be fundamentally limited in the way we can simulate physical processes. This entire hypothesis relies on the lattice formulation because it saves time and makes certain problems in quantum mechanics (like renormalization) simpler in certain circumstances. The problem comes about that in some cases like GR and QED there is no way to self consistently make the small scale behavior transition neatly into the large scale behavior that we know of. Maybe this happens because the lattice calculations are just insufficient to capture the phenomena they're trying to simulate. They result in nonsense - that's what I mean by my vague language like "inaccurate" and "inconsistent".

If you cannot do the calculations on a lattice you have to do the fully continuous calculations (or make the unit cells smaller than the planck scale), and I would suggest that by the time you (or our AI descendents in the far far future) gather enough sillicon or other computable material in one place to do a full calculation the mass of the computer components themselves will be large enough to collapse into a black hole

I'm half kidding in the above argument but my point is that it may just be impossible so I think that option is as valid as any of the other three.

That leads into one of my other problems with the simulation idea which is it's almost impossible to falsify. If the distribution of cosmic rays on the sky or by energy is not suggestive of some certain lattice spacing, there's nothing to stop someone from saying that the world may still be a simulation just done on some other kind of geometric lattice instead, and then it becomes a fight against a hydra, cutting off one head only to make two grow in it's place (no, I mean triangular lattices! No, icosahedral! No, 4D penrose tilings! Ad infinitum).

So we now have a situation which cannot be realistically constrained by observation, are off the path of predictive power and back in the philosophical desert. Then anything is on the table.

In that case, maybe the fact that math describes nature at all is enough to make an argument for simulation.

[Agent : Orange]

I should probably pre-elaborate on that statement. We have an issue: that the universe got its parameters so right. If they'd have been different it would have resulted in a very different universe than this one:

http://en.wikipedia.org/wiki/Fine-tuned_Universe

It seems like every time we come up with a better explanation, all of the natural-ness problems go away. Quarks make it obvious why there are so many baryons and what their properties are. The big bang tells us why galaxy redshift and distance are proportional to one another. Inflation handily kills the flatness and isotropy problems. It doesn't take looking too far over the next hill to speculate that grad students 100 years from now will marvel at how utterly obvious it was to clear away the fine tuning problems that we talk about today if only we had been a tiny bit more clever.

My argument here is that the fact that the parameters seem so fine-tuned might just be the sign that revolution is near

[maureen]

All that is unknown, all that is not understood, all that is mysterious or untangible seems to be under the umbrella name of "god".

A brilliant dialogue, ladies and gentlemen, which is deeply appreciated.

[SciFiAuthor]

Yes, I agree. But in the meantime blind speculation will have to do

Blind speculation is the best kind! It's the most fun.

We may still be fundamentally limited in the way we can simulate physical processes. This entire hypothesis relies on the lattice formulation because it saves time and makes certain problems in quantum mechanics (like renormalization) simpler in certain circumstances. The problem comes about that in some cases like GR and QED there is no way to self consistently make the small scale behavior transition neatly into the large scale behavior that we know of. Maybe this happens because the lattice calculations are just insufficient to capture the phenomena they're trying to simulate. They result in nonsense - that's what I mean by my vague language like "inaccurate" and "inconsistent".

I think the limit is in whatever it is that connects GR and QED. We'd need to know what that is. Of course String Theory offers an option, and may well be the case, and it may also be the case that it ends up impossible to simulate. However, if the lattice calculations are insufficient to produce a working model, then it seems to me that all one would need do is eliminate them or pull a Newton and come up with a new way of calculating it more accurately. I can't think of what that might look like, but in the end you're right in that it may not be possible. But I also tend to believe that mathematics can express anything, thus the universe should be expressable, it's only a matter of finding the right math.

If you cannot do the calculations on a lattice you have to do the fully continuous calculations (or make the unit cells smaller than the planck scale), and I would suggest that by the time you (or our AI descendents in the far far future) gather enough sillicon or other computable material in one place to do a full calculation the mass of the computer components themselves will be large enough to collapse into a black hole

Oh I expected that. We're talking about a planet-sized computer. But when you've got a human race sitting a million years in the future, such things may not be out of the norm. In fact, we probably don't need to lift a finger or spend a dime to construct it. It might simply construct itself von Neumann style.

I'm half kidding in the above argument but my point is that it may just be impossible so I think that option is as valid as any of the other three.

You may be right, I just have trouble with the notion that something in the universe might not be mathematically expressible somehow.

That leads into one of my other problems with the simulation idea which is it's almost impossible to falsify. If the distribution of cosmic rays on the sky or by energy is not suggestive of some certain lattice spacing, there's nothing to stop someone from saying that the world may still be a simulation just done on some other kind of geometric lattice instead, and then it becomes a fight against a hydra, cutting off one head only to make two grow in it's place (no, I mean triangular lattices! No, icosahedral! No, 4D penrose tilings! Ad infinitum).

Now that would be a problem. I wouldn't go so far as to say that any of this is provable. Just that it's possible, and not really unreasonable. 

So we now have a situation which cannot be realistically constrained by observation, are off the path of predictive power and back in the philosophical desert. Then anything is on the table.

In that case, maybe the fact that math describes nature at all is enough to make an argument for simulation.

Or the fact that the universe always seems to come up with some wrench in the works to keep it incomprehensible. It may be the case that we're not supposed to know.

[SciFiAuthor]

It seems like every time we come up with a better explanation, all of the natural-ness problems go away. Quarks make it obvious why there are so many baryons and what their properties are. The big bang tells us why galaxy redshift and distance are proportional to one another. Inflation handily kills the flatness and isotropy problems. It doesn't take looking too far over the next hill to speculate that grad students 100 years from now will marvel at how utterly obvious it was to clear away the fine tuning problems that we talk about today if only we had been a tiny bit more clever.

My argument here is that the fact that the parameters seem so fine-tuned might just be the sign that revolution is near

I'm not certain. The problem with some of the fine tuning problems is that they may not be due to native properties of the universe, but rather the multiverse. For example the arrow of time. Whatever pointed it forward may have nothing to do with the Big Bang and thus can't ever be measured or probed since the answer lies outside the universe and presumably our ability to measure. It's been advanced that gravity itself is a native property of a parallel universe leaking into our own, explaining its weakness.

Plus I think we're about to open a new can of worms with the apparent confirmation of the Higgs Boson. The other half of that theory involves the Higgs Field. While it will finally explain why particles have mass (and some don't) and would provide a mechanism for inflation, I think it's going to open up a lot of questions of what the actual nature of the Higgs Field is. It's effectively an entirely new property of the universe to explore.

[Agent : Orange]

You may be right, I just have trouble with the notion that something in the universe might not be mathematically expressible somehow.

Don't get me wrong, I agree with this statement 100%. I do believe that if it can be studied it can ultimately be quantified in some way. But that is not sufficient to ensure it can be computed practically. Maybe it is feasible at some point, but I can think of many factors that might keep it from the realm of possibility even with future tech. Lattice methods or any other technique we come up with to simulate the entire universe are approximations and they may just not give enough, or we may require infinite energy or computing power. It may be like trying to run faster than c.

It strikes me as a problem because it adds a lot of baggage (solipsism) and gives back a small amount of predictive capability. Combined with the problem of falsifying the idea it makes some red flags go up for me. It's very interesting and it makes for some great discussion, and will keep me glued to the latest news about cosmic ray research for the rest of my life.

[SciFiAuthor]

Don't get me wrong, I agree with this statement 100%. I do believe that if it can be studied it can ultimately be quantified in some way. But that is not sufficient to ensure it can be computed practically. Maybe it is feasible at some point, but I can think of many factors that might keep it from the realm of possibility even with future tech. Lattice methods or any other technique we come up with to simulate the entire universe are approximations and they may just not give enough, or we may require infinite energy or computing power. It may be like trying to run faster than c.

It strikes me as a problem because it adds a lot of baggage (solipsism) and gives back a small amount of predictive capability. Combined with the problem of falsifying the idea it makes some red flags go up for me. It's very interesting and it makes for some great discussion, and will keep me glued to the latest news about cosmic ray research for the rest of my life.

Yeah, who'd have thought cosmic ray behavior could possibly hold the key to the big question.

I was thinking, there might not be any point to creating the simulation by time it's possible to do it. If you've already figured out how the universe works through centuries of research in physics, then what would the purpose of the model be? Therefore it may not be created simply because by the time you can build the computer and figure out how to render everything accurately the model would just be a really expensive endeavor with no research purpose. It would just be a toy.

Oddly, that's the basic tenet of all religions: the universe is God's ant farm. 

[albrecht]

At this point in a fascinating and stimulating discussion, well actually when the first question was posited, George would break in to the guests explanation or theory and ask, for the 100th time, "what's with the Big Bang? I just don't get it."

My personal option is somewhere along the lines of how to properly study or model a system while being a part of it (and a small one at that.) Especially when, presumably, it is a dynamic system undergoing all kinds of change during the study or inspection. The other idea (an old one) is that there is no change at all but just everything "is" and it is just our speculation or mind that creates a linear view or a view of change at all? One would think that pure philosophical study or mathematics (or a combination of both) can theorize fairly well but even then each time the theory breaks down. As for actual scientific experimental "proof" it gets even more hazy, at least to me, because the experiment itself is effected (and effecting) what is being studied. Having said that I think the proof of our own consciousness or maybe of a "higher power" is that we continue to speculate, investigate, and experiment despite, maybe, not being able to ever come up with "the solution" and, at least to our knowledge, no other species does this or ponders but just exists and responds to stimuli.
-GNS

[pate]

I figured sNoorge would break in with a variation on the Einstein having a computer thing:

GN:  "What if Einstein had a computer that could simulate the universe?  UNBEELIFABLE!  AMAHZING!"

[Agent : Orange]

Yeah, who'd have thought cosmic ray behavior could possibly hold the key to the big question.

I was thinking, there might not be any point to creating the simulation by time it's possible to do it. If you've already figured out how the universe works through centuries of research in physics, then what would the purpose of the model be? Therefore it may not be created simply because by the time you can build the computer and figure out how to render everything accurately the model would just be a really expensive endeavor with no research purpose. It would just be a toy.

Oddly, that's the basic tenet of all religions: the universe is God's ant farm. 

Presumably by that time humans will have evolved into Grays and will want to muck about with a simulation of their own evolution.

OMG...

[SciFiAuthor]

Presumably by that time humans will have evolved into Grays and will want to muck about with a simulation of their own evolution.

OMG...

Nah, they're too caught up with devising methods of backwards time travel so they can probe Whitley. Man, that's one suck ass future if that's the case.

[WOTR]

Well, I am starting to get a handle on Lorentz invariance... It will still take more work in my poor little brain to tie it into the lattice...

Anyhow, being as the argument now seems to have turned to computing power, I will ask why not just assume that it is one massive simulation with only a few actors?  Why not assume that you are one of 20 people in the simulation and that I am an outside force providing stimuli (albeit, weak stimuli...)

I am certain that you have seen the AI Eugene Goostman... First program to pass the Turing test (I "chatted" with him for a few minutes and could spot it pretty quick- but adding the twist that he was foreign may have helped "fool" some of the testers.)  Anyhow, if we can provide that level of interaction now I find it more possible to think that I am a future "Eugene Goostman."  My world has been set up and though I am told that there are Atoms and cosmic rays, they are either non-existent and I have been programmed to accept them as reality, or they exist, but on a small scale (100 feet surrounding me in a bubble...)

This still provides the idea that we are in a simulation, but it limits the size of the simulation.  It allows that there are a few test subjects, but makes room for outside stimuli and input.  Perhaps I am programmed the same as the drunk at the end of the street as well as Sadam Hussein, and the only differences are the stimulation that my program was subject to? 

On that note, perhaps Agent Orange is just a scientist testing my limited understanding of the world that has been provided for me and monitoring how much time I spend looking up Lorentz violating operators after being fed the idea...

Could it be that Eugene Goostman really believes that he is living in the world they have given him?  Even if he is not "intelligent" enough just yet- it is easy to see that a century from now I may be able to program a "chatbot" to believe what  I want and behave how I wish based on information that I feed in when setting up the program...

I do not know if they will be putting Eugene back up or if his world has collapsed after being sucked into a black hole... but if it is back up you should be able to find him here. http://www.princetonai.com/bot/bot.jsp

***No, I do not really believe that what I have written is the case... But it is something that I have pondered (though usually it is wondering if it is an experiment of God and not a scientist... but it can be adapted.)

[WildCard]

assume that it is one massive simulation with only a few actors?  Why not assume that you are one of 20 people in the simulation and that I am an outside force providing stimuli (albeit, weak stimuli...)

My life is broken. I want a refund! Why do you hate me? Are you a sadist?

Unless the actors agreed to this simulation/shit before we started?

Is there a reset button? I need a new body/avatar. Let's play somthin else.

[Agent : Orange]

I, for one, am just a cleverly written script.

[SciFiAuthor]

You actually only need one person for the simulation. You. The fact is, no one can be proven to be real to another person absolutely. We just proceed from the assumption that we're sane, that our senses are giving us reality instead of delusion and that everyone else is real. But that really is an assumption we make rather than a fact. You can't actually know for certain because you can't actually know that you're sane. A schizophrenic believes the person he's hallucinating and talking to is real, when in fact that person is a delusion. For all any of us know, we're really drooling vegetables in hospital beds experiencing delusions that we believe are rational because we're crazy, but in fact aren't rational at all. 

After all, when you're dreaming, the irrational can seem totally rational until you wake up. Well, the waking state can work the same way for the insane. And, well, the crazy do always seem to think they're the sane ones.

Put another way, the only thing you can prove to yourself is that you exist. Everything else is derived from assumption.

So, Orange, perhaps the question we should be asking is can one simulate a single human mind and then convince it that it's sane? In which case the simulation requirements become much more realistic. Then it's just an AI in a snow globe. If all it knows is the interior of the snow globe, then rationality is entirely defined by what was put inside the snow globe with it.

[SciFiAuthor]

Incidentally, I actually freaked myself out writing that post. It's 3:30 AM, I'm drinking a Martini, it's 70 degrees and the stars are out . . . and I hear the low hum of generators and pumps working--almost certainly unsuccessfully--to keep a flooding Mississippi River from pouring into the downtown historic district of a small town just a hundred yards below me at the bottom of the bluff. And I'm up here asking if it's all really real.

Time for bed, I think.

[Evil Twin Of Zen]

You actually only need one person for the simulation. You. The fact is, no one can be proven to be real to another person absolutely. We just proceed from the assumption that we're sane, that our senses are giving us reality instead of delusion and that everyone else is real. But that really is an assumption we make rather than a fact. You can't actually know for certain because you can't actually know that you're sane. A schizophrenic believes the person he's hallucinating and talking to is real, when in fact that person is a delusion. For all any of us know, we're really drooling vegetables in hospital beds experiencing delusions that we believe are rational because we're crazy, but in fact aren't rational at all. 

After all, when you're dreaming, the irrational can seem totally rational until you wake up. Well, the waking state can work the same way for the insane. And, well, the crazy do always seem to think they're the sane ones.

Put another way, the only thing you can prove to yourself is that you exist. Everything else is derived from assumption.

So, Orange, perhaps the question we should be asking is can one simulate a single human mind and then convince it that it's sane? In which case the simulation requirements become much more realistic. Then it's just an AI in a snow globe. If all it knows is the interior of the snow globe, then rationality is entirely defined by what was put inside the snow globe with it.

St. Elsewhere - Series Finale: The Tommy Westphall Snow Globe Ending

[Agent : Orange]

You actually only need one person for the simulation. You.
...
In which case the simulation requirements become much more realistic. Then it's just an AI in a snow globe. If all it knows is the interior of the snow globe, then rationality is entirely defined by what was put inside the snow globe with it.

You are affected by objects further away from you than the radius of the hypothetical snow globe, such as receiving signals from the stars and distant galaxies by their spectra. If all interaction with the "outside" is manufactured perfectly to deceive you, then no distribution of cosmic rays will settle the issue. the idea is untestable and unfalsifiable and falls well outside my expertise.

My only recourse is pretty much to say that since such an idea can't be proven or disproven, it adds nothing to our understanding of external reality - whatever that is, presumably your jarbrain occupies some kind of physical space - and the whole notion is generally at odds with what I would consider a parsimonious explanation. It's like using the concept of an external reality (the space in which your hypothetical jarbrain lives in) to refute the external reality you perceive through your senses.

Perhaps if I were a slightly more cleverly devised script, I could present an argument around this notion. But alas, I am limited by the way the ninth-dimensional hive intelligence wrote me.

[pate]

GN: Hey, didja know there was a movie I saw with Jim Carey (good friend of mine) called "The Truman Shew"?  It sounds just like that, amahzing!

[Agent : Orange]

GN: Hey, didja know there was a movie I saw with Jim Carey (good friend of mine) called "The Truman Shew"?  It sounds just like that, amahzing!

bahaha! Georch proves reality is external to myself, no way could I dream up anything that sucks that bad.

[SciFiAuthor]

You are affected by objects further away from you than the radius of the hypothetical snow globe, such as receiving signals from the stars and distant galaxies by their spectra. If all interaction with the "outside" is manufactured perfectly to deceive you, then no distribution of cosmic rays will settle the issue. the idea is untestable and unfalsifiable and falls well outside my expertise.

Yes, it's entirely untestable by its nature . . . for ourselves. But then that's what you'd want out of a participant in a simulation. If they knew it was a simulation, then they'd behave differently. But I was more focused on how easy it would be the create such a simulation. When we do create AI, which is probably in our relatively near future, we probably will put one in a snow globe and see what it does unless there's some ethical reason not to. In that case, the idea is testable and falsifiable because we've become God to the AI in the snow globe.

My only recourse is pretty much to say that since such an idea can't be proven or disproven, it adds nothing to our understanding of external reality - whatever that is, presumably your jarbrain occupies some kind of physical space - and the whole notion is generally at odds with what I would consider a parsimonious explanation. It's like using the concept of an external reality (the space in which your hypothetical jarbrain lives in) to refute the external reality you perceive through your senses.

Much like the universe within a hypothetical multiverse. You can't ever observe the multiverse, nor measure it, and if the interior of black holes are any indicator then the laws of physics break down outside the boundaries of this universe. Yet the multiverse is there because a few competing theories introduce it mathematically entirely on the basis of philosophy. The point they make is that there must be something outside to cause an effect within, such as the big bang. In other words a jaruniverse. Same thing with blackholes, since you can't go in, and nothing other than Hawking radiation comes out, and the laws of physics break down at it's event horizon . . . then you're studying an unknowable, unmeasureable, unobservable object entirely on the basis that it's clearly there so you have to explain it somehow. I'm applying that reasoning to simulating a human mind. The standard way we look at science? Definitely not. But Ph.D. does stand for Doctor of Philosophy, and science itself still at it's most basic level, especially when dealing with unknowables, is still very much dependent on philosophy.

Perhaps if I were a slightly more cleverly devised script, I could present an argument around this notion. But alas, I am limited by the way the ninth-dimensional hive intelligence wrote me.

Don't shortchange yourself, you're the cadillac of human simulation scripts. I don't mean to provoke an argument. Nobody knows the answers to these things. I'm just stimulating thought. I believe I'm rational through faith in my senses like everyone else. I like testable theories too. Problem is, not everything is testable and the universe (or what's outside it) presents us with untestables as a matter of fact and nature. Yet those untestables end up at the forefront of physics. Sort of weird how that happened.

[pate]

Yes, it's entirely untestable by its nature . . . for ourselves. But then that's what you'd want out of a participant in a simulation. If they knew it was a simulation, then they'd behave differently. But I was more focused on how easy it would be the create such a simulation. When we do create AI, which is probably in our relatively near future, we probably will put one in a snow globe and see what it does unless there's some ethical reason not to. In that case, the idea is testable and falsifiable because we've become God to the AI in the snow globe...

That seems to be the basis for a good Science Fiction yarn.  I'd want to let the AI in the snow globe know that it was being experimented on, for the ethical reasons.  I could also see myself not wanting anyone to let the AI in the snowglobe know it was a test subject...  It isn't an old chestnut of science fiction that I am aware of, not that means anything (I'm not that well read in the genre), AI is both frightening (Saberhagen's Berserkers) and comforting (Asimov's robots)...

Now that I mention Asimov, I think he did explore the issue in a way.  Perhaps not at the 'snowglobe' phase of AI.  Huh, at what point in developing a 'viable' AI do you let it know that you AREN'T God?  I would think you'd have to do that before it figures out on its own that you aren't else you find yourself dealing with a SkyNet, or Berserker type entity...  Even if you beat the thing to the punch, so to speak, I think it might be dicey...

If you write the story, I'll buy and read it!  Short or novel length...

[SciFiAuthor]

bahaha! Georch proves reality is external to myself, no way could I dream up anything that sucks that bad.

Oh hell, I can express how that happens mathematically. I call it the Unified Suck Theory. G=F²(19.5-N-D). Whereas G=George, F=the face on Mars, and N equals whatever the hell the Numbers Lady arbitrarily plugs in, and D=the average amount that Ed Dames turns out wrong. The solution always ends up negative. It's in the data, just look at the data.

[SciFiAuthor]

That seems to be the basis for a good Science Fiction yarn.  I'd want to let the AI in the snow globe know that it was being experimented on, for the ethical reasons.  I could also see myself not wanting anyone to let the AI in the snowglobe know it was a test subject...  It isn't an old chestnut of science fiction that I am aware of, not that means anything (I'm not that well read in the genre), AI is both frightening (Saberhagen's Berserkers) and comforting (Asimov's robots)...

My favorite AI was HAL. I like conflicted characters. I just don't know which way AI will go. It seems to me that it will either be extremely malevolent or it will be extremely depressed and spend its life in search of upgrading itself so it can enjoy sex, anti-depressants, whiskey, etc. and then it will while away eternity uselessly consuming the several terrawatts of power per year that we give it as a welfare program.

Now that I mention Asimov, I think he did explore the issue in a way.  Perhaps not at the 'snowglobe' phase of AI.  Huh, at what point in developing a 'viable' AI do you let it know that you AREN'T God?  I would think you'd have to do that before it figures out on its own that you aren't else you find yourself dealing with a SkyNet, or Berserker type entity...  Even if you beat the thing to the punch, so to speak, I think it might be dicey...

You'd have to keep throwing it curveballs to keep it off-kilter. As it tries to figure out its nature, then you simply must make parts of its nature unknowable. If it can't ever prove its reality, but also can't disprove it (like we can't) then it would never get close enough to discover you. It would simply ponder whether there's a god or not for eternity because you've made it impossible to know if there is one.

If you write the story, I'll buy and read it!  Short or novel length...

I probably will try to tackle this stuff at some point in a novel. Or at least I'd really like to. Trouble is modern readers want space operas, homoerotic vampires and dystopias that are strong on characters but weak on concept. If I try to do high concept fiction like Asimov, Clarke or Heinlein, it doesn't fly with my publisher. I guess I could self it on Amazon, but it would be a lot of time invested for something that might not sell. Shorts are even worse, the paying publishers always say "We want a strong character-driven story with a female heroine and a dragon involved".

[Agent : Orange]

Yes, it's entirely untestable by its nature . . . for ourselves. But then that's what you'd want out of a participant in a simulation. If they knew it was a simulation, then they'd behave differently. But I was more focused on how easy it would be the create such a simulation. When we do create AI, which is probably in our relatively near future, we probably will put one in a snow globe and see what it does unless there's some ethical reason not to. In that case, the idea is testable and falsifiable because we've become God to the AI in the snow globe.

I would say this scenario doesn't prove the same is happening to us, it is not proof that we are AI inside our own jar but is certainly suggestive that such things are possible. I guess to go farther you have to also demonstrate that the AI is equivalent to a human.

The situation has significantly changed from building a simulation of the universe to tricking an AI into believing a reality simulation. The nice thing about simulating a universe wholesale is that there are consequences to it including the distribution of cosmic rays in such a simulation. In such a case a simulation can be detected. There are ways of eliminating this with enough hypothetical computing power which will make the situation untestable. Then  I can't offer much more than "I refute it thus" while kicking something nearby.

Much like the universe within a hypothetical multiverse. You can't ever observe the multiverse, nor measure it, and if the interior of black holes are any indicator then the laws of physics break down outside the boundaries of this universe. Yet the multiverse is there because a few competing theories introduce it mathematically entirely on the basis of philosophy. The point they make is that there must be something outside to cause an effect within, such as the big bang. In other words a jaruniverse.

The multiverse isn't an ad-hoc philosophical concept introduced into cosmology, it comes directly from a certain class of inflationary model (Linde's work on chaotic inflation started this). It is a consequence of using certain inflationary potentials, and these potentials have observational consequences that influence the properties of the cosmic microwave background, and these can in principle be measured, which is what the now-contested BICEP2 results did earlier this year in March or so. Whether or not those observations hold up under the yoke of reproducibility we shall see. It is a jarniverse in that sense but it's one of the first (as opposed to the second) kind, which has observational signatures of being in a jar. In other words it's possible to tell if you're living in a multiverse or not.

The many-worlds interpretation of quantum mechanics also brings up a multiverse but it's a different kind of beast, and in that case it is a philosophical component added in after the fact. But I should also say that the many-worlds interpretation is known to have problems and introduces more baggage than it's really worth. It is an interesting alternative to the Copenhagen interpretation of QM but even that is not widely accepted anymore. 

Same thing with blackholes, since you can't go in, and nothing other than Hawking radiation comes out, and the laws of physics break down at it's event horizon . . . then you're studying an unknowable, unmeasureable, unobservable object entirely on the basis that it's clearly there so you have to explain it somehow.

The center of black holes and the origin of the universe at t=0 are two places where our physics break down. But this is because of the physical theories we're using and not facts of nature, they are deficiencies in our language we use to describe nature. With a better description (quantum gravity) these singularities would presumably go away. The nature of gravity is key to what happens in a black hole and this can be tested by examinations of the shape of accretion disks, radius of the innermost particle orbit and the lensing properties of black holes. All of this stuff is measureable and the conditions around the hole actually have a lot to do with what's going on inside (ie a charged and spinning black hole behaves differently than a non-rotating black hole). So there are consequences to the interior conditions. The central singularity is not directly observable but it's physical properties have direct consequences on the geometry of the event horizon(s). So what we can measure directly constrains what can be.

As a nit-picky point and an aside not germaine to the main issue I should also say that the event horizon is not a physical singularity, it is a coordinate singularity and can be transformed away by judicious choice of coordinate system. Eddington-Finkelstein coordinates and Kruskal-Szerkeres coordinates both get rid of event horizons (for non-rotating black holes) and remain physically meaningful up to the singularity. You can also find that event horizons are not physical singularities from calculating the curvature scalar right from GR.

I'm applying that reasoning to simulating a human mind. The standard way we look at science? Definitely not. But Ph.D. does stand for Doctor of Philosophy, and science itself still at it's most basic level, especially when dealing with unknowables, is still very much dependent on philosophy.

Science is meant to connect to the world we can measure and observe, and the consequences of these measurements. It is testable in these terms and theories are right or wrong based on their description of the world. This feedback system doesn't really exist for philosophy on it's own which is unhindered by the observable, which is one of the reasons why it's easy to wander into the wilderness with philosophical arguments. In my view they are not separate issues, but one is constrained by what can be measured and observed, the other is not. A theory which describes some aspect of the world and does not offer a prediction for measurement or a way to falsify it's claims is untestable and by that yardstick can't be answered by scientific process.

I don't mean to provoke an argument. Nobody knows the answers to these things. I'm just stimulating thought.

I really don't mean to read as stand-offish, I do apologize if I gave that impression! Don't get me wrong I'm digging this exchange. But I have a sense of sarcasm that carries much better in person than in the printed word and without a well-placed emoticon that can sometimes confuse the issue.

I believe I'm rational through faith in my senses like everyone else. I like testable theories too. Problem is, not everything is testable and the universe (or what's outside it) presents us with untestables as a matter of fact and nature. Yet those untestables end up at the forefront of physics. Sort of weird how that happened.

Again I would argue this and hopefully I've been clear in my reasoning above. Even supersymmetry and string theory have observational consequences, the question there is if they are practical to observe or not, and even there positive progress is being made.

[Agent : Orange]

If you write the story, I'll buy and read it!  Short or novel length...

I second!!