That’s why it doesn’t matter if God plays dice with the Universe
March 4, 2014 5:25 AM   Subscribe

Discovering Free Will (Part II, Part III) - a nice discussion of the Conway-Kochen "Free Will Theorem".

The article includes a proof of the Kochen-Specker Theorem which I found clearer than some more technical accounts [wikipedia] of the result. There's also a link to Conway and Kochen's "surprisingly readable" article in the Notices [pdf] of the AMS.
posted by Wolfdog (92 comments total) 41 users marked this as a favorite
 
Although the Free Will Theorem can't prove if we have free will, it does have a fundamental consequence: if the Universe is deterministic, and a particles behaviour is always described by a function of the past, then we can’t have free will. And Conway is convinced that we do: "I can’t prove we have free will but I still believe that we do."

Ok then.
posted by shivohum at 5:45 AM on March 4, 2014 [3 favorites]


"The Free Will Theorem says, that subject to these axioms, if indeed the experimenters have free will, which means that which button they are going to push is not a predetermined function of the past, then the particles have the same property. The answer they are going to give is not a predetermined function of the past," says Conway. "We phrased this, evocatively, as: if we have free will then so do the particles."

This sounds a lot like process metaphysics: free will becomes a certain unpredictability or indeterminacy built into the fabric of reality, which human beings interpret in spiritual and ethical ways.

There's been a lot of work in philosophy since Conway-Kochen on free will, determinism, and the compatibility of those two things, but I think that everyone should read Strawson's "Freedom and Resentment." It just doesn't look like "unpredictable" is what we want out of free will.

At the very least, we want to be able to blame and praise people for their actions, which is to say that we want to hold them responsible for what they do and don't do. The idea that every action has the same random weighting and past independence as pushing this or that button ignores the fact that sometimes the button is attached to a gun pointed at a human being and sometimes the button is attached to keyboard writing a great novel. We're not infinite monkeys, and we shouldn't be satisfied with an account of free will reducible to that little joke.
posted by anotherpanacea at 5:55 AM on March 4, 2014 [12 favorites]


"We phrased this, evocatively, as: if we have free will then so do the particles."

This sounds a lot like process metaphysics


It sounds a bit like panpsychism, but let's not go there.
posted by Segundus at 6:03 AM on March 4, 2014 [1 favorite]


Oh, this is the thread where I get to expound on determinism, quantum physics, the evolution of consciousness, and irony. But I apparently choose not to.
posted by DarkForest at 6:11 AM on March 4, 2014 [9 favorites]


Personally I feel like it makes less sense to think about free will in terms of "Are human actions a result of deterministic processes that predetermine a particular outcome" and more in terms of "Given perfect information about all of the input to a human and how exactly humans work, could you predict their actions to a high degree of accuracy". On a probabilistic level, current actions are at the very least not independent of past inputs. The decision of what restaurant you go to tonight is dependent on experiences at previous restaurants, for example. It's just a question of how much accuracy a good prediction model of all of that would be able to achieve. The deterministic/non-deterministic view of free will mostly just boils down to "Can humans act completely at random?" which is not what most people are talking about when they discuss free will.
posted by burnmp3s at 6:14 AM on March 4, 2014


if the Universe is deterministic, and a particles behaviour is always described by a function of the past, then we can’t have free will.

See, this is a kind of 'free will' that no one (except perhaps philosophers) particularly cares about. Don't we want our choices to be the result of our material thinking machine interacting with the material world? What kind of 'freedom' would we be talking about if it is wholly independent of the physical universe in which we are instantiated?
posted by leotrotsky at 6:16 AM on March 4, 2014 [8 favorites]


“The fundamental reason why these things aren’t predetermined,” says Conway, ”is that no consistent set of answers exists.”

This is a statement about the fundamental nature of reality. It sometimes feels as if Nature is sort of "patching" things as more of it is revealed in order to avoid being inconsistent. Like the cosmic censorship hypothesis as well.
posted by vacapinta at 6:17 AM on March 4, 2014 [1 favorite]


It sounds a bit like panpsychism, but let's not go there.

It's SORT OF like panpsychism: the difference is whether we're talking about consciousness or willing. Nobody (I don't think) is saying that particles have phenomenal consciousness, just that they are radically indeterminate in their response to stimuli, that the fundamental basis of reality is precarious rather than stable.

Which is still dumb. But actually I find panpsychism a little more palatable than Whiteheadism just because panpsychists are more about forcing us to reconfigure our conception of human consciousness than they are about attributing radical intelligence to the universe. The rock's "consciousness" is just its responsiveness to the environment, which most involves falling and getting crushed and occasionally having a chemical reaction of some sort. The human's consciousness then becomes something like the rock's, only with more ways to respond. The famous example is the thermostat: is it conscious? Well, it knows if the environment is too hot or too cold, and it responds. Are we humans like complicated thermostats, or is the comparison inapt?

Just noticed that I seem to be a thermostat that detects stupid arguments about free will and responds: couldn't I, in principle, be replaced with a not-too-bright ELIZA program? I mean, just add a random variable like "Falls for the metaphysical trolls 80% of the time" and throw in some stuff about my family and you've got a pretty good facsimile.
posted by anotherpanacea at 6:18 AM on March 4, 2014 [5 favorites]


Why, Mr Anderson, why?
posted by lalochezia at 6:37 AM on March 4, 2014




A fully deterministic universe is not incompatible with free will. If the deterministic algorithm displays sensitive dependence on initial conditions (which it sure as hell seems to) then it can't be predicted because the system itself is the simplest model that can predict its own behavior.

This conclusion of chaos theory is far more devastating to traditional Western religious notions than anything ever discovered by physics or geology. It basically says that not only does God not know the fate of every sparrow that will fall, the only way even God could learn a sparrow's fate is to build the universe and let it run -- and He would be as ignorant of its ultimate conclusion as we are, until it actually concludes. Indeed, unlike traditional Christian metaphysics, this actually gives god a reason to create the Universe -- something that is conspicuously lacking in the more traditional accounts.
posted by localroger at 6:38 AM on March 4, 2014 [13 favorites]


anotherpanacea: The human's consciousness then becomes something like the rock's, only with more ways to respond.

I'd say consciousness is about experience or 'interiority', not responsiveness, otherwise sufferers of locked-in syndrome would be classified as unconscious.

localroger: If the deterministic algorithm displays sensitive dependence on initial conditions (which it sure as hell seems to) then it can't be predicted because the system itself is the simplest model that can predict its own behavior.

Being not able to foretell a behaviour's outcome is not the same as that behaviour being 'free'.
posted by Gyan at 6:44 AM on March 4, 2014 [4 favorites]


Maybe the universe is retrospectively deterministic. Like sausage.
posted by ian1977 at 6:52 AM on March 4, 2014


Being not able to foretell a behaviour's outcome is not the same as that behaviour being 'free'.

How do you tell the difference?
posted by localroger at 6:54 AM on March 4, 2014 [1 favorite]


FWIW I think all behaviour is determined, but free behaviour is behaviour that cannot be determined without psychological data.
posted by Segundus at 6:57 AM on March 4, 2014 [1 favorite]


I'd roughly agree with localroger whenever anyone identifies freedom with a measure of unpredictability, but..  All this debate over free will arises historically from religious bullshit, like afterlives and omnipotent gods, that sensible people ignore today, even sensible religious people.

Imagine you encounter say a moral proposition that historically religious people might differ about if they differed on free will. Afaik you'll actually know simply from your educated modern perspective which position is correct : Individuals are morally, and sometimes legally, responsible for the foreseeable consequences of their actions (free will), but crime should be treated epidemiologically because tha'ts the most effective approach (determinism).

See also The Triviality of the Debate Over 'Is-Ought' and the Definition of 'Moral' by Peter Singer.
posted by jeffburdges at 7:05 AM on March 4, 2014 [3 favorites]


How do you tell the difference?

Maybe you can't. But let's assume for a moment that humans have free will in this non-predictable sense. Then suppose Peter points a gun at Paul and says "Read this speech aloud" and dictates a speech.

The content of the speech may be freely created, let's suppose, in that we've assumed that Peter has free will. But that doesn't mean that the instrument of action, Paul, has free will. Paul's coerced in this.

So the existence of non-determined action in some part of the system, even if it does indicate free will, does not mean that an actor down the causal chain is free.

So it could be -- and in fact I think it is the case -- that the universe as a whole, or God if you prefer, has free will, but that no human does.
posted by shivohum at 7:13 AM on March 4, 2014


We debate this a lot in my Buddhist class.

1) Your current set of circumstances is the result of choices in the past.

2) In this one moment you have an infinite of choices.

3) BUT... out of this infinite possibility of choice, the specific choice you choose is based on your personality, habits, world view, values and current emotional state.

3a) And these things also depend on your previous choices.

therefore

4) we have no free choice, even though it totally feels like we do; there was an initial impulse a gabillion years ago and we are just running on energy: kinetic-potential-kinetic-potential-kinetic.....

Then we look at the monk and he gently laughs at us and says there is a choice in there; there are many many influences in our minds that sway us one way or another, but ultimately we still get to choose, and enjoy the consequences of our choices.
posted by St. Peepsburg at 7:14 AM on March 4, 2014 [3 favorites]


Do they mention actual neurology anywhere beyond the first article, or is it just more dancing about architecture?
posted by Pirate-Bartender-Zombie-Monkey at 7:16 AM on March 4, 2014 [1 favorite]


localroger: How do you tell the difference?

The former is a fact about epistemological faculty, the latter is about ontology.

Freedom can't be positively identified other than the intuitive sense of free will. One can simply think of all the conceptual ways for non-free behaviours to exist and discard all those possibilities as being the case, and by elimination, hold freedom to be the case.
posted by Gyan at 7:18 AM on March 4, 2014 [1 favorite]


Man, that article was going so well until the last couple of headings, when it went all 'You can't prove you don't have free will so lets make up some metaphysical bullshit'.

It's interesting to talk about the role of the 'observer' in quantum mechanics, something which is often overlooked. For those of you who don't know, in the standard interpretation of Q.M. a particle is in a superposition of all possible states until it's observed, when it collapses randomly into one state (i.e. in the article, the spin in along axis isn't defined until we make some kind of measurement, then suddenly it is).

What an observation actually means is an area of (philosophical, rather than physical) debate, and I was hoping this was heading towards a discussion of 'there's no such thing as an observer, as there can't be free will (defined as concious will somehow effecting the way particles act)', which sounds interesting, even if I'm not quite sure how that would play out.
posted by Ned G at 7:24 AM on March 4, 2014 [1 favorite]


But don't determinism and free will both exist along side each other peacefully within the Many Worlds Interpretation?
posted by qwickset at 7:27 AM on March 4, 2014


That’s why it doesn’t matter if God plays dice with the Universe

I suspect God is playing Cards Against Humanity with the Universe.
posted by Foosnark at 7:27 AM on March 4, 2014 [11 favorites]


I think this means the ideas of Epicurus on the motions of atoms were right after all. Take that, Cicero!
posted by GenjiandProust at 7:31 AM on March 4, 2014 [6 favorites]


I guess the question boils down to does god play dice with the Universe, or does s/he play pool?
posted by Twain Device at 7:36 AM on March 4, 2014


But don't determinism and free will both exist along side each other peacefully within the Many Worlds Interpretation?

Not really, do you mean Everett's many worlds? In that framework reality branches each time a quantum state is measured into parallel universes. So if you were to look at the entire tree of all possibilities you could pick the line you'd follow if you had 'free' choice, but there's no way of knowing which branch you're on as the other universes are causally closed to us. Also, the idea of free will is dependant on a paradigm that isn't fundamentally random, and I don't think you can just define a single route that you are you on, without positing something metaphysical (a soul or similar idea).
posted by Ned G at 7:37 AM on March 4, 2014


Didn't you read that FPP, Twain Device? Nobody plays pool anymore.
posted by Pirate-Bartender-Zombie-Monkey at 7:38 AM on March 4, 2014


Then suppose Peter points a gun at Paul and says "Read this speech aloud" and dictates a speech.

The content of the speech may be freely created, let's suppose, in that we've assumed that Peter has free will. But that doesn't mean that the instrument of action, Paul, has free will. Paul's coerced in this.


Paul is really not deterministic here though. He can choose not to read the speech -- perhaps convinced that Peter won't actually shoot him, or knowing that the gun isn't loaded, or that he may be able to smooth-talk his way out of this. Or he may begin reading the speech and then make a lunge for the gun partway through, or something like that.
posted by Foosnark at 7:41 AM on March 4, 2014


Not really, do you mean Everett's many worlds? In that framework reality branches each time a quantum state is measured into parallel universes....
I thought many worlds supported the probability wave collapsing, not a branch creating (i..e everything with a probability >0 happens)
posted by qwickset at 7:41 AM on March 4, 2014


I confess, haven't got to read it yet (at work), just reusing the worlds oldest physics joke.
posted by Twain Device at 7:42 AM on March 4, 2014 [1 favorite]


Foosnark: I suspect God is playing Cards Against Humanity with the Universe.
I never really understood ____________ until I found ____________
The Big Bang; Road head
posted by andycyca at 7:44 AM on March 4, 2014 [3 favorites]


What an observation actually means is an area of (philosophical, rather than physical) debate

I favor the idea that the wave function is forced to collapse when the computational expense of maintaining coherence becomes excessive. That would have rather amusing implications for the practicality of quantum computing.
posted by localroger at 7:46 AM on March 4, 2014 [1 favorite]


It basically says that not only does God not know the fate of every sparrow that will fall, the only way even God could learn a sparrow's fate is to build the universe and let it run -- and He would be as ignorant of its ultimate conclusion as we are, until it actually concludes.

Imagining that a transcendent god would experience time the same way we do seems like exactly the same category error as imagining God lives on a particular mountain somewhere.
posted by straight at 7:50 AM on March 4, 2014 [5 favorites]


At the moment what the bleep do we know? is showing in 31 parts on youtube. It is currently in a spin:Up state but your observation may vary. It's real state is half alive and half dead.
posted by bukvich at 7:56 AM on March 4, 2014


Imagining that a transcendent god would experience time the same way we do...

So this transcendent god, does it transcend math? Because information theory doesn't really say anything about time; what it says is that a certain amount of computation must be performed to calculate the system's state at a certain point. We call the progression of that calculation U(t+1)=f(U(t)) "time" but even if we don't "experience time the same way" something has to calculate f() and iterate the state of the universe.

What chaos theory says, quite convincingly, is that for some systems (of which the Universe appears to be a good example) there is no way to predict what U(t+1000) will look like without performing the calculation for all its intermediate states U(t+1)..U(t+2)... etc. So even though the system is finite and deterministic, it is its own simplest model and any model that can predict it is, in fact, the Universe.

So while God might not be trapped in the flow of time within the finite Universe as we are, he has no choice but to experience the universe itself the way we do, one moment at a time in linear order. He might get to take divine bathroom breaks and pause to catch the finale of Breaking Bad but he can no more peek into the future to catch a spoiler on our finale than we can. He can only do that by doing the computational work to make the future happen. And if he does that he has, by implication, actually created the Universe over that interval of time.
posted by localroger at 8:00 AM on March 4, 2014 [8 favorites]


If the deterministic algorithm displays sensitive dependence on initial conditions (which it sure as hell seems to) then it can't be predicted because the system itself is the simplest model that can predict its own behavior.

I don't think it's obvious that you would need to have 100% prediction accuracy of a system to say that a system doesn't have free will. A lot of real-world systems are basically noisy input -> deterministic behavior based on the input -> noisy output. It's the part in the middle where free will would be involved. You wouldn't have to be able to predict every specific movement of a given Roomba for example to figure out that Roombas in general follow a fairly basic algorithm to decide their movements with no free will involved.
posted by burnmp3s at 8:03 AM on March 4, 2014


I thought many worlds supported the probability wave collapsing, not a branch creating (i..e everything with a probability >0 happens)

As I understand it, and I'm by no means an expert, many worlds theory is a way of getting around the measurement problem. Basically, when we're talking about an unobserved quantum object, it's in a superposition of all it's possible states. So for a spin 1/2 particle we can say it has probability 'a' of spin up, and 'b' of spin down. When we observe the particle, we only observe it as up or down, but the probabilities a and b have given the chance we see it with this spin. So the question is, what's happened here? Normal Q.M would say that the wave-function has collapsed on observation, but it's tricky to know what this means. The Copenhagen interpretation, which is the most popular among physicists, can be paraphrased as 'don't ask difficult questions, we observe this happening with no deviation from our theories, so it's obviously just the way the universe works.' Many worlds interpretation gets around the idea of wavefunctions collapsing by saying both happen, the quantum particle is still in a state of spin up and spin down, it's just that we can only observe one of these states now, but in another parallel universe the other state is observable. So everything with probability >0 does happen, but the universe branches so we can't access it.
posted by Ned G at 8:13 AM on March 4, 2014


I don't think it's obvious that you would need to have 100% prediction accuracy of a system to say that a system doesn't have free will.

The problem is that this argument really only works in one direction. If something isn't free, like a Roomba, then that unfreedom will eventually become apparent, and we can check our work by making predictions. Scientific Method FTW.

But what about something that doesn't appear to be unfree? Suppose I program the Roomba to avoid taking the same path twice. It would still clean the floor but it would take a different route, defying all efforts to make any prediction other than that the floor would be clean within an hour or two. Is that Roomba free? Does it make a difference whether the Roomba gets its random paths from a 64-bit RNG which is deterministic but will never repeat within the life of the Universe, or from actual radioactive decay which even physicists can't predict? Suppose one of the random results is that it gets tired and refuses to clean the rest of the floor until tomorrow?

Ultimately we would only be able to prove such a Roomba is unfree by taking it apart and exposing the algorithm by which I had directed its actions. If we were hunter-gatherers entirely ignorant of electronics and programming, would it be possible to tell whether the Roomba was free? I suspect we'd be building statues of it.

So really, even if we can't achieve 100% predictability, "freedom" is in essence about the theoretical possibility of 100% predictability. The unfree Roomba can't be predicted by hunter-gatherers, but they might observe that it was built by humans and that there might be someone who knows more about its interior workings than they do. That both the Roomba and a human might have their courses altered by a rock in their path doesn't change the potential for one to be free and the other not.
posted by localroger at 8:15 AM on March 4, 2014 [1 favorite]


You can choose a ready guide in some celestial voice
If you choose not to decide, you still have made a choice
You can choose from phantom fears and kindness that can kill
I will choose a path that's clear
I will choose freewill
posted by Talez at 8:17 AM on March 4, 2014


The Lucretian swerve: The biological basis of human behavior and the criminal justice system

It is the author’s contention that a belief in free will is nothing other than a continuing belief in vitalism—something biologists proudly believe they discarded well over 100 years ago...

I suspect that we inherit a belief that free will is perfectly logical, and therefore not worthy of questioning. Note that the way we think is influenced by the inheritance of both cultural ideas (memes) as well as genetic material (21), and in some cases, ways of thinking may survive, somewhat in contrast to the logic, or lack of that is associated with that process. The way we in society think about free will (and religion) is likely to be an example of such a process—the line of thinking may have survival value, despite the fact that it is nonsensical and unsupported by any evidence.

posted by Blazecock Pileon at 8:21 AM on March 4, 2014 [4 favorites]


Many worlds interpretation gets around the idea of wavefunctions collapsing by saying both happen, the quantum particle is still in a state of spin up and spin down, it's just that we can only observe one of these states now, but in another parallel universe the other state is observable. So everything with probability >0 does happen, but the universe branches so we can't access it.

So one could argue that the probabilities of all current events are founded in the probabilities of previous events all the way back to the big bang, which is to say, if everything that could happen did/does/will happen, then those things have already been determined.

That begs the question, which branch path is an individual experiencing at any given time and why that one among an infinity of others? Couldn't one argue that this is determined (for lack of a better word), by an individual's free will/mindset/intentions/etc...
posted by qwickset at 8:37 AM on March 4, 2014


[T]he soul acts according to fixed laws; and is, as it were, a spiritual automaton (automatum spirituale).--Spinoza / On the Improvement of the Understanding.
posted by No Robots at 8:41 AM on March 4, 2014


Ultimately we would only be able to prove such a Roomba is unfree by taking it apart and exposing the algorithm by which I had directed its actions. If we were hunter-gatherers entirely ignorant of electronics and programming, would it be possible to tell whether the Roomba was free?

This is basically my point, that the heart of the matter of free will is the process by which actions are decided, not the actual ability of a third party observer to figure out that process. There is some set of actors in the world that we understand well enough to prove that they don't have free will, but not being able to figure out how a process works is not the same thing as that process involving free will.

So really, even if we can't achieve 100% predictability, "freedom" is in essence about the theoretical possibility of 100% predictability.

Again, I don't really think makes sense as a definition of free will. Suppose there is some source of actual non-deterministic randomness in the system that every actor in the system can use as part of their decision-making process, from humans to washing machines. Any deterministic actor that used such a source as part of their decision process would be fundamentally impossible to predict, because you could never predict that input. Does that mean that every randomized actor, even a mechanical device, has free will? And if one part of the system has free will, how do you prevent that from making any other part of the system have free will? For example if I have free will and because of that you can't predict the future position of my shoe, does that mean that my shoe also has free will?
posted by burnmp3s at 8:43 AM on March 4, 2014


Ok, so we're not mentioning Dennett or compatibilism? Got it. I'll show myself out.
posted by Mooseli at 8:46 AM on March 4, 2014 [2 favorites]


"Consider what effects, that might conceivably have practical bearings, we conceive the object of our conception to have. Then, our conception of these effects is the whole of our conception of the object."

-C.S. Peirce's pragmatic maxim

In conclusion, debates about free will are inconsequential.
posted by Pyrogenesis at 8:55 AM on March 4, 2014 [2 favorites]


Mankind's current understanding of the universe seems to suggest that we don't have free will. So, either we don't have free will, or we do and our current understanding is incorrect or incomplete.
Given that throughout history and without exception every previous attempt to understand and explain the universe has been incorrect or incomplete I'm going to assume that we have free will and maybe one day we'll be able to understand why we do.
posted by rocket88 at 8:57 AM on March 4, 2014 [2 favorites]


Tangentially related: Numberphile recently posted an interview with Conway focusing on his "game of life".
posted by Proofs and Refutations at 9:01 AM on March 4, 2014


So while God might not be trapped in the flow of time within the finite Universe as we are, he has no choice but to experience the universe itself the way we do, one moment at a time in linear order.

Even people in time don't have to perceive the past one moment at a time in linear order. I don't see why "the future" would be any different (or even have a meaning) for a being outside of time.
posted by straight at 9:01 AM on March 4, 2014 [2 favorites]


One problem I have with "particles taking free decisions":

Everything is made of, as far as we know, quarks, leptons and bosons.

Do those particles all "take free decisions?" Do neutrons and protons also? How about atomic nuclei? Atoms? Molecules? Proteins? Organelles? Cells? Organs? Organisms? Communities? Civilizations? Planets? Solar systems? Galaxies? Galactic clusters? The universe as a whole?

How do I get to exercise free will when my brain is made of smaller things that are all exercising free will, and those smaller things are made of yet smaller things exercising free will? Either I am at the wheel, or the wheel is turning itself.

Of course it's almost as bizarre to consider the opposite: that somehow, the coincidence of quarks, leptons, and bosons behaving as they must results in a Rube Goldberg machine that's currently typing all this dreck and posting it on MetaFilter.
posted by Foosnark at 9:04 AM on March 4, 2014 [1 favorite]


Foosnark: In a way I agree with you, I can't see a way how quarks and leptons following rules can create a being with free will. In order to do that conciousness (which is weird, but not magical in my opinion) would have to be able to effect the rules that the quarks and leptons follow. As this seems so unlikely as to be almost impossible to me, I don't think free will can exist, despite the illusion persisting.

Qwickset: It's also possible to look at the problem from the other direction too, that there's a version of you in a parallel universe thinking they are the true qwickset experiencing the only reality that they know about.
posted by Ned G at 9:30 AM on March 4, 2014


All this casual jumping from quarks to a concept without further ado is just so... it's like a biologist who says: "Look, here's a carbon atom. Therefore there is no such thing as "organism". QED." Apparently there's nothing between an elementary particle and a philosophical concept that's worth noticing. They are basically the same thing!
posted by Pyrogenesis at 9:38 AM on March 4, 2014 [1 favorite]


Re: Dennett, there was this long discussion a month or so ago coming at the free will question from another angle. People who are interested in this might be interested in that.
posted by LobsterMitten at 9:40 AM on March 4, 2014 [1 favorite]


Ned G: Yes, this is exactly what I understand the MWE to imply.
posted by qwickset at 9:52 AM on March 4, 2014


These arguments always seem to assume that ideas like causation and natural law are completely clear and settled fact. They are just as problematic philosophically, in my opinion, as is the will.
posted by thelonius at 10:01 AM on March 4, 2014


So, back to TFA,

Am I reading this incorrectly, or are they basically just saying that because universal path dependency is impossible, no determinism can hold? Why can't we just say that within a given frame, things can behave deterministically?

I'm having trouble conceptualizing how any of this actually relates to free will in any meaningful sense.
posted by graphnerd at 11:31 AM on March 4, 2014 [1 favorite]


Any deterministic actor that used such a source as part of their decision process would be fundamentally impossible to predict

This completely misses several points. As a thing with an independent existence (as opposed to just being part of "you") your shoe can be shown never to do much of anything unless it's following your foot.

I find this argument has an almost 1:1 correspondence with the potentially much simpler question, "Does an extent of the Mandelbrot Set exist if nobody has ever bothered to set a computer traversing it?" Like our future actions (assuming the Universe is deterministic) the entire Mandelbrot Set is implied by the ridiculously simple program that draws it. But you cannot inspect the Mandelbrot algorithm and intuit what the set will look like; to actually see what a part of the fractal looks like you must do calculations to realize it.

So any five people who all set their computers to traversing the same bit of the Mandelbrot fractal will all get the same pre-determined result, but the first to do so will have no way to have any idea what the result will look like. In this sense the "free will" of Mandelbrot calculation is the choice of extent. You can pick an area that's likely to be all zero, or an area that's likely to be all one, or one of the more interesting edge areas. That decision transcends the fractal itself; the set might be fixed, but it's also potentially infinite, and the decision to realize a certain part of it would reflect the uncertainty of a free being doing exploration.

A similar situation might exist with the Universe. Whatever it might mean for god to "perceive" something, if god has access to U(t+1000) it implies that either god or its personal computer or cloud data service started with U(t) and performed the calculations to iterate it out to U(t+1000), in order, and without skipping steps. There is no other way for U(t+1000) to ever exist. Now, god's reasons for setting this calculation in motion could be very similar to the mathematician's for setting his computer to realizing some corner of the Mandelbrot Set; and it's one possible answer to the question "How did all those magic numbers in physics get their values."

Now one could argue that the elements of such a computation are still unfree beads on a wire, but I think when you get to the point where even god can't see the wire the difference ceases to have any meaning.
posted by localroger at 11:58 AM on March 4, 2014


Am I reading this incorrectly, or are they basically just saying that because universal path dependency is impossible, no determinism can hold? Why can't we just say that within a given frame, things can behave deterministically?

Well I think that's only the Kochen-Specker theorem. In contrast, the free will theorem is a logical argument that uses KS to place a consistency constraint on what any concept of "free will" is allowed to be. One philosophical consequence is that if you "believe" in free will, this theorem should make you to reexamine what you think you know about it, etc, otherwise you are just being philosophically unsound, etc.
posted by polymodus at 12:26 PM on March 4, 2014


Also, I hadn't read to Part III but there's a mini section discussing a conceptual nuance between freedom, and the randomness/determinism dichotomy.
posted by polymodus at 12:33 PM on March 4, 2014


So even though the system is finite and deterministic, it is its own simplest model and any model that can predict it is, in fact, the Universe.

So you're saying that if God were a being that could actually model the entire Universe in his mind, that this model would, in fact, be the Universe?

Johnathan Edwards (who ought to be famous for a lot more than Sinners in the Hands of an Angry God) wrote an essay on the Trinity that makes a similar assertion. Edwards argues that if God knows himself omnisciently, that would require God to have a 1:1 model of himself in his mind, a model which would be identical with, which would be God. He says it is in this sense that the divine Word of God exists at the same time begotten and yet equal--of one being--with God.
posted by straight at 1:37 PM on March 4, 2014 [1 favorite]


straight, that's a very interesting point. I find it amusing to contemplate God recursively being equivalent to His conception of Himself at an infinite, and possibly circular, range of levels.

For me, a question which is more fundamental than this one of free will is "what is the Universe made of." Is it particles, waves, mathematical expressions, ideas in the mind of God? The one thing we can say with some certainty -- both the ONLY thing and a SURE thing -- is that the Universe contains information. We may be unsure of everything else, and even of how much, what kind, or how reliable that information is, but a universe in which we can remember things and correlate those memories with new experiences is a universe which itself contains information, and also in which we personally contain information.

And that says some surprisingly robust and significant things about what the Universe (and we, whether we are "part of it" or not) can and cannot do.

I have on occasion made the point that Judaism and its derivatives (including Christianity and Islam) are the second of three tiers of metaphors about the Universe. The first tier, encompassing nearly all pagan religions, models the universe as a living thing or product of living things, with the universe itself being born or hatched or otherwise created through some analog-biological process. The interesting thing about this view is that it does not endow the creator with inevitable powers; biological mothers, after all, do not have omnipotent control of their offspring.

In the Judeo-Christian metaphor, though, the universe is a made thing, like a sword or a pot. It is the product of skill and craft and thus implies the existence of a craftsman; further, it only performs as well as its builder's skill allows, and as long as its creator actively uses it. It is from this that we get the idea that god must must have vast skills and knowledge, down to "every sparrow that falls," because if sparrows are made things like pots then they do nothing at all unless the potter continuously animates them.

The third tier, which has not yet arrived, will have to recognize that systems made of information are capable of being "greater" (for some axis of "greatness") than their creators. Even today we as mortal humans are now capable of creating things which do things we cannot do and act in ways we cannot predict; why would we consider god to be less capable than we?

This is of course a deeply heretical view which has terrified the keepers of religous thought for centuries; the creation of clockwork automata which could march in circles for a few hours without human input was so terrifying they were made illegal in some medieval places.

After Turing and Mandelbrot, we have to consider the idea that god might not only be a clockmaker, he might have made a clock even he does not completely understand, and its ultimate working could be as much a surprise to him as it is to we as part of the clockwork. It's only a crazy idea if you have never written a computer program that didn't do what you expected.
posted by localroger at 2:18 PM on March 4, 2014 [5 favorites]


I'd say consciousness is about experience or 'interiority', not responsiveness, otherwise sufferers of locked-in syndrome would be classified as unconscious.

I just saw this: I'd say that experience is a kind of response. That is, if I have locked-in syndrome, but seeing a movie still entertains or bores me, then yes, I'm responding, even if I'm not responding verbally. You could almost certainly detect that response on an MRI, and even if we couldn't we'd only claim that the person was conscious by saying that there was an "undetectable" reaction, that is, by assuming measurement error. I don't think it would make sense to say that someone could be conscious without having that internal reaction, that experience, of the world or themselves or a dream or something. In contrast, if I'm in a coma or a vegetative state, then I'm *not* responding to the environment, and then I really am not conscious. I may have interiority in that case, though I don't quite know what that means and I suspect it's a meaningless concept: I can't have an interiority without some relationship to an exteriority, since that transaction is what creates experience. Even in reflection I respond to myself, take a part of myself (an idea or an impression, a memory of some external thing) and set it at odds and start a conversation with it.

I assume that interiority as such, "radical" interiority as it were, would be an inside with no outside or internal variegation, and while I can say things about that in a metaphysical mode, pretending to be a Christian or an enlightened meditator, I doubt that those things actually represent states of affairs.
posted by anotherpanacea at 2:29 PM on March 4, 2014


As a thing with an independent existence (as opposed to just being part of "you") your shoe can be shown never to do much of anything unless it's following your foot.

So? You said that "freedom is in essence about the theoretical possibility of 100% predictability". The fact that you can predict the shoe's location when it's not on my foot doesn't prove anything. Being able to predict its behavior for the percentage of the time that it's not interacting with something that has free will is not the same thing as being able to predict it 100% of the time.

A similar situation might exist with the Universe. Whatever it might mean for god to "perceive" something, if god has access to U(t+1000) it implies that either god or its personal computer or cloud data service started with U(t) and performed the calculations to iterate it out to U(t+1000), in order, and without skipping steps. There is no other way for U(t+1000) to ever exist.

I'm not disagreeing with any of this, in fact my concept of the notion of free will doesn't really involve the question of an omniscient being observing the universe because I don't think one exists.
posted by burnmp3s at 2:33 PM on March 4, 2014


Conways dismissal of Randomness (on part 3 of the article) doesn't make sense to me.

Proposal: The actions of the particles can’t be functions of past history, but that still allows that they could be random at the time.

Conway's answer:
part 1) The way backgammon tournaments are run is a way of showing that there's no real difference between a random number chosen at a specific time, or a random number pre-chosen and then read off a list at a later time.
part 2) Therefore the actions of the particles can't be explained by randomness.

I have not idea how he is getting from (1) to (2) there.
posted by memebake at 4:53 PM on March 4, 2014 [1 favorite]


Last months free will thread

From the last thread, I think I finally worked out what I think of the free will thing. I don't really know what I think until I try and write it down though so here goes:

Whether the universe is fully Deterministic or involes Randomness doesn't have anything to do with Free Will. But I think reflecting on Determinism makes people very aware that there appears to be little room left for 'Free Will' (in the traditional sense). But if you think about it carefully, exactly the same can be said for a non-deterministic universe where genuine 'randomness' occurs. Because randomness also doesn't leave any room for traditional 'Free Will'.

Determinism is not the same as predictability. If the universe is a giant rude-goldberg machine evolving deterministically, it is still literally unpredictable. Because to 'model' the universe and try and see where it was going, you would have to simulate every atom and sub-atomic particle and every photon etc and you would need a machine bigger than the universe itself. So the universe may be deterministic, but its like a one-off program that has never been run before - no-one can tell the details of where its going to end up.

So if the universe is a giant super complicated web of interconnected cause and effects, whats left for free will?

Trouble is the traditional concept of free will is hopelessly confusing. I think the compatibilist answer is to redifine free will as something like:
a) everything that happens is the result of a giant super complicated web of cause and effect
b) that includes your brain as well as everything else in the universe
c) draw an imaginary dotted line around your skull
d) look at the things you do as an organism, and measure what proportion of the 'causes' for those events come from inside your skull as opposed to outside your skull.
e) If most of the events are being caused from inside your skull then lets call that "free will".

Hence with this definition, we can say humans have "free will", and so do ants probably, but that biro on my desk doesn't.

--------------
the Free Will Theorum that is the subject of this FPP seems fairly mundane:
"if we have free will then so do the particles"
the opposite also applies:
"if particles do not have free will then neither do we"
Its talking about the traditional magical mysterious property called Free Will that is neither determinism nor randomness. And it simply says that if we have it then so must all particles. The compatibilist view would be the other way: That concept of magical Free Will is hopelessly confused and impossible to define, the particles clearly dont have it and so neither do we.
posted by memebake at 5:14 PM on March 4, 2014


burnmp3s: I'm not disagreeing with any of this, in fact my concept of the notion of free will doesn't really involve the question of an omniscient being observing the universe because I don't think one exists.

Your comment made me realise something: The reason that omniscient beings keep cropping up in these arguments is that Determinism is only relevant from the point of view of an omniscient being. If you're in a Deterministic universe and _not_ omniscient, then the determinism is irrelvant to your experience of things.
posted by memebake at 5:35 PM on March 4, 2014


is God a Taoist?
posted by RobotVoodooPower at 7:37 PM on March 4, 2014


anotherpanacea: I can't have an interiority without some relationship to an exteriority, since that transaction is what creates experience.

Dreaming?
posted by Gyan at 8:32 PM on March 4, 2014 [1 favorite]


Localroger wrote:
a certain amount of computation must be performed to calculate the system's state at a certain point[....] for some systems (of which the Universe appears to be a good example) there is no way to predict what U(t+1000) will look like without performing the calculation for all its intermediate states U(t+1)..U(t+2)... etc.
I hope I didn't mess up your meaning by abbreviating this.

I don't think we can actually be sure that there aren't algorithmic shortcuts to every problem. We haven't found them, but if they exist they would necessarily be known to an omniscient being. In any event, you're constructing a straw deity. Omniscience means that the result of a calculation is known before it is calculated.

As an analogy, consider an old-fashioned celluloid movie. Take Casablanca, for instance. Rick and Ilsa's relationship changes and deepens over the course of the movie. That change takes time; it is the process of development that makes it meaningful. In the movie, Rick has free will: his choice isn't inherent, but something that he arrives at through contemplating the circumstances presented to him. For us, though, Rick's choice is a static thing: we are outside the movie and can look at any part of it without changing it.

That's the sort of viewpoint implied by omniscience. I'll stipulate that a sufficiently sensitive physical problem may be incalculable: the only way to know Thursday's weather is to wait for Thursday and experience it. But that doesn't mean that God acting on Monday is ignorant of the weather on Thursday. He's aware of Thursday's weather, just as we are aware of Rick's state of mind at the end of Casablanca.
posted by Joe in Australia at 1:59 AM on March 5, 2014


Dreaming?

Is that interiority? When I dream I dream about people and places in the outside world or in fiction, which are only available to dream about because of experiences I had with external objects.
posted by anotherpanacea at 4:34 AM on March 5, 2014


Experiences you had with exteriority. The dream itself does not involve transactions with external objects, only references to them, unlike seeing which requires transactions with photons i.e. external objects, in real time.
posted by Gyan at 5:18 AM on March 5, 2014 [1 favorite]


It seems like your claim is that interiority is just time-shifted exteriority, which is fine, but it's not ipseity and it does seem to involve experience and reaction, which I thought you claimed it didn't.

Basically, I'm not sure why it doesn't involve transactions with external objects if, as you say, it does involve prior transactions with external objects.
posted by anotherpanacea at 5:29 AM on March 5, 2014


My claim is that experience is a form of interiority and that's the hallmark of consciousness. Dreaming is contingent on some earlier involvement of exteriority but the process of dreaming itself does not require transaction with exteriority

Just to be clear with terms, interiority means first-person phenomenal experience and exteriority just refers to the external world and its contents i.e. a term to define the negative space of all that is held to exist but outside the mind.
posted by Gyan at 5:48 AM on March 5, 2014 [1 favorite]


I am happy with calling consciousness "first-person phenomenal experience." The word philosophers usually use for this is "qualia," not "interiority" precisely because "interiority" seems to only have meaning as a part of the binary with "exteriority" and that doesn't seem to be how the experience works: rather, experience is a transaction between inside and outside, or between inside and inside (where the inside is an import from the outside.) If interiority means phenomenal, then does exteriority mean "non-phenomenal"? That doesn't seem right.

I worry, though, about this: in what sense is "phenomenal" a metaphysically loaded term? In what sense does it assume that human experiencing is different in some categorical sense than other sorts of experiencing or transaction, like the thermostat? Are we justified in assuming that?
posted by anotherpanacea at 6:05 AM on March 5, 2014


qualia is a property of a phenomenal experience, the redness of red, not the fact that there is such a thing as a visual experience of colour. Interiority is chosen to signify that the conscious experience is privileged to the experient and hence is "interior" to them.

The thermostat, putatively, does not experience warmth or cold. The mercury rises/falls, or the resistor function..etc changes in ways described by physics. But the thermostat is not 'responding' to a sensory affect.
posted by Gyan at 6:25 AM on March 5, 2014 [1 favorite]


I agree with your first paragraph, I just don't think the term "interiority" is necessary to capture "first-person-ness" and I worry that it creates an artificial dichotomy. But your first paragraph doesn't support your second paragraph. Certainly, people who are dreaming and locked-in do have experiences. Those experiences are in the first-person, and those experiences themselves constitute responses, just as I can have an affective response to an artwork before I actually verbally say what I think. In that sense, responsiveness and what you're calling interiority are compatible.

I'm not sure that any of that resolves whether, in fact, there's something it's like to be a thermostat. You can claim there isn't, as you do here, but I don't think you can do so authoritatively, for all the problems that arise when you try to get at the "problem of other minds" generally. That's why panpsychism continues to hold some interesting challenges, while the original subject of the post does not.

I recommend Eric Schwitzgebel's paper "The Crazyist Metaphysics of Mind." No kidding, he's doing good work showing how our intuitions and common sense on these topics tend not to resolve into a coherent theory of mind, so our kneejerk rejection of, say, panpsychism isn't as dispositive as we would like. I much prefer crazyism to mysterianism, if only because McGinn is a jerk, but also because it doesn't deny the possibility of progress, just alerts us to the challenges.
posted by anotherpanacea at 9:47 AM on March 5, 2014


Your comment made me realise something: The reason that omniscient beings keep cropping up in these arguments is that Determinism is only relevant from the point of view of an omniscient being. If you're in a Deterministic universe and _not_ omniscient, then the determinism is irrelvant to your experience of things.

The other reason it comes up is when people try to distinguish different kinds of free will and whether a deterministic universe rules them out.

Some would say that if--in principle--an omniscient deity could predict your actions perfectly based on knowledge of the current state of the universe, this implies things about the causes of our decisions and whether we can be said to have free will, even if there is no such deity.

There is a separate issue where some would say that an omniscient deity could--in principle--know your future actions, not because it's possible to predict them based on the current state of the universe but because, existing apart from time, the deity observes them in the same way the deity observes what is, for us, the present and the past. And there are arguments about what implications this would have if true (or even if it's just theoretically possible regardless of the actual existence of the deity). (One version of this is referred to as eternalism or the "block universe" theory.)

I think Joe in Australia and I were pointing out that localroger was confusing these two different senses in which it might be possible for an omniscient deity to know what is, for us, the future. (Although he may believe, as some people do, that if one of these were true, the other would also necessarily be true.)
posted by straight at 10:27 AM on March 5, 2014 [1 favorite]


thanks straight, reading this thread it seems to switch between omniscient beings as a tool for thinking about determinism, and theological-type discussion of what omniscient beings may or may not be able to do, which is confusing.

Some would say that if--in principle--an omniscient deity could predict your actions perfectly based on knowledge of the current state of the universe, this implies things about the causes of our decisions and whether we can be said to have free will, even if there is no such deity.

Yes, and what localroger says above is that to do this trick, the omniscient being would essentially be running a model of the universe in fast forward to see how it was going to unfold. But he (rightly) in my view points out that such a model of the universe - if it was going to be properly accurate - would be just as detailed as the universe itself, and hence the model of the universe becomes indistinguishable from the real one. The 'simulated' people inside the model would have the same experiences as the 'real' ones, etc etc. (an idea that often crops up in sci-fi as it happens).

There is a separate issue where some would say that an omniscient deity could--in principle--know your future actions, not because it's possible to predict them based on the current state of the universe but because, existing apart from time, the deity observes them in the same way the deity observes what is, for us, the present and the past.

Right, but here we seem to switch from using an omniscient being as a tool to explore determinism etc to just making things up about imaginary beings. I mean, I could posit an omniscient being that has a car made of bananas, or that can change the laws of logic at will so that it wins any argument. But that doesn't seem relevant to anything except exploring 'what kinds of things can we imagine if we go hog wild'.

Joe in Australia has another slightly different counterargument to localroger

I don't think we can actually be sure that there aren't algorithmic shortcuts to every problem. We haven't found them, but if they exist they would necessarily be known to an omniscient being

That's an interesting one. To me (degree in maths etc) it seems self evident that there is no algorithmic shortcut to predicting the unfolding of a chaotic/dynamical system. But I'd have to hunt around to see whether there's some proofs of that. There are proofs that head in that sort of direction, like the Period 3 Implies Chaos one. But, I dunno, I find 'there might be secret shortcut to calculating {massively complex chaotic system}' to be a very very very big stretch.
posted by memebake at 10:52 AM on March 5, 2014


anotherpanacea: those experiences themselves constitute responses

If you are including experience under responses, there's no disagreement, but that's an unusual use of the term.
posted by Gyan at 11:04 AM on March 5, 2014


In addition to what memebake said it is worth asking why we think an "omniscient being" is a reasonable thing to think exists at all.

This goes back to that "god as maker" tier of religious metaphor; the natural world is full of vast amounts of detail, and if it was created in the manner of a pot or sword then this implies a vast amount of effort and skill on the part of a creator. Such a creator would obviously have to be much vaster than the universe, and capable of keeping all those details straight. He must also be a pretty swell guy to have put all that work into making us and our world, who wouldn't want to buy him a beer?

But one of the largest implications of chaos theory is that the universe is quite capable of looking the way it does without the effort of such a vast and conscious creator; details can self-organize out of relatively simple systems in ways that even a conscious creator cannot predict, even in a fully deterministic system. This is the exact memo which so terrified those leaders who made clockwork automata illegal, because it removes all the arguments for the creator to be so much better than us in every conceivable way.

This is only a thing which has really been known to anybody since the 1980's, and to people outside of certain mathematical specialties since the 1990's. It basically renders every argument which has ever been made about the nature of god moot. God could just be the equivalent of a bored teenager fiddling with a system he didn't build himself and doesn't really understand. There is no reason to suppose that god even understands how the universe works, much less is comparable to it in scale.

Having control over the magic numbers and the off switch are of course important powers if such a being exists, but they do not imply much else about the guy.
posted by localroger at 12:00 PM on March 5, 2014 [1 favorite]


Right, but here we seem to switch from using an omniscient being as a tool to explore determinism etc to just making things up about imaginary beings.

No, it's a question about the nature of time and of the universe. Is the future "fixed" in the same sense that we consider the past "fixed"? (And what, if anything, do either of those statements really mean?)
posted by straight at 2:09 PM on March 5, 2014


I don't know much about information theory. Do we have an agreed-upon account of what it means to know something? That seems like a necessary precursor to having an opinion about whether a being could know everything.
posted by straight at 2:16 PM on March 5, 2014


straight, we can calculate (and sometimes measure) the amount of information represented by a system or which can be transmitted over a channel. "Know" is a pretty fuzzy word, but the transmission bandwidth of a channel is surprisingly sharp. One of the first things to emerge from Claude Shannon's work was that the radio spectrum is a finite resource, which is a thing that was suspected but not really known previously.

Information theory doesn't have much to say, though, about whether information is "accurate" or whether it represents "knowledge." It says with some authority that the observable Universe is finite -- it encompasses a finite dimensional span, and objects within it can only be measured to finite accuracy. Several people, including myself as a teenager, have arrived at a figure of 1081 bits give or take an order of magnitude or two as an estimate of how much information the Universe represents. It could, of course, be much larger than that but that's all that we can even theoretically access. It's all that we can reliably know exists (and we don't really know that it exists all that reliably; we can say with some confidence the part we could ever access isn't bigger, but it could actually be much smaller if it was playing any of a variety of tricks we'd never be able to detect on us).

As for whether information exists at all, the practical standard for that is to compare two stores to see if they're the same. This requires two stores, of course; we could take for instance the Universe itself as we perceive it and our mental model of it. Inasmuch as these two things align information can be said to exist; when they diverge information can be said not to exist or to be faulty in at least one of the two stores. As it happens most of us store a vast amount of information about the Universe which regularly aligns with our expectations, but most of us also experience faults which suggest that there are errors. Whether those errors are in our mind or in the Universe is a question that leads to some very interesting contemplation of things like Occam's Razor.
posted by localroger at 5:29 PM on March 5, 2014


memebake wrote: a model of the universe - if it was going to be properly accurate - would be just as detailed as the universe itself, and hence the model of the universe becomes indistinguishable from the real one.

The Conway-Kochen theorem (described in the FPP) says that even someone with a fully detailed model of the universe cannot be omniscient regarding future events. Particles "choose" their 101 spin at the moment they are measured, and even an omniscient being can't know the result of a measurement that hasn't taken place. So there is no such thing as omniscience–by–simulation when the outcome depends on the result of a 101 spin measurement.

But would a model of the universe be in any sense "real"? Leaving aside the question of whether it could be used to make predictions, are you suggesting that a model of someone's mind would "be" that person? Would it be wrong to shut the simulation down? If we stipulate that it would be wrong to wantonly kill an ant for no reason, would it be equally wrong to kill a simulation of an ant?
posted by Joe in Australia at 5:40 PM on March 5, 2014


An omniscient being could know all events even if such events were based on randomness if time were unreal and past, present, and future already eternally existed and such a being could see the future since in some sense it tenselessly is.

An omnipotent being could know all events even if such events were based on randomness because it could control them.

Who chooses how the roulette wheel of randomness spins? If that is an intelligence (as it almost certainly is), it too can see and control everything.

Just some musings.
posted by shivohum at 5:50 PM on March 5, 2014


Who chooses how the roulette wheel of randomness spins? If that is an intelligence (as it almost certainly is)

Um that's a pretty big leap there. While I am somewhat amenable to the idea that randomness is what might be considered a security vulnerability in that it could be controlled in nonrandom ways that would pass muster with any error-checking going on, that's not at all certain and for the most part it still has to be random to act believably random.l
posted by localroger at 7:19 PM on March 5, 2014


An omnipotent being could know all events even if such events were based on randomness because it could control them.

Yeah yeah but we're not here to do theology. We're talking about omnipotent/omniscient beings because we're trying to explore free will/determinsim etc

Please replace omniscient being with "giant supercomputer sitting outside the universe somehow" if that makes it easier. Its all about imagining a being that could exploit determinism and seeing where that leads. But if you make you're imaginary being too powerful (e.g. it can transcend math or whatever) then the thought experiment breaks down.
posted by memebake at 1:52 AM on March 6, 2014


Joe in Australia: But would a model of the universe be in any sense "real"? Leaving aside the question of whether it could be used to make predictions, are you suggesting that a model of someone's mind would "be" that person? Would it be wrong to shut the simulation down? If we stipulate that it would be wrong to wantonly kill an ant for no reason, would it be equally wrong to kill a simulation of an ant?

Sortof off topic, but yes if you accept the Hard-AI view of consciousness (and for me its either that or dualism and I don't feel much on the dualism argument) then it follows from that that the treatment of AIs and similations becomes an ethical issue.

Sometime in the next few centuries, treatment of sophisticated AIs might become a civil rights issue. E.g. what if super-realistic AIs are used in computer games? At a high enough level of fideltiy, simulated suffering would become indistinguishable from real suffering. If a dying person uploads their consciousness to a computer, do they still get a vote? What if they make 100 copies of themselves? It all gets quite complicated.

See Accelerando by Charles Stross, or Crystal Nights by Greg Egan, or (terrifyingly) Surface Detail by Iain Banks for explorations of these themes....
posted by memebake at 2:00 AM on March 6, 2014


straight: There is a separate issue where some would say that an omniscient deity could--in principle--know your future actions, not because it's possible to predict them based on the current state of the universe but because, existing apart from time, the deity observes them in the same way the deity observes what is, for us, the present and the past.

me: Right, but here we seem to switch from using an omniscient being as a tool to explore determinism etc to just making things up about imaginary beings.

straight: No, it's a question about the nature of time and of the universe. Is the future "fixed" in the same sense that we consider the past "fixed"? (And what, if anything, do either of those statements really mean?)

OK, I think I follow this. You're kindof throwing 'what is time?' into the equation and using the pretend omniscient being to explore that concept as well.

Some serious people argue that Time is an illusion (Douglas Adams adds: 'Lunchtime, doubly so') and the 'universe' is just made up of a continuum of frozen instants and we just think we're experiencing time because we're stuck in the middle of it all. Its actually very hard to argue against a metaphysics like that.

But in this debate we're exploring our own apparent sense of 'free will', what it really means, and how it might work in a determinstic universe. Ultimately the debate is about the human point of view. Throwing 'time might not exist' into the debate is not a foul move as such, but its a bit like bringing up solipsism or something: its a metaphysical trump move that can't be counterargued and pulls the foundations out of the debate so everything becomes moot. Time appears to exist for us, as in the future doesn't seem to exist until we get to it, so we're exploring free will and determinism from that point of view.
posted by memebake at 2:19 AM on March 6, 2014


memebake: the 'universe' is just made up of a continuum of frozen instants and we just think we're experiencing time because we're stuck in the middle of it all

We don't experience all the instants at once. We experience a slice and then "progress" to the next one. So Time recovered, kinda?
posted by Gyan at 2:33 AM on March 6, 2014 [2 favorites]


We experience a slice and then "progress" to the next one. So Time recovered, kinda?

Julian Barbours theory is that the instants exist independently of each other without any causality, so there isn't really any 'progress' from one to the next - even though we think we can detect causal patterns. Meta-physics that do away with time have to made radical changes to cause and effect as well.
posted by memebake at 2:59 AM on March 6, 2014


Progress here just means changeover, not development of cause-effect. Even if the series of Nows are not related to each other, you only experience a finite set of Nows before "moving on" to the next set.
posted by Gyan at 3:56 AM on March 6, 2014 [1 favorite]


Time appears to exist for us, as in the future doesn't seem to exist until we get to it, so we're exploring free will and determinism from that point of view.

Also, the universe really strongly resembles a dataset which is processed by an iterative function; again, stuff nobody appreciated until the 1980's. This would mean that time does have a direction, defined by computation, and that there is no sense in which "all moments exist simultaneously" as one moment has to be generated before it can serve as the source for the computation which generates the next.

Now this "computation" might be performed by some really simple analog system (particles bouncing around under the influence of simple rules, for example) but if someone wants to make the case that this is not how the universe is constructed, they have an awful lot of explaining to do with regard to why it looks the way it does at macro scale.
posted by localroger at 7:26 AM on March 6, 2014


« Older The Collected Wisdom of Bill Murray   |   The Audience is Listening (when you're done with... Newer »


This thread has been archived and is closed to new comments