The Simulation Hypothesis

Nick Bostrum at Oxford published a paper in Philosophical Quarterly in 2003 entitled Are you living in a computer simulation? In it he discusses the possibility that, at some point in the future, it will be possible (with sufficient computer horsepower) to run a simulation of a world that is so detailed that its inhabitants are not merely automata - they are conscious and unaware that they are in a simulation.

This is more than just a science fiction premise. He manages to come to a fairly unexpected conclusion. One or more of the following three are necessarily true, he concludes, if this simulation technology ever becomes possible. (his term for a society capable of this "posthuman.")

1. The human species is very likely to go extinct before reaching a "posthuman" stage.

2. Any posthuman civilization is extremely unlikely to run a significant number of simulations of their evolutionary history (or variations thereof)

3. We are almost certainly living in a computer simulation.

The reasoning that got him there is a little tricky but completely credible - the original scholarly paper is linked above, and a highly simplified pop-sci version is here.

And this sits somewhere in-between. Beyond the delight I always take in discovering something (to me) completely counter-intuitive the implications of his conclusion are interesting to ponder. So give it a read and ponder away. Thoughts, anyone?

Once we're done huddling around the bong I should get to the reason I posted this here. I think the simulation argument has ramifications that go far beyond Sci-Fi frippery. In particular the Problem of Evil, never fully addressed by Christians but kind of punted off to the side, seems interesting again in this new context. Invoking "Free Will" suddenly seems to force a Christian into some other, somewhat uncomfortable positions.

Anyway, at the very least it's a good quick read. If philosophy or math aren't your things I recommend the extremely simplified second link or the only slightly more difficult third. And here's the inevitable Wiki.

And in case anyone has gotten this far thinking that I or the author of the original paper are under the impression that we are, here, right now, living in some kind of simulation, I'm not, and neither is he. As he puts it, the argument is that the probability that we are in a simulation is not only non-zero, it's a lot higher than you think.

 

Ah yes, the matrix conjecture returns in fancier clothing.

It is an interesting idea, this hypothesis that we are all essentially sims living in an advanced program, here's the problem with it: there is not only no evidence for it at all but also evidence against it.

As a staunch empiricist, I have to reject three conjectures out of hand: the dream (put forth first by Descartes'), the matrix, and now the computer simulation.

If one is going to make a claim that reality itself is not reality, I certainly expect there to be evidence behind such a bold claim.

Of course, the proponents of this theory are correct in that we would not be able to know, in some sense, if reality is just a simulation. At the same time, however, one can claim that anything, say, for example, that invisible unicorns project a complex hologram from their tails that is earth and it would be the same. The inherent problems with claims like these is that they are impossible to test and can explain almost anything. They haven't been disproven precisely because, by their nature, they cannot be much like God except, I would argue, even less capable of critique.

Yes, the human race has had its fair share of entertainment trying to recreate human life in some simulations, that's all fine and good, but it doesn't prove anything; so far nothing we are familiar with can recreate consciousness. So far all I hear are people saying "oh yes, it will be able to in x number of years", but where is the proof for that? Here is the problem, while our computers can program behavior routines which are extremely complicated; we cannot program consciousness. We can, yes, make realistic looking people, people who tend to act on their own based on external stimuli in complicated looking ways. But all of this is programmed! If no commands are programmed, nothing will happen and ultimately there is no robot or AI which can act outside of its programming. Consciousness that exists among people cannot be explained by computer technology.


So what does the evidence point to?


It is necessary to note several things:

1) Our universe shows very little evidence of design.

2) Morally, how could one justify the suffering of millions of people many times over?

3) Nothing in history is inevitable, there is no future that will definitely be reached where computers can create consciousness; we simply do not know if that is possible or not. Assuming that such a technology far beyond anything known today will exist in the future is absurd. Again, the present progresses in ways which are, in many cases, unpredictable. This reasoning brings to mind those who predicted flying cars and other unrealistic inventions in the early 21st century. Imagination does not equal reality and just because someone can think of a technology developing in a certain way in the future does not mean that it inevitably will.


Finally, even if this is true, just like the matrix conjecture, what would it matter? If it is true, every being that exists in the simulation has consciousness and all its trappings and thus the rules of how to treat other sentient beings still apply. Again, though, the evidence does not suggest that this is true and there is no reason to believe it in the first place since there is no evidence for it!

 

i guess we could be a simulation no way to tell . no reason to suspect we are though. not yet any way . if some 1 ever invents such a simulation i guess its likely weer in one as i thinck there would be only 1 genuine reality that could produce multiple simulations. i dont get number 1 and 2 though

 

The big problem for me is that the argument is circular. We can make computer simulations because we use math to describe the universe. Then we make a few simulations that look like our universe, and claim that because we can simulate the universe, the universe is a simulation. Of course our simulations look like our universe -- we made them look like our universe.

 

I would recommend reading the paper before you dismiss it as being the same as the matrix argument or the Cartesian Dream. It's not at all the same, and what makes it interesting doesn't hinge on the testability of whether or not we're in a simulation. Even its author isn't proposing that we are in a simulation. In fact the conclusion is something a lot more subtle. It's a short paper and very rewarding - I recommend at least scanning it before you dismiss it with the admittedly much less interesting matrix style stuff.

 

That's not the argument at all. Did you read the paper? He lays out the probability that, if at any time in our future people start making simulations then it's more likely that we are in one ourselves than not because the number of simulated people will so vastly outnumber the real ones. Despite this he's not trying to convince anyone that we are in a simulation.

 

duplicate post.

 

I think numbers 1 and 2 account for the two possible ways there might not be simulations in our future. Either we blow ourselves up (or get hit by an asteroid or something) or for some unknown reason by the time we achieve the technical knowhow we aren't interested in running such things. Those two account for there not being any simulations, and anything else means that there will be simulations, and if that's true, he shows how it's likely that we're in one.

 

There is a lot to chew on but at the beginning I would say Prop 2 is questionable because of the assumption of what defines an advanced species. Saying it is unlikely minds like ours would be simulated by an advanced intelligence wrongly assumes it would be advanced in every area.

I can imagine a another galaxy where the intelligent life forms had an innate computing ability but lacked any recognition of morality or even the concept, but then stumbled upon the idea during their advancements like we stumbled into a technological explosion. In this scenario our minds would be the advanced civilization because we wrestle with morality on a breathing basis.

Let me also admit I'm not 100% confident I completely understand the simulation hypothesis as defined by the links in the OP. Mainly because it is said all three props cannot be false.

On a personal note, I do believe we are in a simulation of some kind. When I look around at the insanity of our actions as a species I cannot accept this is reality because if it is, I'm jumping off the Sagamore bridge just after dinner. Basically, the pain of our existence does not justify our existence. If we stop existing when our frail bodies wither, there is no point.

 

But what if we are the first ones capable?

 

I was kind of shaky on the necessity of at least one of the three being true when I first read the paper years ago. At least I think I remember feeling this way - I can't remember the details. But now, considering it again, they seem reasonable to me. Or at least the first two take into account plenty of others that are sufficiently similar that they don't need their own case. Given the reputation of the guy, where he works, and the journal it was published in I'm working on the assumption that the reasoning is rock-solid and any issues I might have are due to my limitations, not the paper's. I definitely need to chew it over a bit more. The differences between the actual paper and the two simplifications are vast, even ignoring the math. I have to admit that I really admire the reasoning that can take me so quickly from considering the possibility of one day creating a simulation to agreeing that if we ever make one, no matter how far into the future, then it's likely that we're in one. Or nested into many.

 

how about the first 1 with the math? Have ya ever thought about who that must be, theologically?

and no matter what the thread paper renders, the math does share how 'we' can create. ie....basic evidence: mankind created the 'word' itself

within the use of mass, energy and time, mathematically there is a returning to the origins of all three.

the entanglement of energy in time is that ongoing 'force' (it is the gravity, actually).

and mathematically, all is a part of the other


Look into Virial Theorem to get an idea how that works conceptually within todays paradigm. It is incorrect but it can assist in how the parts are combined (mass, energy, time) to represent the whole.

 

I don't question his credentials but philosophy has a pattern of upending traditional thought so I always approach every theory on the assumption it can be seriously challenged.

Once I get a better grip on his trilemma I'm sure it will be more appreciated and the ironic aspect is I have no doubt we are in a simulation but how it was created and maintained is the problem.

In the very least, even a cursory glance shows he has succesfully argued it is more likely than not we are in a simulation. For a while I adhered to dualism but over the past few years have been convinced idealism is closer to the truth.

 

Oh absolutely, but I'm willing to give something like this the benefit of the doubt at least until I've got a handle on it. It's been so widely read that somebody would have pointed out the glaring flaw if it existed. So I'll just keep pondering for now until either it sinks in or I've found a flaw in a 10 year old widely read paper written by an Oxford professor.

 

no simulation.

we are entangled to everything else, at the same time


what transcends is a 'light' thing


That is how the concept of 'other' is within our conscious experience. We can know, "I" within itself.


news from a couple days ago

it's a brain twist

2 bridges to build are;

a. entanglement is the gravity of nature

b. a property of 'light' (em) and naturally, a part of existence


proof: your mass is entangled to this post, in time. Now ask yourself, why were you here!

 

Entanglement is too ambiguous and is trying to justify what we do not understand much in the same way "nature" has been a blanket explanation.

 

Interesting video helping to explain the OP in visual terms:

http://m.youtube.com/watch?desktop_...e=related&feature=related&v=_ucqYJzK1rk&gl=US

I had been contemplating a similar problem regarding journalism. Mainly, that at some point, CGI would be so available and authentic, how could we tell if a story was actually from a human news source? In short, let's say there are 8 broadcasts of CNN and they all claim to be the original. How could we tell the difference?

 

Fishmatter, I apologize for assuming a bit much before; regardless I just read the article myself and I do believe it is extremely problematic in that it assumes to know the way technology will advance in the future. I think that it is impossible to know and ultimately renders the entire argument as speculation rather than one which is meaningfully possible.

Right now, technology cannot reproduce human consciousness and I still believe this is the primary barrier that cannot be overcome. All our evidence indicates that consciousness is a purely biological process that cannot be replicated with machinery or technology. For sims, there is no getting outside of the programming of the designer; it just doesn't happen and they, ultimately, are just ones and zeroes. Sims are not self-aware either, they never will be because they don't have consciousness; one important element of the biological experience which cannot be captured or transferred using technology.

 

His argument does not rest upon exactly how technology will advance and extrapolating current technology is completely legit and the video I just linked helps explain why. Especially the part of having a brain and a computer in a box, asking questions while both are in a sealed box, and not being able to tell which is human and which is the computer.

Our brains are simply complex computers and being biological does not negate the basic principles of electrical processes. Claiming Sims will never be able to be self aware is simply absurd given the technology we are aware of today. At the current rate of progress in computing processes, the technology will be available in as little as 50 years. It is this threshold that is mind blowing because it signifies we are more likely to be in a compsim right now.

 

Seeing how there is an attack on the argument based on biological versus artificial, it may help to explore why biological processes are not beyond human construction. People often view the biological as separate from other processes based mainly on the creator. If a human created it mechanical and artificial it must be, so anything not created by people must automatically be biological.

Using the right equipment, we can look deep enough into a human finger to find mechanical actors taking shape not much different from an assemply line or an internal combustion engine. If we can construct a finger to look and operate exactly like a human finger to the point nobody can physically notice the difference using all available equipment, it proves we are inherently mechanical and not biological. Since we are mechanical we are artificial thus proving we must have been created. Our ability to create artificial objects is meaningless because birds and spiders create artificial objects all the time.

 

SkyStryker,

While I understand that we create objects which have the apparent features of biology, none of them match exactly and for very good reason. Ultimately, what I'm arguing is that the basic experience of consciousness, that fundamental self-awareness that allows us complete freedom, cannot be replicated by technology. Technology can program complicated behavioural routines, it can even program randomness into the equation to make the sim seem autonomous, ultimately, however, it cannot program something that can function outside of its programming. If the programming isn't there, then it isn't going to happen. You can also program robots to learn, but only within their programmed abilities. If you have a robot programmed to lift items and translating language, it is not suddenly going to develop its own language and start interacting with people freely.


Also, I'd like to repeat a few arguments I made earlier, namely that

1) There is absolutely no evidence of this massive claim (extraordinary claims require extraordinary evidence!)

This is why I reject this argument to a degree, there is no evidence of it; at all. It claims to tell us about the very nature of our existence without any substantiating evidence. This makes me skeptical.

2) This argument assumes that technology will inevitably progress in a specific way, when such a progression is not inevitable and perhaps not even possible.

Ultimately, we ourselves came up with the concept of the simulation and that it can be put into a computer in the first place. Its origin lies within recent times, and so it seems over-reaching to say that such a recently invented concept now accounts for the entire universe.

We also cannot predict how far technology will go in the future or what it will be able to do with it. As the author pointed out, humanity may well go extinct before such technology is developed. But to even say it will be developed to the extent where consciousness can be transferred into purely digital creations is fallacious, there is no way to know that it will or that it is even possible for it to occur at all! Without proof that it can happen, this argument fails.


To re-summarize. This is an idea which claims to know the nature of the universe, but which offers no evidence for it besides that simulations exist even though these simulations are used merely for entertainment and cannot come anywhere close to transferring or capturing human consciousness. Without any basis for believing that this will be possible in the future besides technological optimism, the author argues that we are "almost certainly" in a simulation unless humanity is wiped out before being "post-human". All this is claimed confidently without a shred of empirical evidence. Right now, everything can be explained sufficiently by reference to the natural world; this is an unneeded hypothesis which has no evidence to back it up and makes too many assumptions to be considered meaningfully possible.

 

Also, there actually are replies from other philosophers who have different objections to the argument, here are a few I found upon searching through journal articles which cited Bostrom's article. He is not without challenges from his fellow philosophers.

1. Weatherson, Brian. Are you a Sim? Oxford Philosophical Quarterly. Vol 53, No. 212. July 2003.
2. Bruekner A. The simulation argument again.

I'm not sure the second argument disagrees with Bostrom, but Bostrom and Weatherson have a back and forth with Weatherson arguing that Bostrom ultimately does not provide a sufficient basis for many of his claims. I think it is fair to say that the argument isn't watertight and has holes of its own which need, at the least, be hashed out further before it can be accepted as a real possibility in the face of the blunt fact that there is no evidence for the thesis at all.

 

Update: in a reply to one person who replied to him, it turns out Bostrom actually is not interested in showing that we are living in a computer simulation at all.

 

You may be right, of course, but the consensus amongst philosophers of mind, neuroscientists, and the entire transhuman movement is that the hard problem is merely that: hard. But not insurmountable.

What evidence is there that this is a problem that can't be overcome? We hardly know enough about how the mind works to form an opinion, surely? Do you just mean that currently it's far, far beyond both both our understanding and our computational abilities? If yes then I agree, but this isn't evidence that figuring it all out is impossible.

Remember, it doesn't matter if it takes 50 years or 5 million years until someone tries a simulation. The argument stands regardless, as it does if an alien race decides to have a go. It doesn't matter, the moment simulations become possible the likelihood is that we are inside of one. Not that I think we are - this is brain exercise, not a proof by deduction.

 

Again, I think you're rejecting a claim that isn't being made. I'm not claiming we're in a simulation, and neither is the author of the paper. It's the implications that arise logically from the possibility that, for me, are interesting to ponder. But I'm certainly not looking for ways to prove it one way or the other - it's just a thought experiment which will take you to some interesting places, theologically, if you can stop trying to prove we're not in a simulation.

And while I disagree that there's anything special about biological as opposed to mechanical machines and we will eventually be able to crate artificial intelligences, even that doesn't matter and has no bearing on why I find the whole thing interesting.

Tell you what - I'll even concede all your points until I get more information, ok? Doesn't change a thing for me because, as is often the case with thought experiments, a willing suspension of disbelief is required.