Are We Really Real?
May 21, 2007 3:53 PM Subscribe

Are You Living In A Computer Simulation? The Simulation Argument is a regularly-updated site, organized by Nick Bostrom, examining the hypothesis that we are currently living in an "ancestor simulation" run by a future "post-human" society.
posted by amyms (163 comments total) 29 users marked this as a favorite

I refute it thus...

(kicks virtual rock)

d'oh
posted by SBMike at 3:55 PM on May 21, 2007 [1 favorite]

Of all the implications of the Simulation Argument, my favorite is that all we ever have is cybersex.
posted by cortex at 3:57 PM on May 21, 2007 [6 favorites]

Bill Hicks: It's just a ride.
posted by hortense at 3:59 PM on May 21, 2007 [1 favorite]

His definition of what is a "computer simulation" is pretty board. The whole goal of science is to reduce the world into something understandable that follows rules. In essence we are pushing to find a model that we can simulate in some fashion. As we have more and more success in this endeavor it will become easier and easier to imagine that some powerful "computer" could theoretically simulate it.

The argument becomes circular though in that if we are a simulation of some advanced beings, are those advanced beings actually being simulated by even more advanced beings? What distinguishes these advanced beings from are normal definition of a all knowing and all powerful God/Creator?

Occam's Razor needs to be applied at some point, because there are a lot of these types of unanswerable questions that one can waste their time on, many of which have more impact on our real/simulated lives that this one.
posted by bhouston at 4:01 PM on May 21, 2007

I heard that Ben Affleck wanted to "re-cut" Good Will Hunting so the entire movie would be the same except that he would have a voiceover at the end: "THIS WHOLE MOVIE HAPPENED IN THE MATRIX!!!1!"
posted by dogwalker at 4:01 PM on May 21, 2007 [4 favorites]

Philosophical masturbation.
posted by Mr.Encyclopedia at 4:03 PM on May 21, 2007 [4 favorites]

Director of the "Future of Humanity Institute" at Oxford? Whose side is he on?
posted by tepidmonkey at 4:03 PM on May 21, 2007

(Damn, why does Digg and Reddit allow for comment editings but not MeFi?)
posted by bhouston at 4:03 PM on May 21, 2007

tepidmonkey: Bostrom is a "transhumanist" who I believe wants to "upload" himself at some point or something like that.
posted by bhouston at 4:04 PM on May 21, 2007

For those who want to get to the meat of it, his argument is thus:

1. Let's admit the possibillity that in the future, we can simulate the past.

2. If we can create one simulation, we can create countless simulations.

3. Since, there are countless simulations but only one "original" what are the chances we are living in that one original reality? Pretty much zero.

Thus, we are living in a simulation.
posted by vacapinta at 4:06 PM on May 21, 2007 [1 favorite]

cortex said: "Of all the implications of the Simulation Argument, my favorite is that all we ever have is cybersex."

But, that hypothesis must be false, because cortex is here moderating Metafilter... Unless he's moderating Metafilter while having cybersex... Yikes, these hypotheses are always so complicated.
posted by amyms at 4:07 PM on May 21, 2007

But... he'd just be uploading himself to a Virtual Machine!
posted by Artw at 4:07 PM on May 21, 2007

So Nick Bostrom is dhoyt?
posted by orthogonality at 4:10 PM on May 21, 2007

Oh, I routinely start my day with the Reginald Barclay Method of verifying reality.

"Computer, End program."
[[Look around, nothing changes, start my day satisfied]]
posted by John Kenneth Fisher at 4:13 PM on May 21, 2007 [14 favorites]

Dude's ripping me off. I totally wrote that story in 10th grade English.
posted by redhanrahan at 4:14 PM on May 21, 2007 [2 favorites]

Actually, this is all just happening in a reality bubble between pages 13 and 14 of Philip K. Dick's "Ubik."
posted by Afroblanco at 4:18 PM on May 21, 2007 [4 favorites]

It seems to me a more interesting point is that, though the argument doesn't REQUIRE it, it implies that not only would we be living in a simulation, but that we likely are a simulation.
posted by John Kenneth Fisher at 4:18 PM on May 21, 2007

* searches intarwebs for cheat codes
posted by UbuRoivas at 4:23 PM on May 21, 2007 [4 favorites]

This would make a great movie.
posted by gc at 4:24 PM on May 21, 2007

But like, whoa, what if, like, we're all living, like, in the matrix, and not even knowing it? Whoa. Here, pass that bong over this way...
posted by UbuRoivas at 4:26 PM on May 21, 2007

Are You Living In A Computer Simulation?

You mean besides Metafilter?
posted by jonmc at 4:28 PM on May 21, 2007 [2 favorites]

But like, whoa, what if, like, we're all living, like, in the matrix, and not even knowing it? Whoa. Here, pass that bong over this way...

Pretty much. It's a fun thing to think about to kill some time, in the same sense that it's fun to debate the mental sanity of Batman, or when and if slavery would have been outlawed in a succesful Confederate States of America, or the effectiveness of photon torpedos and phasers on a SuperStar Destroyer... but in the grand scheme of things, yeah, maybe we are. I still have to pay my bills though. I'm not criticizing amyms or Bostrom, cause it IS fun to think about, and I hope to see a good discussion of the theory here, but I've met people who get a little too worked up on these questions, and I've always felt that they should spend more time worrying about bettering where we are, and we deal with people in the here and now, and less thinking about that which may or may not exist. If that makes sense.
posted by John Kenneth Fisher at 4:34 PM on May 21, 2007 [4 favorites]

(super-crazy, before 1900, and very effective at impact point, but the sheer size differences would make it very difficult to do any real damage.)
posted by John Kenneth Fisher at 4:36 PM on May 21, 2007 [3 favorites]

Ok, what's the value to an advanced civilization of running an ancestor simulation, when most answers could be had far more cheaply (quickly, using less computing power, whatever metric of "cheaply" you want to use) with a less detailed simulation.

If it's purely sentiment ("let's let our ancestors 'live' again"), why include starving babies and ethnic cleansing and the pains of childbirth?

Or is everyone simulated? I mean, I know I'm real and have internal states (or at least I can apparently recall them), but all of you could be Eliza and Racter and Hal for all I know. Perhaps we're characters in a novel, and most of us are spear-carriers and the anonymous crowd murmuring "rhubard, rhubard".

I can see how we could, to some degree of fidelity, simulate, say Jefferson, but how do we simulate the 27th dirt farmer down the road from Monticello? At least we have his name, his occupation, the dates of his christening and burial and some account of taxes he paid.

But what about the Bangladeshi refugee dead a hour after his birth? (Especially now that the Catholics have abolished Limbo!) Do we even know he existed? Will the post-human society simulate (sumulate?) some statistical average of dying Bangladeshi infants?

Any decently faithful simulation requires knowledge of the past we not only don't have, but have no clue how to discover: what were the thoughts, desires, interactions of some, what 50 billion people past and present?

If that faithfulness is unattainable, what's the value of the detail that the simulation would have to attain for us to mistake it for real?
posted by orthogonality at 4:37 PM on May 21, 2007 [2 favorites]

Somebody's got waaaaay too much time on their hands.
posted by The Light Fantastic at 4:37 PM on May 21, 2007 [1 favorite]

The way I understand it, he's not discussing reality, but rather (the reasonableness of possible) claims about reality. The former would indeed be mere "philosophical masturbation" (not that there's anything wrong with that per se) but the latter investigation does have some relevance outside of the merely speculative. I find it an interesting argument, and not one that's easily dismissed if you accept his basic premises (which, of course, is a big if).
posted by treepour at 4:39 PM on May 21, 2007

I have finally been educated stupid.
posted by WolfDaddy at 4:40 PM on May 21, 2007 [5 favorites]

Any decently faithful simulation requires knowledge of the past we not only don't have, but have no clue how to discover: what were the thoughts, desires, interactions of some, what 50 billion people past and present?

Who says it has to be faithful? Maybe they loaded it up with an as accurate as they could get 1700, or cavemen, or amino acids, and just let it simulate. Maybe we look nothing like real human history, but it's a study of parallel development. Start 50 identical simulations and let them develop for a million (simulated?) years.

Another point, computing power may well be cheap enough that they might as well add all the detail that we see. And do we KNOW how accurately we reflect this theoretical reality? So, we say we feel love. We feel hate. We feel hunger. We feel pain. Maybe we just feel rough estimations of those real feelings and emotions. how would we know? Maybe they haven't been able to simulate the key emotion of BlarkVol at all, and they just skipped adding it to us at all.
posted by John Kenneth Fisher at 4:42 PM on May 21, 2007

Perhaps these simulations are really easy to run, given adequate processing power and properly designed models. I wouldn't be terribly shocked to learn that we're all part of Grand Theft Auto: The Naughty Oughties on some kid's Playstation 13.
posted by mullingitover at 4:46 PM on May 21, 2007 [3 favorites]

Just as long as it's not in Windows
posted by niles at 4:46 PM on May 21, 2007

You are living in a simulation. I'm just working in it, for an internship. You humans, can you figure out the meaning of life in the next thousand years of simulation time? I want to leave the office sooner.

I swear I'll personally delete from the mainframe the first one that tries a Douglas Adams joke.
posted by darkripper at 4:49 PM on May 21, 2007 [2 favorites]

Does this mean that Second Life is real?
posted by strangeleftydoublethink at 4:50 PM on May 21, 2007 [1 favorite]

John Kenneth Fisher said: I'm not criticizing amyms or Bostrom, cause it IS fun to think about, and I hope to see a good discussion of the theory here, but I've met people who get a little too worked up on these questions, and I've always felt that they should spend more time worrying about bettering where we are, and we deal with people in the here and now, and less thinking about that which may or may not exist. If that makes sense.

Of course it makes sense, and I totally agree. I posted it because, like you said, it IS fun and interesting to think about, and to read other people's thoughts on it.
posted by amyms at 4:52 PM on May 21, 2007

The key point of the argument is that if you can believe our civilization will eventually be able to make a virtual universe simulation environment that is indistinguishable from what you experience now (simulation of people and physical laws), then you can think of a recursive series of those simulations. The first simulated universe makes another simulated universe inside it and so on. Then, barring any other evidence, you (and I) exist in one of these arbitrary universes. Since there seems to be an awful lot of them due to the recursive property, there is a high probability that the universe you actually exist in is one of the simulated ones, rather than the original.
It seems that the argument is sound, but I doubt if such a universe simulation is possible.
posted by demiurge at 4:55 PM on May 21, 2007 [2 favorites]

If this were a computer simulation it would mean everything would be quantized, not just energy. Position and time would also be quantized.

There are inherent and fundamental differences between how an analog universe presents and a digital universe would present, and if we were in a digital universe we'd know it by now. No amount of digits of precision in a digital simulation can make it act the same as an analog universe. Everything we see and have observed is consistent with an analog universe -- and inconsistent with a digital universe.

In particular, if this was a digital simulation, then Relativity would not be correct. There would be an absolute frame of reference -- the one the digital simulation was using. And because of that, the things we observe wouldn't be as we observe them.
posted by Steven C. Den Beste at 4:56 PM on May 21, 2007 [6 favorites]

I think he vastly underestimates the computational difficulties associated with simulation, at least on the scale being suggested.

But it's ultimately impossible to know if you are "real," or just a "brain in a vat" (or shadows on a cave wall, or projections of a magic lantern, or in the Matrix).

If you really can't discern the difference (assuming such a simulation is possible -- I doubt that it is) -- what the hell does it matter? It doesn't.
posted by teece at 4:57 PM on May 21, 2007 [2 favorites]

I'm still trying to prove that I'm not in a constantly accelerating windowless elevator.
posted by StickyCarpet at 4:57 PM on May 21, 2007 [2 favorites]

Position and time would also be quantized.

Whether that statement is true or false in this universe is still an open question.
posted by teece at 4:58 PM on May 21, 2007 [3 favorites]

I think he vastly underestimates the computational difficulties associated with simulation, at least on the scale being suggested.

But that presumes that the complexity of our universe is no less than that of a universe in which this simulation would be running. Which, from our perspective, seems reasonable, but what would we know? We're making game (if crude) attempts at creating artificial intelligence—when and if we do manage it, is it plausible to presume that it might have an understanding of its existence as being in a universe much simpler than our own? Why not?
posted by cortex at 5:01 PM on May 21, 2007 [2 favorites]

I think he vastly underestimates the computational difficulties associated with simulation, at least on the scale being suggested.

For starters: Military Simulation Techniques & Technologies.
posted by SweetJesus at 5:07 PM on May 21, 2007

But that presumes that the complexity of our universe is no less than that of a universe in which this simulation would be running.

I think that is a very interesting point, cortex, but I'm not sure my statement presumes that. Assume that some other intelligence from a vastly more complicated universe simulates us. Why would their vastly more complicated reality make it any easier to simulate our already pretty complicated reality? Maybe it would -- maybe it wouldn't.

We can't answer these questions; ultimately I think they are the exact same questions religions ask. The simulators and God would be one and the same. Which is not to dismiss such ideas -- I am fascinated by them -- by they are not a fashion of idea we have any method of proving or disproving.

Would we be simulated with a digital computer? An analog computer? Is this universe the computer? You really can go as far down the rabbit hole as you want with this stuff. Objective reality and truth are ultimately unknowable, if I remember my philosophy correctly.
posted by teece at 5:08 PM on May 21, 2007 [1 favorite]

Ok, what's the value to an advanced civilization of running an ancestor simulation, when most answers could be had far more cheaply (quickly, using less computing power, whatever metric of "cheaply" you want to use) with a less detailed simulation.

Who says it's about "answers"? It could just be The Sims + AI, run on a massively powerful computer, and might apply itself to all kinds of entertainment purposes, not only research. With enough processing power, it might be like the Hindu idea that one hundred thousand billion earth years is one second in the mind of Brahma. If you accept that processing power could increase to that extent, then one wouldn't really need any particular reason to set up a simulated universe, especially if its various possibilities could be computed in mere megahertz.

As for the Bangladeshi babies, hell, when I messed around briefly with SimEarth, I inadvertently set off all kinds of cyclones & tsunamis. Shit happens.

Apart from that, who says that there is only one simulated universe? Why not have branching, fractal-style parallel simulated universes, with minutely changing parameters, such that a flap of a butterfly's wing causes a polio germ to infect & kill a Bangladeshi baby, whereas in a parallel branch universe the butterfly flapped a hundredth of a second later, the polio germ missed the baby, and she grew up to be the President of the United States of Eurasia?
posted by UbuRoivas at 5:09 PM on May 21, 2007 [1 favorite]

Somebody's got waaaaay too much time on their hands.

What if everyone had nothing but free time on their hands, all the time?

Sorry if I just blew your mind.
posted by _sirmissalot_ at 5:10 PM on May 21, 2007 [4 favorites]

Also: This thread is itself a replica.
posted by vacapinta at 5:12 PM on May 21, 2007

I refuse to believe anyone would waste computer time simulating me. Though, I suppose, reality could be just one big distributed screen saver. Maybe it's call Reality@Home.
posted by chairface at 5:15 PM on May 21, 2007

Also: This thread is itself a replica.

No, it's not. It's a simulacrum.
posted by UbuRoivas at 5:28 PM on May 21, 2007

How is "simulated reality" necessarily any less real than "real rality" (whatever that is)?
posted by ZenMasterThis at 5:29 PM on May 21, 2007 [1 favorite]

Two arguments against this (insert your favorite pejorative) idea.

-If it is instinguishable from reality; then it is no longer a simulation as simulations are not real Therefore you are not living in a simulation. It is basic math.

-A hypothesis must be testable; therefore Simulism isn't a one. This is no more than an aethesists version of intelligent design. The dinosaurs are just there to throw us off the scent of our simulator overlords. If the simulation isn't run by an aethesist then that's pretty f'd up. What if a god(s) messes with the sim?
posted by humanfont at 5:35 PM on May 21, 2007 [1 favorite]

A simulation doesn't have to run on a digital machine, and computers don't have to be digital. In Hitchhiker's Guide, the easiest way to make a powerful problem solving computer turned out to be "use a planet". But the planet was real, not a simulation. The planet and all the life on it were the computer and the software. I suppose that when you have access to horsepower like that, you can, like, totally run Doom on it.
posted by Area Control at 5:40 PM on May 21, 2007

Ok, what's the value to an advanced civilization of running an ancestor simulation, when most answers could be had far more cheaply (quickly, using less computing power, whatever metric of "cheaply" you want to use) with a less detailed simulation.

Hey dude, we are the less detailed simulation.

[logs off MeFi, watches American Idol, thanks Bush and Jeebus for having such a great reality]
posted by Avenger at 5:45 PM on May 21, 2007 [3 favorites]

A hypothesis must be testable; therefore Simulism isn't a one.

I don't think its masquerading as an empirical hypothesis. And, from my understanding at a least, it's not so much "Hey, maybe we're living in a simulation," but rather "Hey, maybe its more rational, given certain presuppositions, to assume that we are." I think the claim is more epistemological than ontological.
posted by treepour at 5:46 PM on May 21, 2007 [1 favorite]

"And man said, Let us make God in our image, after our likeness..."
posted by sergeant sandwich at 5:55 PM on May 21, 2007

Like I said: philosophical masturbation.
posted by Mr.Encyclopedia at 5:59 PM on May 21, 2007 [1 favorite]

Get real. Bill Hicks dresses like a dorkfart.
posted by gorgor_balabala at 6:09 PM on May 21, 2007

I actually live in a pop-up book.

*pops up*
posted by jonmc at 6:10 PM on May 21, 2007 [6 favorites]

In Iain M. Bank's The Algebraist, there's a religion specifically devoted to the idea that we live in a simulated universe. The devotees believe that, once enough of the population believes, the folks running the simulation will do something about it.

Then you've got the Tipler thing. And the Andew J. Wilson story "Under The Bright and Hollow Sky," wherein the question is asked, "What if the people running the simulation are particularly cruel?"
posted by adipocere at 6:23 PM on May 21, 2007

HOW DO YOU FEEL ABOUT LIVING IN A COMPUTER SIMULATION?
posted by Cyrano at 6:24 PM on May 21, 2007

Like I said: philosophical masturbation.

You should try it. It feels good and relieves sexual tension.
posted by treepour at 6:25 PM on May 21, 2007 [1 favorite]

Cogito ergo sum

Stand back, Mr. E, here comes the money shot.
posted by Bokononist at 6:27 PM on May 21, 2007

A hypothesis must be testable; therefore Simulism isn't a one

"Simulism" as described here is not testable, but then it isn't presented as a hypothesis to begin with.

In differential geometry, it is possible to derive extrinsic and intrinsic properties from a manifold.

In laymen's terms, you can be an ant on the surface of an orange, and if you're smart enough and have the right rulers, you can figure out things about the surface you're on. These intrinsic properties don't required that you see the "outside" of the manifold.

Instead, it might be possible to learn that we're "inside" a simulation through its measurably intrinsic properties. We may just not yet have the language for describing those properties.
posted by Blazecock Pileon at 6:36 PM on May 21, 2007 [2 favorites]

Decartes and his demon are crying, wishing the Matrix wasn't seen as the starting point of the world-is-fake-phenomena.
posted by geoff. at 6:37 PM on May 21, 2007 [2 favorites]

The article makes one major assumption: that computers (or programs or the two together, if you prefer) will one day be able to think the way we do. That is that they will have intentional consciousness.

That is at best a questionable assumption; many would argue that it is obviously false.

Yes, that is quite an assumption. However, I'm more or less a pantheist who believes that consciousness and matter are two sides of the same coin, so it's not hard for me to grant him that premise.
posted by treepour at 6:39 PM on May 21, 2007

This is a new variant of the 'brain vats on Jupiter' aka Matrix argument. One of the more entertaining twists I've seen was an SF novel in which super-advanced civilizations were attempting to prove or disprove this theory by running "timing-channel attacks" on the substrate of the Universe.... the galactic equivalent of trying to break out of your virtual machine. :) Kind of an amusing one-off.

Might have been by Cory Doctorow, now that I'm thinking about it.
posted by Malor at 6:40 PM on May 21, 2007 [1 favorite]

You can't argue against this. You have to assume that the simulators will take any means necessary in order to 'contain' the simulation and therefore all data is compromised. The strongest "argument" against this is that one would hope that the super-powerful human beings of the future who can simulate universes and have overcome all constraints on intelligent existence (that is, they're almost certainly living inside a simulation themselves) ... really have something better to do on their Saturday nights than watch us. Cuz if that's not the case the future fucking sucks.
posted by nixerman at 6:43 PM on May 21, 2007 [3 favorites]

Malor, no, that was by our very own Charles Stross, at the very end of Accelerando.
posted by adipocere at 6:43 PM on May 21, 2007 [1 favorite]

Like I said: philosophical masturbation.

Careful! You will go blind!
posted by voltairemodern at 6:44 PM on May 21, 2007

Occam's Razor needs to be applied at some point

Bollocks. What's the simplest explanation: solipsism, or the common-sense view of reality?

You'd best replace that razor with hume's fork.
posted by dreamsign at 6:45 PM on May 21, 2007 [1 favorite]

This is a new variant of the 'brain vats on Jupiter' aka Matrix argument. ...

At this point, I don't think the argument is very new. I recall reading it in the other MeFi thread, the one vacapinta linked above, and I believe even then it was a few years old.
posted by voltairemodern at 6:47 PM on May 21, 2007

Dude, the whole universe could just be like one tiny atom in the fingernail of this giant living in a whole 'nother super mega-universe. Man.
posted by Mister_A at 6:47 PM on May 21, 2007

Sounds like civilized debate of this paper is enabling a polite Messiah Complex. If we are living within a simulation and Nick The One Bostrom believes he can find a way to leave it then that would make him mighty special indeed.
posted by CynicalKnight at 6:52 PM on May 21, 2007 [1 favorite]

I can move stuff with my mind, through the agency of my hands. I AM TEH CHOSENED ONE BITCHAZZ!!11!1
posted by Mister_A at 6:55 PM on May 21, 2007

adipocere: oops, yes, you're right. I think he mentions it in the book at least twice, actually, once about midway through. But I'd lost the name...thanks for the reminder. :)
posted by Malor at 6:57 PM on May 21, 2007

Hey dude, we are the less detailed simulation.

Quite right. In fact, we're two quick and dirty simulations run simultaneously in the same result space. This is why we will never have a Grand Unified Theory of Everything.
posted by Sparx at 6:57 PM on May 21, 2007 [2 favorites]

I'm an 8 bit character in a 64 bit simulation.
posted by isopraxis at 6:59 PM on May 21, 2007

> I recall reading it in the other MeFi thread, the one vacapinta linked above, and I believe even then
> it was a few years old.

Considerably more than a few. Brueckner's “Brains in a Vat” appeared in the Journal of Philosophy v. 83 in 1986.
posted by jfuller at 6:59 PM on May 21, 2007

The simulation...

When I was a kid, I believed that, when I turned around, the world behind me disappeared... the universe existed only in my mind.

Might be true...

thanks for the fpp.... very good stuff.....

Now...all of you be careful...don't get involved in anything too complicated....when I go to sleep, you're all going to disappear....

fyi.. I'm on EST, bedtime is in about an hour or so...
posted by HuronBob at 7:02 PM on May 21, 2007

Now...all of you be careful...don't get involved in anything too complicated....when I go to sleep, you're all going to disappear....

Yes, but when you wake up again, we resume exactly where we left off.
posted by UbuRoivas at 7:08 PM on May 21, 2007

Steven C. Den Beste :In particular, if this was a digital simulation, then Relativity would not be correct. There would be an absolute frame of reference -- the one the digital simulation was using.

Actually this not not so! Today's massive oline games have issues that are based on the real-world problems of lightspeed and space/time separation. The obvious algorithm that synchronizes everything on the main server doesn't work because the communication latencies are too large. They actually use a type of lazy synchronization where things only occasionally synchronize on a local basis when they interact: each player has his own virtual time and space 'frame' which is known only by his own local computer.

This sort of algorithm fits in with certain problems in quantum physics where it appears that interactions happen superluminally (which violate relativity). But you might be able to explain this paradox with 'lazy reality' that synchronizes just-in-time similar to the games...or in this case visa versa!
posted by Osmanthus at 7:11 PM on May 21, 2007 [4 favorites]

vacapinta: thanks for the meat (of it). I laughed, it sparks a douglas adams kind of attitude.
posted by uni verse at 7:41 PM on May 21, 2007

Im in yer simulashuns
stealing yr bucket
posted by uni verse at 7:43 PM on May 21, 2007 [1 favorite]

At this point, I don't think the argument is very new.

Plato was bringing up similar ideas in his cave allegory some 2000+ years ago. So no, these things are not even remotely new.
posted by teece at 7:47 PM on May 21, 2007

Also: This thread is itself a replica.

Wow, vacapinta, yours didn't show up when I ran a search of the URLs... Maybe we really are in a simulation and it purposefully prevented me from finding the other thread. *plays spooky music*...

I hope the admins will let my post stay (the Simulation Argument site has been updated, and added to, quite a bit in the five years between vacapinta's post and mine) because I'm really enjoying reading the discussion.
posted by amyms at 7:50 PM on May 21, 2007

This argument was more interesting when one possible source of our deceptions was an evil demon. Maybe if the simulation was generated by a possessed computer, this idea would be more interesting to me.

isopraxis, is it full of Goombas?
posted by Tehanu at 8:06 PM on May 21, 2007

But how many layers of meta-simulation are there? As our computing power increases, and we begin to simulate consciousnesses, then we will indirectly multiply the computational load on those who are simulating us. And when our sims figure out the simulation thing and start running their own simulations...

I like to imagine that when I run a simulation, one of the events I'll most look forward to is the day when my sims figure out their situation and try to communicate with me. So perhaps we should do likewise and figure out some way to wave and smile at our simulators. (And maybe we should try to be extra cute so they don't get bored and decide to switch our simulation off).
posted by gregor-e at 8:23 PM on May 21, 2007

If this essay is true then I'm pressing A-B-A-B-Up-Down-START. I never wanna come down man.
posted by Peter H at 8:29 PM on May 21, 2007

This reminds me of a similar argument (I can't remember where I saw it and my Google-fu has failed) that the human race will meet a catastrophe soon. The argument is that since the majority of all people born during a period of exponential growth will be born near the end of that period, it would be improbable that we live in a time far from the end of exponential growth. I remember it being framed as a proof that the world is about to end, while really all it argues is that exponential growth will end soon. No surprise there. Like this simulation nonsense, the conclusion has too many conditionals to be relevant. It's wankery and nothing more.
posted by obvious at 8:32 PM on May 21, 2007 [2 favorites]

Afroblanco, page 13 and 14 are on the same leaf.

Also: I wish images were turned on so I could snark like a Sim.
posted by infinitewindow at 8:33 PM on May 21, 2007

This is sort of a cool argument precisely because it's not a "brain-in-a-vat argument" in the standard Cartesian way. The purpose of the argument is not to claim that we aren't justfied in believing anything about the external world but to claim that we are justified in believing a specific thing about the external world---only it's something different than we expected.

It does seem like there are some problems with the argument. First of all, it uses a posteriori justification for claiming that our experiences are invalid. If the computers that we expect to be powerful enough to simulate ourselves in the future are themselves simulations, why can we infer anything about the computers in the "actual" universe from them? As well, there's the whole Putnam-style content-externalism argument to contend with.

Also, the argument being made in the paper is a lot weaker than it lets on. I'd think most people would accept the second fork of their trilemma and not the third; namely, that posthumans are extremely unlikely to run ancestor-simulations (after all, would you if you were a post-human? In fact, post-humans can run ancestor simulations without the "simulation" condition obtaining: because if we are in a simulation, we know that it is the sort of simulation such that the experience of being in the simulation gives no (overt) evidence of the simulation. There is no reason the simulation necessarily has to be like this, though.

There's also the Searle-style objection about symbol manipulation, Turing tests and whatnot. But I find the Chinese Room Argument a bit old hat, and I don't agree with it anyway.

Lastly, and most glibly, there are actually four options inherent in the paper:

It's very hard for a civilization to become posthuman
Posthumans don't run ancestor simulations
We are ancestor simulations
There's a mistake in the argument somewhere

On the basis of observing the entire history of philosophy, I'm sticking with #4.
posted by goingonit at 8:35 PM on May 21, 2007 [1 favorite]

gregor-e writes "As our computing power increases, and we begin to simulate consciousnesses, then we will indirectly multiply the computational load on those who are simulating us. "

Which will just cause the simulation to take longer to calculate every time-step. We, in the simulation, won't -- can't possibly -- notice that it takes longer for the sim to calculate its state at N+1 from state at N.

(And not even that, if it actually calculates each sub-atomic particle: after all, our computers are just arrangements of atoms, and in principle calculating the next-step of an atom of sand is the same as calculating the next-step of an atom of silicon.)

gregor-e writes "So perhaps we should do likewise and figure out some way to wave and smile at our simulators."

Traditionally, this is done by sacrificing a son (or daughter) to the Great Simulator. According to some metaphysicians, the Great Simulator responds by making His only begotten Son a sim, and sacrificing that sim for us.
posted by orthogonality at 8:44 PM on May 21, 2007 [1 favorite]

A simpler limit can be found by looking at the size of the universe, and the sheer limit to how much information could be processed in it: about 10¹²⁰ "bits." (10⁹⁰ practically, sez the article, once you account for entropy).
While that's a staggering number for walking-around processing, it's easy to conceive of problems that'd bust you up to E like on a cheap Texas Instruments pocket calculator. The incredible resolution required for a full simulation of EVERYONE (let's avoid solipsism for a moment) would easily be beyond the physical capabilities of our universe.

Of course then, our mass limits are all simulated too, yadda yadda...
posted by klangklangston at 8:49 PM on May 21, 2007

I wonder if before computers existed people just thought we were all just characters in a book?
posted by Peter H at 8:51 PM on May 21, 2007 [2 favorites]

SCDB: Digital isn't really the most accurate word, what you want is computable. There are some important differences.

If you're interested in an exploration the implications of having a computable universe, regardless of simulations, I can reccomend Max Tegmark's work, notable his paper The Mathematical Universe, where he talks about the universe itself as math, and the implications of this in discussion of multiverse theory, and possible ways to test these ideas.
posted by Arturus at 8:54 PM on May 21, 2007

There are inherent and fundamental differences between how an analog universe presents and a digital universe would present, and if we were in a digital universe we'd know it by now.

stephen, you're assuming that a far advanced race wouldn't be using analog computers

The article makes one major assumption: that computers (or programs or the two together, if you prefer) will one day be able to think the way we do. That is that they will have intentional consciousness.

and oddman, you're assuming that WE have intentional consciousness ... some have argued that what we have is a mere simulation of intentional consciousness

A simpler limit can be found by looking at the size of the universe, and the sheer limit to how much information could be processed in it: about 10 120 "bits."

and klangklangston, you're assuming that the universe you're measuring isn't a simulation ... for all we know, it could be the equivalent of one of those wristwatch pet toys in the REAL universe

as for me, i'm assuming that i don't know nothin'
posted by pyramid termite at 8:59 PM on May 21, 2007

I like to imagine that when I run a simulation, one of the events I'll most look forward to is the day when my sims figure out their situation and try to communicate with me. So perhaps we should do likewise and figure out some way to wave and smile at our simulators. (And maybe we should try to be extra cute so they don't get bored and decide to switch our simulation off).

If those guys exist, they're maybe reading this right now, like I read the newspaper in SimCity 2000. You could just say hi.
posted by TheOnlyCoolTim at 9:01 PM on May 21, 2007

Meh, at least this beer is damn tasty.
posted by arse_hat at 9:33 PM on May 21, 2007

Traditionally, this is done by sacrificing a son (or daughter) to the Great Simulator.

Maybe that's what happens to all the self-linkers when they are banhammered... Maybe the Great Simulator eats them.
posted by amyms at 9:48 PM on May 21, 2007

I like to imagine that when I run a simulation, one of the events I'll most look forward to is the day when my sims figure out their situation and try to communicate with me.

So how could you communicate with them in such a way that they would know it's YOU, and that they're not just, say, hallucinating?

And, for that matter, how would you know that they're really trying to communicate with you and not just engaging in some sort of artistic or otherwise irrational behavior?

Things get a lot more complicated when you suppose that the simulations really do have intentional consciousness.
posted by treepour at 9:49 PM on May 21, 2007

This is sort of a cool argument precisely because it's not a "brain-in-a-vat argument" in the standard Cartesian way. The purpose of the argument is not to claim that we aren't justfied in believing anything about the external world but to claim that we are justified in believing a specific thing about the external world---only it's something different than we expected.
goingonit

You're looking at the wrong Cartesian argument. It's not the "brain in a vat" or the "demon" argument; this is a rehashing of his dream argument. If you want to put it in terms of the "reasonableness of possible claims about reality" and not about the nature of reality itself, then:

1. There are dreams that vividly mimic real life, to the point that we cannot distinguish the two and that, while dreaming, we feel that we are awake.

2. Also, there are dreams in which we think we wake up, but in reality are still dreaming, which we realize when we do wake up.

3. Given this, there is a strong possibility that what we are now experiencing is simply a dream, and not "real reality".

No need for speculation about future computational capabilities or motives. Rationally, there is no way to show that it is not at least possible that we are dreaming. teece is right. Objective reality and truth are ultimately unknowable. It's an interesting thought exercise, but ultimately little more than that.
posted by Sangermaine at 9:59 PM on May 21, 2007

Sangermaine, the dream arguments suppose that a dream is an experience and not just a memory of having had an experience. We remember having dreams but does this mean we actually experienced them? There may be no such thing as "having a dream" -- only "remembering having had a dream."
posted by treepour at 10:07 PM on May 21, 2007

A bug in the simulation might explain why one sock out of every pair eventually disappears off the face of the earth.

(on preview: hello, simulators! fix this disappearing sock bug, already!)
posted by UbuRoivas at 10:18 PM on May 21, 2007

so, in this simulation, they simulated the Wachowski brothers and their three movies?

kind of a spoiler, isn't it?
posted by drjimmy11 at 10:29 PM on May 21, 2007

Jesus, someday I am going to bomb Time Warner for their crappy service outages...

treepour,
We know that nightly we experience periods of no consciousness. We sleep, at some point our perception of the world stops, then starts again when we wake. Even if dreams are just memory constructs, within the construct we feel as if we are living it. That just seems to add another layer but leaves the dream argument the same. How would we know whether the dream was experienced or just a false memory of experience? The possibility would exist that it was an experience, and that what we experience now is also such a dream experience. There's no logical way to prove or disprove it.
posted by Sangermaine at 10:30 PM on May 21, 2007

The dream argument, and this computer simulation argument, rests on probabilities. If a non-zero probability exists that we could be experiencing a dream or a computer simulation, then it can't be logically ruled out.
posted by Sangermaine at 10:31 PM on May 21, 2007

hi, i'm on metafilter and i could overthink plato's beings
posted by sergeant sandwich at 10:43 PM on May 21, 2007 [5 favorites]

Simulated plagiarism.
posted by Blazecock Pileon at 10:43 PM on May 21, 2007

Apart from the inherent cyberpunk geek factor, how is this really any different from Descartes's "evil demon" hypothesis? I mean, it's pretty much the same thing with a Keanu Reeves movie overlay, right?
posted by infinitywaltz at 10:50 PM on May 21, 2007

Oh, and since the Matrix keeps getting tossed around in this thread, I thought people might like this interview with Jean Baudrillard from Le Nouvel Observateur in which he says the movies had nothing to with his ideas and are at best a misinterpretation. Also contains this great confluence of philosophy, cinema, and anal sex: "What really is striking in The Matrix 2 is there is not the tiniest irony to help the spectator taking this monumental special effect in the rear."
posted by Sangermaine at 10:54 PM on May 21, 2007

pyramid termite, I make it a policy to ignore thinkers that are so caught up in the grip of theory that they feel free to deny the obvious. I do have intentional consciousness (they are people who doubt this? names?) and I also have free will (this is the more common candidate for the "you only think you do" nonsense). They only think they have valid arguments, but the validity is just an illusion (see that kind of cheap response is just unsatisfying).
posted by oddman at 10:55 PM on May 21, 2007

莊周夢蝶 Zhuang Zhou dreamt he was a butterfly

...

If we are living in a simulation, does that change life any? What should I do differently if I'm living in a simulation as opposed to reality? I don't see any way to transcend this reality, and unless I have verifiable proof that the human condition is better in some other reality, I see no reason to bother trying to actually escape this one.
posted by jiawen at 10:58 PM on May 21, 2007 [2 favorites]

Crap, I thought my Descartes reference was all clever and shit, but a bunch of people beat me to it. Serves me right for not reading the whole thread first (I did a quick search beforehand, but I used "Descartes" and everyone else used "Cartesian").

If whoever is running the simulation could edit this whole interlude to make me sound smarter, that'd be great. KTHXBYE
posted by infinitywaltz at 10:59 PM on May 21, 2007 [1 favorite]

I took them as they came...

But... he'd just be uploading himself to a Virtual Machine!

What I call reality IS a simulacrum. The whole escape idea is the problem. Out of ? to ?. Just because I am here and I am thinking and living and rational and alive, how is this to mean I am more worthy than what I am? (Which is, what I am.)

It seems to me a more interesting point is that, though the argument doesn't REQUIRE it, it implies that not only would we be living in a simulation, but that we likely are a simulation.

Now this is what it this all is to me. The entire thing is a simulation. But since it’s our reality, it’s reality, too. But only to us. That which has created us/all this thinks of this as whatever it is to them. Since there is meaning in my life – love – then whoever or whatever this is here “for” is irrelevant to me. From my perspective, we are created and not necessarily for the ends for which we imagine, but this is not to invalidate those ends.

Tricky? Let’s see. What if we are created by what we call monsters for the purpose of the production of pain suffering and hurt which “they” use for nourishment. But in order to make the farm that we are/reality is “go,” there has to be a method for us to continue. Or maybe the longer the suffering lasts the better.

In any case, it is not beyond conception since to do such a thing as the simulation of an entire universe that is real to that which makes it up would require a degree of spontaneous adjustment, and spontaneity requires an element of unpredictability (“the improbability drive,” I believe it is called). So in my simulation that I call reality, what I call Love happened. And while it is anathema to what created us in the first place, it is also something “they” give no thought to. (If it was YOUR simulation that you were using for YOUR ends, would you worry that the machine had developed a “spirit”? I mean, we don’t now ourselves…)

I've met people who get a little too worked up on these questions, and I've always felt that they should spend more time worrying about bettering where we are, and we deal with people in the here and now, and less thinking about that which may or may not exist. If that makes sense.

Well, in order for this to happen, one must land upon a reason for being. In all honesty, it SEEMS as if the cards are stacked against us. And worrying over “monsters” that may or may not exist can be pointless, but it can also lead to inner knowledge. Then, regardless of the situation, everything works out.

Think of it as if it were being a people or kind in bondage – by that which created us. We ourselves don’t give credit to living things we create (or breed, or incubate, or etc.). They are not our “equals.” We mostly just use them as we see fit. And what we call nature is what people call fractal, that is, the same at ever outwardly or inwardly spinning magnitudes. (Including quanta.)

But as bondaged people, then we look for other ways out. If my body is bound, I use my mind. If my mind is bound, I use my spirit. If my spirit is bound, I use my imagination. And so on. Eventually, it gets to a freedom that cannot be tainted by anyone or anything.

Another point, computing power may well be cheap enough that they might as well add all the detail that we see. And do we KNOW how accurately we reflect this theoretical reality? So, we say we feel love. We feel hate. We feel hunger. We feel pain. Maybe we just feel rough estimations of those real feelings and emotions. how would we know? Maybe they haven't been able to simulate the key emotion of BlarkVol at all, and they just skipped adding it to us at all.

I agree with the point about computational power. That seems almost intuitive at this point. But when it comes to simulated emotion, I disagree. And another thing is that regardless of the nature of our universe, simulacrum or not, for it to be autonomous, contact with anything other than it would have to be limited. No two-way talking mirror. No talking with the denizens here at all, in fact, for spontaneity and autonomy to happen.

Certainly, maybe we are not autonomous. But I know that love is real in the way that I know, so that is an unacceptable premise. It is my “illusion” that love is what god is, although I do not give this quality (or quantity) which I call god the effect of self-awareness. Love is alive to me, and it is what it is – the purpose of being – but it is not demanding or crusading or have needs or desires. So regardless of what we or our universe was created for, what purpose it actually serves is another matter. The miracle of spontaneous adjustment – a quality which is necessary for a universe to “work.”

And I think the idea isn't that we're going to recreate actual historical events, but rather set up the initial conditions of the simulation and let it unfold on its own.

And those conditions unfold on their own and lead to their own conclusions. Again the uncontrollable and desirable condition of spontaneity.

Perhaps these simulations are really easy to run, given adequate processing power and properly designed models. I wouldn't be terribly shocked to learn that we're all part of Grand Theft Auto: The Naughty Oughties on some kid's Playstation 13.

But what is interesting about it is that while we have been created as a “videogame” to “farm” whatever quality our creators would desire – in precisely the same way we play videogames for the production of “thrills” or pleasure – that a videogame would be infinitely more interesting if it actually did something. Autonomy is that something. Without it, it’s just a videogame. With it, it’s wild!

I'm not criticizing amyms or Bostrom, cause it IS fun to think about, and I hope to see a good discussion of the theory here, but I've met people who get a little too worked up on these questions, and I've always felt that they should spend more time worrying about bettering where we are, and we deal with people in the here and now, and less thinking about that which may or may not exist. If that makes sense.

Well, the first step is to realize there is no there, that we are here, and we are staying here, and regardless if we are some kind of funtime production facility or farm or whatever for our creators, there are reasons to exist. That’s why we give live it’s chance.

The key point of the argument is that if you can believe our civilization will eventually be able to make a virtual universe simulation environment that is indistinguishable from what you experience now (simulation of people and physical laws), then you can think of a recursive series of those simulations. The first simulated universe makes another simulated universe inside it and so on. Then, barring any other evidence, you (and I) exist in one of these arbitrary universes. Since there seems to be an awful lot of them due to the recursive property, there is a high probability that the universe you actually exist in is one of the simulated ones, rather than the original.
It seems that the argument is sound, but I doubt if such a universe simulation is possible.

This is the “philosophical masturbation” (sic) on the part of the author of the posted article’s (extremely predjudiced) viewpoint. There is no reason to believe that while we are emulating our conditions with our own direction and lifestyle (animal farms), that we would ever do the same thing. For one, the physical limitations of our universe are more than likely conditions of the simulation, and the laws and scientific principles we understand are inapplicable outside the scope of the simulation. It’s like these are just the rules of our videogame. Mario’s World is resemblant to our own world, but only superficially. Just carry this to the other extreme.

If you really can't discern the difference (assuming such a simulation is possible -- I doubt that it is) -- what the hell does it matter? It doesn't.

As long as one is true to reality, is perhaps doesn’t. But once we are able to imagine something better, than it immediately becomes of paramount importance.

Why? Our so-called condition resists bettering. If this is inherent, then we may be able to better spend our energies in making the things that do work for us, work better.

Assume that some other intelligence from a vastly more complicated universe simulates us. Why would their vastly more complicated reality make it any easier to simulate our already pretty complicated reality? Maybe it would -- maybe it wouldn't.

We can't answer these questions; ultimately I think they are the exact same questions religions ask. The simulators and God would be one and the same. Which is not to dismiss such ideas -- I am fascinated by them -- by they are not a fashion of idea we have any method of proving or disproving.
Why do God and simulators have to be the same? Only if God is thought of as the creator, or as controlling, but there is no reason to make this assumption. Personally, I imagine the creators to be very much like us – that is, without answers – but beyond this I am aware of a greater aspect of myself, and lastly I understand God to be a life that is developing as more and more love is produced but without self-interest.

A simulation doesn't have to run on a digital machine, and computers don't have to be digital. In Hitchhiker's Guide, the easiest way to make a powerful problem solving computer turned out to be "use a planet". But the planet was real, not a simulation. The planet and all the life on it were the computer and the software. I suppose that when you have access to horsepower like that, you can, like, totally run Doom on it.

Exactly.

Hey dude, we are the less detailed simulation.

But real in our own way. That’s the beauty of it. Our creator’s create, for better or for worse, and instead of being all videogamey, we have our own integrity. Obviously, if we COULD do this we WOULD, too.

I don't think its masquerading as an empirical hypothesis. And, from my understanding at a least, it's not so much "Hey, maybe we're living in a simulation," but rather "Hey, maybe its more rational, given certain presuppositions, to assume that we are." I think the claim is more epistemological than ontological.

In differential geometry, it is possible to derive extrinsic and intrinsic properties from a manifold.

In laymen's terms, you can be an ant on the surface of an orange, and if you're smart enough and have the right rulers, you can figure out things about the surface you're on. These intrinsic properties don't required that you see the "outside" of the manifold.

Instead, it might be possible to learn that we're "inside" a simulation through its measurably intrinsic properties. We may just not yet have the language for describing those properties.

And…

Today's massive oline games have issues that are based on the real-world problems of lightspeed and space/time separation. The obvious algorithm that synchronizes everything on the main server doesn't work because the communication latencies are too large. They actually use a type of lazy synchronization where things only occasionally synchronize on a local basis when they interact: each player has his own virtual time and space 'frame' which is known only by his own local computer.

This sort of algorithm fits in with certain problems in quantum physics where it appears that interactions happen superluminally (which violate relativity). But you might be able to explain this paradox with 'lazy reality' that synchronizes just-in-time similar to the games...or in this case visa versa!

....and that’s just on the top of my head. Reality is cool because it is a simulacrum and it’s real, too. That’s good programming even if it’s happenstantial.

In the end, there are some inexplicable qualities to reality that I am only comfortable with by having learned that everything always works out and that the genuine truth makes me laugh.

Considering that everything doesn’t appear to be working out or have worked out at all, that’s something!
posted by humannaire at 11:09 PM on May 21, 2007

It's not a computer simulation. Your mind is simply in the yolk of a very small egg, at the bottom of a very deep well, in the middle of a very cold desert, on the face of a very lonely planet, orbiting a very distant star. (Don't wake up.)
posted by kid ichorous at 11:17 PM on May 21, 2007 [2 favorites]

Wait, isn't time actually quantized, in a practical sense, from the standpoint of observability?

I thought physics had defined a definite smallest division of time that can be measured: Planck time.

This is not the same as the theoretical chronon, of course... but still, if there's a definite limit to observable measured time, that would seem to cover the quantization angle; we can only measure time in discrete tiny amounts, dividing reality into snapshots separated by Planck time - our perceivable level of resolution - while the Universal Processor has a much finer resolution or even runs unquantized.

And remember, our perception of reality using our brains is not actually continuous and unquantized. Our eyesight, for example, only gives us around 30 frames per second. We're poorly equipped to judge the continuum of reality even with our built-in gear... and even greatly amplified by our clever equipment, we still can't say anything for certain.

Our fundamental human perception of time is entirely subjective; since any processor can run a simulation of any other processor - for instance, you can run a Cray XMP emulator on a 1994 Mac SE... veeeerrryy sloowwwwly - our own existences could be "playing" at any arbitrary rate, maybe one Planck Time frame per thousand "years" of Universal Processor Time.

So it doesn't matter how many reams of data you generate trying to figure out how it all works, you still can't ever really know.
posted by zoogleplex at 11:25 PM on May 21, 2007 [1 favorite]

oddman,
I'm not saying you're right or wrong, but you can't just dismiss questions like this out of hand. What's "the obvious"? People are just trying to logically work out how things are from what can be known about the world. You can't just declare something as ultimately true. A delusional person might tell you that you're a fool for denying the large pink elephants he can clearly see in the yard. If you want names, read the articles on solipsism and determinism. How can you call an argument that is logically sound invalid?
posted by Sangermaine at 11:30 PM on May 21, 2007

Sorry to keep popping in, but I'd also like to say that thinking about this kind of thing doesn't need to be associated with seeking to "escape" or anything like that. It's just trying to find out the way things are, which is a natural urge for us curious humans. Maybe it has no effect on our lives at all, but we still want to know the truth.
posted by Sangermaine at 11:34 PM on May 21, 2007

You can't just declare something as ultimately true.

You just did.
posted by kid ichorous at 11:36 PM on May 21, 2007

jiawen beat me to the Zhuang Zhou reference.
posted by Abiezer at 12:01 AM on May 22, 2007

If countless universes - involving endless randomly generated permutations of laws and happenstance - are being simulated, the grand majority of these universes would necessarily be incoherent, absurd or chaotic. The overwhelming uniformity and banality of our universe is good evidence for its authenticity.

Also, unless our allegedly godlike decedents (see Idiocracy for an equally plausible future!), are also the most sadistic fucks this side of a Tarantino movie or an Old Testament author, there is no way in hell they would ever cause a near infinite amount of murders, wars, rapes, plagues, pains, tears, and deaths for an infinite number of fully sentient human simulations (indistinguishable from them in every meaningful way). This would be the greatest possible crime in the history of the universe. I doubt any human being would be capable of evil this profound and pointless, or any future society advanced enough to have such a capability would be so lawless to permit this.
posted by dgaicun at 12:34 AM on May 22, 2007

It's fairly clear we are in a simulacrum. Black holes are where the simulator tried to divide by zero. *duck*

Ultimately though, it's a meaningless argument; it's no different than arguing for the existence of God. Science is the study of the natural. God, or a simulator is by definition supernatural - beyond and above the rules that govern us, or it/they could not function.

We cannot prove or disprove the supernatural using the rules of the natural. Without proof, we have to assume it doesn't exist, or we start having to believe in invisible pink elephants too; or at least, we assume they don't exist for any current purpose, other than idle speculation.

The delusional man may wish to declare he can see the elephants, but I demand to see some proof (i.e. something other than individual eye-witness accounts) before I accept it.

On a derail point, the old chestnut about 30 frames per second is false. A film running at 30fps fools the human eye because it's taking snapshots of an analog moving reality, and we struggle to tell the difference between a moving item and a blurred item. When creating static images, many can distinguish the difference at over 100fps. Find a CRT monitor (i.e. not LCD), set it to 60Hz (refreshes per second) and focus at a point about 2 foot above it; most people can see the flicker at the bottom of their eye easily. I can personally see up to about 80Hz, I know people who can see higher.

That said, I don't argue the greater point that space/time is possibly quantized, though I don't believe that argument is fully resolved yet, as it comes down as much the method of measurement as it does what is being measured. I'm not much of a quantum physicist.
posted by ArkhanJG at 12:44 AM on May 22, 2007

IDDQD

Ouch! It didn't work.
posted by Elmore at 1:51 AM on May 22, 2007

oddman: The article makes one major assumption: that computers (or programs or the two together, if you prefer) will one day be able to think the way we do. That is that they will have intentional consciousness.

That is at best a questionable assumption; many would argue that it is obviously false.

I fail to see how it can be in any way "obvious"; we're thinking, conscious machines, ergo it's possible to make thinking, conscious machines. Nature does it all the time, so why can't we?

Now, the class of system required for this may or may not be something we can model in a turing machine; hey, maybe we do depend on some pseudo-magical quantum effects or something. Never mind -- we'll just make a different kind of machine with the necessary properties (all the way down to gigantic biological systems if we have to). Why would this not be possible, does God have a monopoly on the required entangled particles or something?
posted by Freaky at 1:59 AM on May 22, 2007

SCDB wrote: 'If this were a computer simulation it would mean everything would be quantized, not just energy. Position and time would also be quantized.

There are inherent and fundamental differences between how an analog universe presents and a digital universe would present, and if we were in a digital universe we'd know it by now. No amount of digits of precision in a digital simulation can make it act the same as an analog universe. Everything we see and have observed is consistent with an analog universe -- and inconsistent with a digital universe.

I'd have to differ here. Our ability to measure the universe is limited, and our ability to calculate many of our physical models is limited - we do them on computers after all. Comparing the two always involves some limit on the precision at one end or the other (except perhaps in the relatively rare case of measuring a property that logically has to be an integer).

In some sense, it has to be consistent with a digital universe because more and more that's the framework in which we calculate the results of physical theories to generate predictions so that said theories can be tested against reality.

In particular, if this was a digital simulation, then Relativity would not be correct. There would be an absolute frame of reference -- the one the digital simulation was using. And because of that, the things we observe wouldn't be as we observe them.

It sounds rather like you're assuming everything must be calculated on a fixed Cartesian grid. In fact, many of our largest scale simulations today use at least an adaptive grid so that things can be simulated on larger scales where they matter less and on smaller scales where we think that level of simulation is more important or we're particularly interested in events. And it's trickier programming but I don't see why there must necessarily be some universal frame over all space, and even if there is it should be possible to calculate relativistic effects and include them - there are relativistic ray tracers available for instance.

Anyway, I'm firmly in the "don't think future people will bother simulating us" camp. The power required would be phenomenal, and frankly I think our lives are just a little too boring for it to be worth doing.
posted by edd at 2:03 AM on May 22, 2007

Ditto, zoogleplex.

Planck time, the Planck Length, and other such quantizations of the real world lead me to believe that it is not quite analog... or even digital on an immeasurable scale.
posted by tehloki at 2:17 AM on May 22, 2007

I personally like the Tegmark idea (Arturus mentioned this) that mathematical existence is the same as physical existence. In this case, any simulation is just another copy of something that already exists as a real universe, and all the simulation is doing is increasing the statistical chance of any self-aware organism happening to find themselves in one of those copies. I doubt this would be very significant against the general fact that smaller mathematical objects will appear more frequently than more complex objects anyway, due to nesting.

It may take the arrival of intelligent machines for it to happen, but I suspect that Tegmark's idea will be convincingly proven within the next hundred years - we'll show that there are patterns within the statistical properties of arbitrary mathematical objects that have exactly the same properties as our own observed physical laws.
posted by teleskiving at 2:35 AM on May 22, 2007

teleskiving - one book I have long been planning to read is Masahiro Mori's (he of the uncanny valley) on his belief that robots have the Buddha-nature.
posted by Abiezer at 2:47 AM on May 22, 2007

sangermaine: The Cartesian argument is that If a non-zero probability exists that we could be experiencing a dream or a computer simulation, then it can't be logically ruled out. But this isn't the same kind of argument as what's in the paper. The argument in the paper isn't about possibility but about probability, which makes it differ in purpose from arguments for external-world skepticism.

But because it's so damn similar, this could be another flaw in their argument. If we're willing to accept that what we see isn't "actually" what's there (i.e. our normal experiential beliefs about reality are false) why should we be justified in believing anything about the external world at all, including that it is simulating us? The standards of proof are higher for the "simulation argument", because we actually have to accept their alternative, not just reject ours.
posted by goingonit at 6:09 AM on May 22, 2007

Freaky - Read Searle's paper on on the Chinese room, or articles on the phenomenological gap.

Sangermain - solipsism isn't something anyone really takes seriously. Further, whether the argument is sound or not is the question at hand. You can't just insist that it is without proof. I have my experiences to show that the conclusion offered is false and therefore I am compelled to believe that the argument is bad in some way (I wasn't really using 'invalid' in a technical way, the argument, strictly speaking might be either invalid or unsound).

The methodological question is simply this which of the fallible faculties, perception or reason, do you ultimately trust more? My stance is that when either offers a tenuous reason to disbelieve something presented as an obvious truth by the other, the faculty presenting the tenuous reason (i.e. tenuous claim) ought to be taken to be wrong.

For example:
1) I feel pain in my hand, though my reason tells me that I have no body and thus no painful hand. I ignore my reason and look for ways to relieve the pain in my very real hand.

2) A see a tower that was cylindrical when I was close to it now looks rectangular when I'm far away. Of course, I don't believe that the tower changed shape as I walked away from it. Why? Because my reason gives me an undeniable conclusion that it couldn't be so.

(In the above, I'm taking undeniable to be a very strong claim. We can always utter phrases that seem to express doubt but it is a much harder thing to actually doubt (not just idly doubt but actually deny the truth of a set of beliefs) things like perceptions and judgments about mathematics and causation.)
posted by oddman at 6:21 AM on May 22, 2007

If this is just a simulation could someone out there please install a patch to fix my migraines? I'm never going to be able to level up at this rate.
posted by MikeMc at 6:26 AM on May 22, 2007 [1 favorite]

oddman,
Whether or not solipsism is taken seriously doesn't speak to whether it's a valid argument or not. The question would be, why do you dismiss it? You say you have experiences to show that the conclusion offered is false, but the experiences are demonstrably not 100% reliable. It essentially comes down to you saying "I know what I know and that's that."

You don't have any reason to take your stance. On what grounds would you reject the faculty presenting the "tenuous claim"? There are many clear examples of times when our senses tell us something that is false, or when reasoning leads to false conclusions. For example, hallucinations occur which both seem real and reasonable. Dreams as well. How would you reconcile these with your stance?

The problem is that proof itself is suspect, and can't be used to prove the point. Saying, "reality is the way it is because that's the way I experience and think about it" is circular.
posted by Sangermaine at 6:35 AM on May 22, 2007

oddman,
Sorry, an addendum. I was thinking about your two examples:

The problem in both of them is that you're making assumptions in both of them that force the conclusion:

1) In this example, you say you would ignore your reason and trust your senses. But why? You would have to already know that your senses are right and your reason is wrong. How do you determine this, other than just deciding it is so?

2) Again, it's the same, with the faculties reversed. You now decide that your reason is to be trusted, and your senses ignored. Again, why?

The question is: how do you determine which faculty to trust in which situation, without already knowing what's true? Both examples assume that you already know what is true, and thus can discard the faculty that goes against it, but that's a false assumption.
posted by Sangermaine at 6:41 AM on May 22, 2007 [1 favorite]

Sangermaine, the key to my stance is that there is a quality to the evidence in each of my examples which makes it indubitable. I do not know what is true ahead of time, I recognize a certain weightiness to my perception in the first example and my judgment in the second example. This weight, of course this is just a poor metaphor, makes me unable to seriously entertain the notion that the evidence in question is false.

Of course the examples are limited. I feel comfortable that I can distinguish between dreams and reality by taking a long view of my experiences and comparing waking bits to dreaming bits. Granted there are times when I cannot in the moment decide whether to hold with my sense or my judgment. In those cases I should abstain from choosing sides (or be open to correcting myself later on).

Is it logically possible to say that I'm systematically in error? I guess so, but I don't think it's rationally possible. That is I don't think a person can exclude from their beliefs all claims that the reality presented to their senses is false. Not really. Not consistently. I just don't think it's psychologically possible (of course I'm thinking of what we might designate as sane people, insane people have malfunctioning minds and that is beyond the scope of my point). This suggested impossibility bears, in my mind, some serious philosophical weight, and robs us of any need to consider solipsism as a viable alternatively.

(A side note, when I say it's not taken seriously, I mean that it's not taken seriously because it is taken by all of the philosopher's of my acquaintance to be philosophically uninteresting and a rhetorical dead end. That is, it's not taken seriously because it's just bad philosophy.)

The tenuous claim, that computers can think like we do, is rejected because it makes a category error: computers are not the kinds of things that can think in the way we do. It would be like supposing that the color purple might some day compose a great opera score.
posted by oddman at 8:59 AM on May 22, 2007

oddman, can you prove to me that you're a conscious human being and not a mere simulation of one? ... and if you can't prove it to me, how can you be able to prove it to yourself?

could you be a philosophical zombie?
posted by pyramid termite at 9:04 AM on May 22, 2007

Wait, isn't time actually quantized, in a practical sense, from the standpoint of observability?

From my understanding of QM (which is much better than most, but I'm no PhD physicist, so YMMV), time and distance are "effectively" quantized in QM, but it's not really settled if they are "actually" quantized, but it's pretty hard to imagine how they would not be really quantized.

On the other hand, in GR, time and space are infinitely differentiable, continuous manifolds.

QM and GR fundamentally disagree. The two can not be combined. And yet, both make fantastically accurate predictions (QM is more accurate, but GR is no slouch). This lack of agreement between the two fields is what makes the center of a black hole so difficult to study -- you must use both there, but the two theories don't play nice together at the high energy/mass, tiny space scale.

But the quanta posited by QM is really small, so it's not like it's super easy to see that we actually live an a digital universe (at least in some respects).

So it's an unsettled and open question what the hell is really going on there. Either there's something in our theories needing amending, or the simulators screwed up.
posted by teece at 9:07 AM on May 22, 2007

Gotta go with #2 on his hit list. In terms of detail - any simulation won’t have quantum indeterminancy. (Hmmm...virtual - virtual particles) Which means the computer either wouldn’t know what’s going to happen (in which case it’s no longer a simulation) or it does, in which case it’d violate the uncertainty principle. Which might otherwise be fine, but humans can “observe” such a thing (quantum measurement) so either it’s eliminated - in which case the universe makes no damn sense at all - or you create virtual minds with limited free will - in which case there’s no point to the simulation. The third option is ancestor simulation pre-quantum theory or even perhaps optics. So we should be living in an earlier age if we’re a computer simulation. Before we were aware of such things. That’s manifestly not the case.
(It’s not merely that there is a change in our frame of reference, but that we can observe the difference - meanwhile there’s a constant reference frame measuring both - that violates the uncertainty principle.)
So really - any simulation is rendered moot by the vast array of possibility that can occur. If this infinitely powerful computer is to have any control over the simulation, it needs to inject some determanistic properties into the program. We’d see those. Or, if it makes us not see them, the program is useless. At some point the simulation becomes pure automation. And then what’s the point of running something where you know every single thing that’s going to happen?
(I think)
posted by Smedleyman at 9:12 AM on May 22, 2007

oddman,
Fair enough. Thank you for your interesting and thoughtful responses.
posted by Sangermaine at 9:21 AM on May 22, 2007

Smedleyman - no, another option is a nonlocal hidden variable interpretation of quantum mechanics. Such a theory is perhaps philosophically unpleasant in some respects, but the nonlocality and hidden variable nature would not be challenging to implement in a simulation, even if the 'real' laws of physics aren't in that form. So I don't think you can argue it away from quantum mechanical observations.
posted by edd at 9:26 AM on May 22, 2007

Which means the computer either wouldn’t know what’s going to happen (in which case it’s no longer a simulation)

Is that really true, Smedlyman?

I've done a little bit of simulation -- I certainly didn't know what was going to happen: that's why I did the simulation. But did the computer? In a sense, I guess so, as it's a deterministic machine. But pseudo-randomness is rather common in computer simulation, and the results of that pseudo-randomness can't be predicted beforehand, generally.

So does the computer really know what is gong to happen in Monte-Carlo simulations, say? Or what about when I model a chaotic differential equation (the Lorenz equations, let's say). I can't predict how a tiny change in initial conditions will affect the curves of the graph. Are you saying the computer can, beforehand? I think it really just works it out as it goes (using the word "know" rather loosely with respect to the computer here).

Lots of computer simulations are run where the outcome will be totally unknown; but the hope is that some general pattern emerges in a series of simulations.
posted by teece at 9:33 AM on May 22, 2007

The tenuous claim, that computers can think like we do, is rejected because it makes a category error: computers are not the kinds of things that can think in the way we do. It would be like supposing that the color purple might some day compose a great opera score.

I find this to be really uninteresting thinking, and generally wrong, oddman.

Barring some spiritual/supernatural justification, our mind is just a biological machine. Our current computers behave very differently, but "thinking like we do" on the part of computers, in the abstract, does not result in a category error.

Computers already think like we do -- they can do rather complicated arithmetic. Humans and computers are the only things I know of that can do that. Computers can't think like us in every way, or the vast majority of ways (and maybe they never will -- that's an open question -- maybe we have to just build a biological mind to think like a human; maybe not, in either event, it's a "computer" in a very real sense). But it is true today that they think like us in some ways.

I'm not trying to say that computers are smart or that current computers are as amazing as the human computers, note.

However, the idea that "thinking" is in the human category and "computing" is in some other category is a totally empty taxonomy, to me.
posted by teece at 9:43 AM on May 22, 2007

oddman: Freaky - Read Searle's paper on on the Chinese room

Is the paper any more interesting than the wikipedia article? I'm pretty damn sure my neurons don't have any understanding of the signals they're passing along and processing, they just do what physics makes them do; I don't really see how that diminishes the experience which runs on top of that substrate.

Also, note I said explicitly that the type of system a mind runs on isn't necessarily something which can run on a Turing machine (i.e. a computer) -- this doesn't stop us making other kinds of machine which *can* run a mind, if that is the case (which I don't really see any reasoning for, but nm). Do you disagree with that assumption? No class of machine we can build could possibly think, aside from those we produce naturally? I feel a strong need to file that idea alongside those I find in Watchtower magazine.

or articles on the phenomenological gap

I'll try to remember to do so when I'm less tired.
posted by Freaky at 10:36 AM on May 22, 2007

I have faith in the simulation theory.
posted by iamck at 11:55 AM on May 22, 2007

“So I don't think you can argue it away from quantum mechanical observations.” - posted by edd

Ok, but if observation on that level could be simulated virtually - would there be quantum randomness - or not - within the system?
Or perhaps as teece says - pseudo-randomness.
And how would their measurement of it align with ours? Would they dictate it?

I think it’s too convenient to say there is no empirical way of eliminating the possibility - without some form of mind control.
Perhaps I’m missing something on QM, but some observations, it appears to me, can’t be objectively verified. That is - the computer can’t violate the uncertainty principle. If such a thing can be simulated - ok, but - (Er, is a virtual virtual particle subject to the same physical laws?)
Hmm, I suppose it boils down to the circled square argument - I don’t think it’s possible to violate a physical property with impunity without it showing. And I suspect violations would occur because of the nature of simulation vs. reality.
Either you care or don’t that the sims know they’re in a sim. That’s predicated on the reason for running the sim in the first place. (see below) And you’d have to allow for that in the sim.

“Are you saying the computer can, beforehand?”

No, I’m saying by this argument if we are in a simulation by ultra-intelligent beings then all simulations of every possible state are/have been run.
There can’t then be real randomness.
Seems to me it’s not possible then for them to run a simulation, at least from their perspective.
Which is where I think the breakdown occurs.

I mean, either you’re running the simulation for entertainment or for information. Given one of the the premises of the argument (the Bland Indifference Principle) a civilization that advanced is unlikely to need the information. (if we’re talking superintelligent beings here, why stop at mid-grade superintelligent beings?)

Given there’s no randomness (in that all concievable states can/has be simulated) there’s certainly no new info to be gleaned, and where’s the fun in running any particular simulation?

It might appear to be random from our perspective, but it isn’t. Of couse it could be fun to limit one’s perspective to that of a sim, it’d make infinity tolerable perhaps.
But then you’d really care if the veil was lifted and you’d want to avoid that. So you’d have to either brainwash people (or yourself) when the truth is discovered/observed or you’d have to make the sim so close to reality that it didn’t make a difference. So then why not just horse around in alternative universes?

Which, I think, is scarier. It’s a real universe, but we’re all self-deluded roles of a superintelligent being hiding from itself.

(Wish I was more up on QM tho. I can smell an empirical argument there, I just can’t express it.)
posted by Smedleyman at 1:22 PM on May 22, 2007

No, I’m saying by this argument if we are in a simulation by ultra-intelligent beings then all simulations of every possible state are/have been run.

Ah, I get it. That makes perfect sense.
posted by teece at 2:00 PM on May 22, 2007

1. Let's admit the possibillity that in the future, we can simulate the past.

2. If we can create one simulation, we can create countless simulations.

3. Since, there are countless simulations but only one "original" what are the chances we are living in that one original reality? Pretty much zero.

Thus, we are living in a simulation.

It is a statistically persuasive argument. And a lot more credible than any religion. But still highly debatable.

I have heard other persuasive arguments...

You live in the future and you want to learn about the past. What's the best way to learn? By living in the past. Not time travel (impossible) but living in a society similar to your own at a particular stage in their evolution (it wouldn't be too difficult to physically insert yourself into another society - at birth - or at another time). Of course, you mustn't know that you are living in the past (or it would not be authentic) and you would be allowed to choose your time and place (late 20C/early 21C seems like a good time - living in the West - pain free - birth of computers (like living at the time of the dinosaurs) - comfortable life).

The simulation argument posits that only one of us is "real". The education argument permits everyone to be real but some of us to be from another time.
posted by bobbyelliott at 2:56 PM on May 22, 2007

Ok, but if observation on that level could be simulated virtually - would there be quantum randomness - or not - within the system?
I'm saying we haven't even shown we have quantum randomness, only that there are very good reasons for thinking it exists, and those reasons are not valid for simulations in some computer that can in principle inspect any part of its own memory. In other words, when you say:
That is - the computer can’t violate the uncertainty principle.
In one sense it doesn't apply to the computer and in another it's perfectly simulating it.

Quantum 'randomness' has not ruled out some underlying nonrandom principle except on the local level. A computer simulation could be nonlocal, trivially so in fact, it seems to me.
posted by edd at 3:17 PM on May 22, 2007

Sangermain, that is, bar none, the most civilized final comment in a discussion I've ever seen on MeFi. Feel free to offer more observations. Also, I admit I may not have been entirely fair to your position, I'll do better if that keeps the discussion going.

Teece, the point isn't that computers and persons do radically different activities. Humans can certainly parse and follow rules and do computations. The point is that computers don't understand meanings. And further computers that are fundamentally like ours, just exponentially more powerful, won't understand meanings either. Because understanding meaning is not a function of computing power. See the discussion between me and Freaky.

Freaky, the wikipedia article does a decent job summarizing the thought experiment and its import. To suppose that the simulation thought experiment is possible we would have to suppose that we'll eventually be able to build machines that are able to understand semantic content. I can't say whether that is possible or not. I have such a poor grasp of what that supposition would entail that I honestly cannot assess its likelyhood. So, I shall remain agnostic on that point. However, these machines, if they are possible, would certainly be self-conscious entities (assuming agency goes hand in hand with understanding semantics). Could these futuristic persons be the kinds of things that run a simulation such as we are discussing? (I don't know. I think it's an interesting question: will self-conscious machines allow us to use them the way we use syntactical machines? It's interesting.)

Finally, I'd like to point out that the simulation thought experiment asks us to accept the eventual development of machines so radically different from anything available to us that we probably wouldn't recognize one if we came across it today. Is this taking things too far? If a thought experiment ask you to conceive of something that is almost completely alien to your experience can you really trust the intuitions that follow from the experiment? For example, imagine I asked you to consider what the world would be like if in the future we evolve in a way that we can taste quantum fluctuations in stars 50 light years away. Do we have trust worthy intuitions about that situation? It seems unlikely. So, if the simulation is asking us to do something similar (e.g. consider an alien post-human culture) should we trust our intutions?
posted by oddman at 4:28 PM on May 22, 2007

And further computers that are fundamentally like ours, just exponentially more powerful, won't understand meanings either. Because understanding meaning is not a function of computing power. See the discussion between me and Freaky.

Right, which is why current AI research spends precious little time pondering computing power, and pretty much all of it's time understanding how "thinking"works.

AI is not about computing power -- it's about understanding how that which we generally define as "intelligence" (ie, us, and certain animals) does what it does. And short of a supernatural explanation, there is no reason to believe it is impossible for a machine to think like us. It may be a "computer" as we think of it today, or it may not be: it's also very important not to get hung up on that point (computers today are analog and silicon based so they can't... Nah, I'm talking about thinking machines, whatever form they may take). And we may never figure it out, even, but that is not at all the same thing as saying it is impossible to build a thinking machine.

Also, you're going to need to provide a very rigorous definition of "meaning" for that statement to have any meaning. ;-)

I really think you're just defining away an issue, not addressing it.
posted by teece at 4:37 PM on May 22, 2007 [1 favorite]

teleskiving: It's also interesting that there's an argument from philosophy of science which converges with Tegmark's mathematical universe hypothesis, called structural realism. The idea is a comprimise between strict scientific realism (The things that science talks about but which we can only observe indirectly are real) and scientific nonrealism (We don't actually have justification for the existence of such things), by saying that what science gives us is justification for the existence of things which are mathematically equivilent to the things that science talks about. The rest of it goes through high level formal logic to try and show that such math is actually fundamental, but I couldn't really understand those bits.
posted by Arturus at 6:04 PM on May 22, 2007 [1 favorite]

I just checked back in on the thread and got caught up on the replies, and I must say I am so impressed with the very thought-provoking and articulate responses... Those of you who have been responding back and forth to each other all day (I'm not going to name names because I'm afraid of leaving someone out) have given me a very enjoyable read, and I thank you. :)
posted by amyms at 8:05 PM on May 22, 2007

Perhaps a question we should ask is, do we have any idea what a Universe-sized quantum computer could do? I've heard all sorts of wild claims about the incredible potential power of a QC made out of only a few molecules, what's the thought on one made of 1.0E80 molecules? :)
posted by zoogleplex at 9:01 PM on May 22, 2007

zoogleplex: I think you get an open ended parallel evoluationary system that is capable of developing sentient life called "The Universe We Live In." :-)
posted by bhouston at 9:27 PM on May 22, 2007

Well, Teece, I've been paraphrasing Searle in this debate. Interestingly he maintains that thinking is, as you say, non-supernatural. In fact he thinks dualism of any stripe is clearly wrong. Yet he also thinks that thoughts cannot be reduced to merely physical processes along the lines of biomechanical or nueral events. In your parlance I believe his point is that the key to human thinking isn't how we do it, but what we're doing.

"Meaning" just refers to the semantic content of an word, i.e. the meaning of a word is just the thing the word is about. No fancy metaphysics or epistemology here. The common sense concept of meaning will do just fine.
posted by oddman at 10:15 PM on May 22, 2007

oddman, wondering if you can say more about how Searle manages to avoid what appears to be a contradiction between denying material reductionism on the one hand, and denying dualism/supernaturalism on the other.
posted by treepour at 12:21 AM on May 23, 2007

I used to have a half-serious idea that most of the population was too stupid, unable to change or be changed (despite all of your attempts), and unable to think critically to actually consist of intelligent beings. My idea was that most of them were really just AI constructs designed to approximate actual people as closely as possible. They would seem realistic on the individual scale, since the 'system' would devote more resources to fleshing them out and developing them a background, but on the large scale they would act as unoriginal caricatures of themselves or their role in the system.

Of course, some people were surely real - yourself, of course, via "cogito ergo sum" (we're not talking intelligent AI constructs here, just 'bots' like in an FPS), and any truly unique, interesting, intelligent, or special person you met. They were the other players in the 'game'.

I called it NPC theory, and never took it seriously, but it was interesting to think about.
posted by Mitrovarr at 2:30 AM on May 23, 2007

treepour, unfortunately I really can't. I don't find his view on this issue particularly compelling. However, this is what wikipedia had to say:

"Searle believes that consciousness "is a real part of the real world and it cannot be eliminated in favor of, or reduced to, something else" whether that something else is a neurological state of the brain or a software program. He contends, for example, that the software known as Deep Blue knows nothing about chess. He also believes that consciousness is both a cause of events in the body and a response to events in the body.

On the other hand, Searle doesn't treat consciousness as a ghost in the machine. He treats it, rather, as an emergent property of the brain as a whole (See holism). The causal interaction of mind and brain can be described thus in naturalistic terms: Events at the micro-level (perhaps at that of individual neurons) cause consciousness. Changes at the macro-level (the whole brain) constitute consciousness. Micro-changes cause and then are impacted by holistic changes, in much the same way that individual football players cause a team (as a whole) to win games, causing the individuals to gain confidence from the knowledge that they are part of a winning team.

The theory is sometimes mistakenly viewed as a kind of property dualism, since, in Searle's view, a person's mental properties are categorically different from his or her micro-physical properties. The latter have "third-person ontology" whereas the former have "first-person ontology." Micro-structure is accessible objectively by any number of people, as when several brain surgeons inspect a patient's cerebral hemispheres. But pain or desire or belief are accessible subjectively by the person who has the pain or desire or belief, and no one else has that mode of access. However, Searle understands mental properties to be a species of physical property -- ones with first-person ontology. So this sets his view apart from a dualism of physical and non-physical properties. His mental properties are putatively physical."

Philosophy of mind isn't really my field. So, I won't analyze this picture. But again, at a gut level, I don't find it attractive. (Primarily because I don't know what to make out of the claim that minds are ontologically irreducible and yet causally reducible.)
posted by oddman at 11:34 AM on May 23, 2007

“Quantum 'randomness' has not ruled out some underlying nonrandom principle except on the local level.”

I see. Yeah, I gotta read up more on QM. Very interesting stuff, but it makes my brain overheat.
posted by Smedleyman at 12:44 PM on May 23, 2007

oddman, thanks for that summary. I read a bit of Searle back in my undergraduate days, and recalled this as a stumbling point. I was hoping you'd be able to shed the light that didn't dawn on me at the time. :) I suppose Searle might say the apparent contradiction is illusory -- but perhaps inevitable, given the nature of what he's describing . . . ?
posted by treepour at 11:08 AM on May 24, 2007

nice summary, vacapinta.
1. Let's admit the possibillity that in the future, we can simulate the past.
2. If we can create one simulation, we can create countless simulations.
3. Since, there are countless simulations but only one "original" what are the chances we are living in that one original reality? Pretty much zero.

this reminds me of the 'proof' of the existence of god, to wit:
God is omnipotent and perfect, therefore must exist because not to exist would be an imperfection. hurr, okay, i will obey your dogma.

An argument that both eats and shits itself.

Back to the Matrix. 1. Even given that we can create ancestor simulations in the future, why bother? Why not just read a history book? Are we an alternate history, then? They decided to run a simulation to find out what would have happened if Hitler had lost WWII? Or if you hadn't stepped on that butterfly?

2. Why do countless simulations? 'Because we can' doesn't mean they must exist. In the 'real' "world", computing power is expensive, space is limited, resources are limited. Infinite simulations? c'mon.

3. What are the chances? The chances of something happening has no bearing on it's truth.
posted by Miles Long at 11:27 AM on May 24, 2007

Even given that we can create ancestor simulations in the future, why bother?

All else aside, that's a toothless argument. Why bother making video games? Why bother with economics simulations? Why bother making documentaries? Why bother climbing Everest?

It's not a given that there exists some far-advanced meta-race simulating us on tremendously powerful hardware, or that a hypothetical people would be running countless simulations; and no one knows the chances, obviously. But if said hypothetical race are anything like the human race that we know experientially, they'd certainly do things just for the heck of it, including run simulations.

This could be the popular fiction of the meta-race; the television, the video game, the gossip column. Given the ready resources, why wouldn't they do it?
posted by cortex at 11:46 AM on May 24, 2007

As I see it, the main reason to simulate self-aware organisms is that if you can arrange it so that most of them will have decent lives, you can potentially improve the average well-being of self-aware organisms across the multiverse in a statistical sense. And it may be that the most efficient way to do that is just to start the thing up and let evolution run its course.
posted by teleskiving at 11:55 AM on May 24, 2007

Hofstadter's recently released I Am A Strange Loop goes some way towards maybe explaining how you might get from dumb symbol manipulation to actual consciousness. And like GEB it's nice and readable.

Some of the ideas don't seem a million miles away from Searle's, but I think he manages to express them in a less easily confused way ;)

Is the Chinese Room experiment trying to express that just symbol manipulation is insufficient for consciousness? That there has to be some higher level, emergent property of it in addition to a simple rulebook on what to do with Chinese characters, but not necessarily *not* based purely on symbol manipulation?

The extension to the thought experiment Hofstadter would seem to suggest is necessary for a "soul" is that, in addition to just juggling symbols according to some rules, you also juggle symbols which represent rules for juggling symbols, and back again, many times, referencing how this sort of capability in purely mathematical systems (like Principia Mathematica, or even a camera pointed at it's own output device) makes them bloom into higher levels of complexity/capability/abstraction.
posted by Freaky at 1:14 PM on May 24, 2007

Treepour, I thought about the apparent contradiction that you were concerned with a bit more and came up with something. Searle might argue that mental properties don't reduce to other physical physical properties because they are not the kinds of things that are composites of other properties. That is they are physical things that don't have the right (or any) parts. So they are irreducible. I imagine that he might think of himself as making a claim to the effect that a bowling ball is no less physical because it can't reduce to tennis balls.
posted by oddman at 7:35 PM on May 24, 2007

« Older When monopolies collide. | Dean and Britta Newer »

This thread has been archived and is closed to new comments

MetaFilter

Are We Really Real?
May 21, 2007 3:53 PM Subscribe

Tags

Share

Are We Really Real? May 21, 2007 3:53 PM Subscribe

Tags

Share

Are We Really Real?
May 21, 2007 3:53 PM Subscribe