Free Will

[eSOANEM]

I'm going to jump in and echo gmalivuk that free will requires some third path.

An observer can, in principle, predict with certainty the outcome of a deterministic process.

An observer can, in principle, predict with certainty the probability distribution of outcomes of a random process.

An observer cannot, in general, predict with certainty any such information about a process which has free will because, were they able to, the decision would clearly not be free.

If you have something that looks like the latter though, Occam's razor will tell you that really you just don't understand the process properly and it is, in fact, either deterministic or random. The alternative is to have some hidden variables and processes; as these are inherently unknowable and unfalsifiable no such theory can be thought of as scientific.

So no, as a scientist, I cannot say free will is real. As a person however, I'll say that it's a lot easier to get on with life assuming that it is

[Tub]

An observer can, in principle, predict with certainty the outcome of a deterministic process.

To prevent anyone from getting the wrong idea, I think it's worth pointing out that determinism does not imply predictability. For any observer inside of the deterministic process we call "universe", no accurate prediction is possible. Laplace's demon cannot be built. Not even in principle.

First, there is no way to determine the current state of the wavefunction. Measurement is just entanglement, and that doesn't give us access to the data we need.
Then we cannot build a computer to store the wavefunction. A computer large enough to store the state of the universe would need to be larger than the universe. And it would be slower, requiring more than 1 hour to run 1 hour of simulation.
But even if you could solve those problems with compression and optimization, you'll end up having the computer inside the simulated universe, so the computer needs to simulate itself and then it gets stuck in an infinite recursion.

The smallest and fastest simulation of our universe is the universe itself. To find out what happens tomorrow, the fastest method is to wait 24 hours.

[Copper Bezel]

Yeah, that's the big difference to me. Ignoring stochasticity, because it's more obviously irrelevant....

Determinism is not fate, it is not an outside force, and it's still orthogonal to any question about "free will" that has meaning on a human level and 99% of the arguments historically that have existed about it. Sophocles and Calvin weren't worried about computability, they were worried about gods. The future in a deterministic universe is predetermined in the sense that if you could restart the universe with identical initial conditions, it would turn out the same way. But nobody's done so and the future isn't known or knowable. It literally has yet to be determined. Whether or not you include stochasticity or ghosts and goblins. The "determine" in "determinism" is a technical meaning devoid of any human-scale significance. The "third factor" sense of "free will" is a transitional concept in philosophical thinking that started as theological thinking and hasn't shaken out all the ghosts and goblins yet.

[ucim]

"Free will" presupposes a thing to have this free will. Think about what this implies... whatever this thing is, you are thinking of it as a thing; that is, holistically. To say "Fred has free will" or "an answering machine has free will" you are thinking of Fred or the answering machine as a kind of indivisible entity that makes decisions of its own accord. That is, it is not a puppet on (external) strings.

The "external" part is important. Whatever it is that is making Fred, an answering machine, or any other thing do what it does must be internal to the entity for "free will" to be a possible thing to happen.

Spoiler:

For those that like to think of souls as the source of free will, it means that this entity would have to posess and embody a soul rather than merely be associated with one. If this putative soul were "outside" (not part of) the entity, then the soul would be a puppetmaster, and Fred would just be dancing to his soul's strings. Fred himself wouldn't have free will.

In any case, souls are problematic for many reasons, not the least of which is that it palms the question off onto something else.

So long as you think of the entity as a "point source", you can think of it as having free will and not get into trouble. But once you look under the hood and examine the mechanism it uses to make decisions, you'll find that it's on rails. The components all dance to the laws of physics.

But the components are not Fred, and that's crucial to the idea. Likewise, an answering machine is more than the sum of its parts.

Once you start looking into the mechanism, you are no longer talking about "Fred" or "an answering machine", even though you might phrase your thoughts and statements as if you were. Free will (as a concept) stops making sense at that point. There's no longer a singular "thing" to have it.

I'll just toss out these links for you (topics here on xkcd)

Definition of free will
Can a computer have free will?

Jose

[madaco]

If this putative soul were "outside" (not part of) the entity, then the soul would be a puppetmaster, and Fred would just be dancing to his soul's strings. Fred himself wouldn't have free will.

Err,

Wouldn't the standard response to that be that the soul is what we are referring to as "Fred" when we say that "Fred has/doesn't-have 'free will' " ?

[eSOANEM]

↶

An observer can, in principle, predict with certainty the outcome of a deterministic process.

To prevent anyone from getting the wrong idea, I think it's worth pointing out that determinism does not imply predictability. For any observer inside of the deterministic process we call "universe", no accurate prediction is possible. Laplace's demon cannot be built. Not even in principle.

First, there is no way to determine the current state of the wavefunction. Measurement is just entanglement, and that doesn't give us access to the data we need.
Then we cannot build a computer to store the wavefunction. A computer large enough to store the state of the universe would need to be larger than the universe. And it would be slower, requiring more than 1 hour to run 1 hour of simulation.
But even if you could solve those problems with compression and optimization, you'll end up having the computer inside the simulated universe, so the computer needs to simulate itself and then it gets stuck in an infinite recursion.

The smallest and fastest simulation of our universe is the universe itself. To find out what happens tomorrow, the fastest method is to wait 24 hours.

Note that I'm talking about processes and not the universe as a whole. This process could be any subset of the universe and, assuming the universe is infinite (and so an observer is guaranteed access to more resources than exist in the subset), it is, in principle, possible for an observer to predict the behaviour of that subset.

[Soupspoon]

You'd need to set up the simulation equipment outside of the observable universe for the fully simulated subset. If not, the flow of electrons in the simulator could conceivably adjust the electrical flux around the target, and around all the particles one assumes are fully quantised in the simulation as they enter into the target area, but not taking into account the simulation feedback. Or minute gravitational waves jiggle things.

All is not lost if there's a way to intercept and record the simulation's 'backwash' and add it back as a usable factor into the simulation before the simulated point in time of the simulation for which the simulated region, in real-time, will be influenced.

But that then supposes an even more efficiently Heisenberg-proof method than the entirely-out-of-light-cone example. (Which already is a problem, as you need a way of conveying the initial state of information from the target to the simulator without that information-moving process itself muddying the no-longer-initial state. I suggest an Event Horizon, but then the usefulness of the observing simulator, stuck the wrong side of an information barrier is questionable. Anyone got Stephen Hawking's ICQ UIN? )

[Tub]

Note that I'm talking about processes and not the universe as a whole. This process could be any subset of the universe and, assuming the universe is infinite (and so an observer is guaranteed access to more resources than exist in the subset), it is, in principle, possible for an observer to predict the behaviour of that subset.

I don't think so.

If you're talking about a classical process like predicting the equilibrium temperature of a glass of water with an ice cube in it, then that's not a deterministic process. You lack information about the initial system that can influence your result. Your prediction is likely to be very accurate, but it's not perfect. There is also no such thing as an isolated system. Your simplified formula would predict that your isolated system keeps the same temperature for the next 10 billion years, but when the sun goes supernova, I'd expect your measurements to dissent by several orders of magnitude.
Ergo: not deterministic, not predictable. Approximations don't count, just like newton's laws of motion wouldn't allow us to discount free will.

If the universe were infinite in size, and you wanted to simulare the wavefunction of one part inside of another part, then it still wouldn't work. You still cannot measure the state of the wavefunction. You can simulate a new universe in the lab, but you cannot simulate the universe we live in, because we cannot know its state.
But let's assume that you could violate the no-cloning-theorem and measure the wavefunction by polarizing your reflector dish with a tricorder. Let's also assume that you can just ignore a part of the wavefunction without affecting your results. You're still not going to predict anything. You need to measure and transmit the state to your remote computer at the speed of light. Then you need to calculate stuff. Then you need to transmit the results back into the original region (submit your lottery ticket or something). You don't need math to conclude: if the computer was outside of the predicted event's past light cone, then the results cannot arrive before the event has passed. If it was (partially) inside the past light cone, you're back to the original problem of recursion. The best you can hope for is the confirmation that your simulated event really did happen, but not a prediction.

[Eebster the Great]

It's worth mentioning that quantum randomness is probably not going to be a particularly significant effect in the macro-level interactions that happen in the brain. It's too big, there's too many particles, and it's too hot.

Unless you are Roger Penrose. His brain is in a constant quantum superposition of "brilliant mathematician and physicist" and "crazy quantum woo practitioner."

[ucim]

Err,
Wouldn't the standard response to that be that the soul is what we are referring to as "Fred" when we say that "Fred has/doesn't-have 'free will' " ?

Then we just examine the soul's mechanism. It's funny how susceptible the soul's decisionmaking process is to the external world (alcohol, drugs, electrostim, pain, optical illusion...) and yet how inaccessible any of its mechanics are. Examined closely there is a fatal contradiction.

Jose

[Eebster the Great]

I'm surprised nobody has brought up the term "compatibilism" yet. I think this is basically what dougly is getting at (and yes, the Sean Carroll article he linked is very good).

Essentially, if we view free will as an emergent (apparent) phenomenon rather than a fundamental property of the universe, then there is no contradiction between free will and determinism. This is in the same way that the apparent randomness of a coin flip is not inconsistent with the deterministic laws of fluid dynamics and ballistics governing its motion. Comparing free will to the Schrodinger equation is, in this view, a category error, akin to comparing a paragraph to the ink it is printed with. We "choose" in every socially meaningful sense, and perhaps in any meaningful sense at all. After all, people in this thread have repeatedly said that the notion of free will is incoherent, yet in practical terms we talk about choices all the time, so maybe that is the best definition.

Personally, I walk the middle road as a radical semi-compatibilist.

[eSOANEM]

↶

Note that I'm talking about processes and not the universe as a whole. This process could be any subset of the universe and, assuming the universe is infinite (and so an observer is guaranteed access to more resources than exist in the subset), it is, in principle, possible for an observer to predict the behaviour of that subset.

I don't think so.

If you're talking about a classical process like predicting the equilibrium temperature of a glass of water with an ice cube in it, then that's not a deterministic process. You lack information about the initial system that can influence your result. Your prediction is likely to be very accurate, but it's not perfect. There is also no such thing as an isolated system. Your simplified formula would predict that your isolated system keeps the same temperature for the next 10 billion years, but when the sun goes supernova, I'd expect your measurements to dissent by several orders of magnitude.
Ergo: not deterministic, not predictable. Approximations don't count, just like newton's laws of motion wouldn't allow us to discount free will.

My words were chosen carefully; "in principle" are some particularly important ones. Given arbitrarily precise initial data, I can produce arbitrarily precise and accurate predictions.

An ice cube in a glass of water is entirely predictable and deterministic in this sense.

[Frenetic Pony]

I'm surprised nobody has brought up the term "compatibilism" yet. I think this is basically what dougly is getting at (and yes, the Sean Carroll article he linked is very good).

Essentially, if we view free will as an emergent (apparent) phenomenon rather than a fundamental property of the universe, then there is no contradiction between free will and determinism. This is in the same way that the apparent randomness of a coin flip is not inconsistent with the deterministic laws of fluid dynamics and ballistics governing its motion. Comparing free will to the Schrodinger equation is, in this view, a category error, akin to comparing a paragraph to the ink it is printed with. We "choose" in every socially meaningful sense, and perhaps in any meaningful sense at all. After all, people in this thread have repeatedly said that the notion of free will is incoherent, yet in practical terms we talk about choices all the time, so maybe that is the best definition.

Personally, I walk the middle road as a radical semi-compatibilist.

This kind of gets at the more interesting concept of, uhhh, free will, that discounts free will as a "fundametal" concept and just replaces it with complexity.

The beginning of the universe (theoretically) is perfectly ordered, perfectly not complex, and thus perfectly predictable. Nothing interesting happens there. The (theoretical) end of the universe is perfectly disordered, and thus in its own way perfectly predictable as well. Again, nothing interesting happens there.

But right here, and now, when we live, interesting things happen. Right now, things are complex. The metaphor I'd thought of is handing someone banana and a rock, and asking which they'd like to eat. Fundamentally, after all, they're the same thing, a collection of protons, neutrons, and electrons bound together. Or you could go more fundamental and say they're a collection of electrons, gluons, up quarks, and down quarks!

Never the less, you'd choose to eat the banana. IE the universe builds interesting things out of base building blocks. Maybe "free will" could just be a huge spike in complexity. The more complex an object, the less predictable its state over a given time is. A bananas state will be slightly less predictable than the rock's state, a human's state will be much less predictable than a bananas state over any given time period. If that's what you want to call "free will", which seems as good a definition as any, then there you are, free will as a consequence of complexity, rather than a fundamental concept. Of course it still separates "objects" into, well, separate things. but hey it's a necessary reduction in computational complexity.

Ah! And as to Tubs "deterministic outcomes" I just realized something (and didn't totally go through it a lot so maybe I'm missing something). But multiple worlds is still fundamentally undpredictable, as you have no way of predicting which universe you end up in. So from the perspective of each individual universe your outcome, even if you had perfect information otherwise (which as stated earlier in this very thread is impossible anyway) would still be indeterminate, and thus from your perspective which universe you end up in is still random.

Edit- Ok, just rounding back on the same point earlier, which is that interpretations can make no predictions beyond the original mathematics. Meaning that no observer can possibly make deterministic predictions better than the defined probability distribution no matter the "interpretations". Which kinda makes interpretations themselves fairly metaphysical.

[madaco]

Err,
Wouldn't the standard response to that be that the soul is what we are referring to as "Fred" when we say that "Fred has/doesn't-have 'free will' " ?

Then we just examine the soul's mechanism. It's funny how susceptible the soul's decisionmaking process is to the external world (alcohol, drugs, electrostim, pain, optical illusion...) and yet how inaccessible any of its mechanics are. Examined closely there is a fatal contradiction.

Jose

In the context, you seemed to be arguing that "even if souls do/did exist, that doesn't/wouldn't allow people free will, because the soul would be controlling the person, not the person being a free-willed-thing itself".

I don't understand why you would make that argument if, when that argument is questioned, you would just go to the "well it doesn't matter if souls existing could allow for free will, because souls don't exist". Why not just make that argument in the first place instead?

Also, optical illusions? Really? Can you imagine a scenario where someone learning about some optical illusions is what causes them to have their probability for "souls exist" drop below 1/2 ? I don't know that I can say that I can't see how that could be any evidence at all, but I have a hard time imagining a reasonable situation where that acts as particularly relevant/significant evidence.

To be clear, I am not here arguing for the position "We have free will because we are free-willed-souls that have a causal influence on our bodies, which is how we take actions in the world". What I am saying is that an argument you gave is one that I don't find convincing.

[Copper Bezel]

In the context, you seemed to be arguing that "even if souls do/did exist, that doesn't/wouldn't allow people free will, because the soul would be controlling the person, not the person being a free-willed-thing itself".

I don't understand why you would make that argument if, when that argument is questioned, you would just go to the "well it doesn't matter if souls existing could allow for free will, because souls don't exist". Why not just make that argument in the first place instead?

Those are your options. Pick one. But I think you misunderstood the point. If the soul is the perceiving agent, then perception is evidence of how it functions, and can be used to study it. If there is a pathway to studying the soul, it is likely reducible to parts; in any case, we can still view it as a functioning mechanism and it doesn't add anything special in regards to free will. You can take this to its dualist extreme and create a magical system entirely separate from the body that performs every single function of the brain, and it's still going to be bound to the causal factors that are ultimately responsible for the behavior of the real brain.

Also, optical illusions? Really? Can you imagine a scenario where someone learning about some optical illusions is what causes them to have their probability for "souls exist" drop below 1/2 ? I don't know that I can say that I can't see how that could be any evidence at all, but I have a hard time imagining a reasonable situation where that acts as particularly relevant/significant evidence.

I don't see that it's in any way different from knowledge of the effects of brain injuries, just on a smaller scale. It indicates very clearly that any soul consistent with observation is a very mechanistic thing (though without the advantage of damage to particular parts of the brain corresponding to particular parts of the soul, etc.)

[ucim]

In the context, you seemed to be arguing that "even if souls do/did exist, that doesn't/wouldn't allow people free will, because the soul would be controlling the person, not the person being a free-willed-thing itself".

Correct. (At least for the version of souls that act as the "point source" of "just who is Fred?", which is kind of the point of souls to begin with.)

I don't understand why you would make that argument if, when that argument is questioned, you would just go to the "well it doesn't matter if souls existing could allow for free will, because souls don't exist". Why not just make that argument in the first place instead?

If the argument is questioned, I would press my argument, because it is sound. My point is that souls do not provide a mechanism for free will, so should not be considered in an argument about free will.

Spoiler:

The a priori argument that "souls don't exist" is not sound. True, there is no evidence for them, but their existence cannot be disproven any more than Russel's Teapot can. I happen to believe that souls do not exist, but their nonexistence does not negate free will. Free will is negated by other means.

Copper Bezel has it right. As in Mrs. Robinson: "Any way you look at it, you lose."

Similarly, arguments about randomness, quantum or otherwise, are equally irrelevant. Choice does not come about by randomness - in fact, randomness is quite the opposite of choice. Predictability is not relevant either (except as evidence that you could open the hood and learn something useful). You make a choice to stay in your lane and not veer into the oncoming pretty truck; it's still a choice even if predicting you'd do so would pretty much nail it.

At its base, free will is a concept whose usefulness comes about in how we think about and interact with things that have (or do not have) it. It's an interface question. Do you push the buttons on the outside, or do you open the hood and poke around with a needle? If you are pushing the buttons, you are letting it choose how to react. But if you open the hood, you are pulling its strings.

Because of this, free will (as a concept) is not a property of a thing, but of how we view a thing - whether we view it holistically or mechanically.

Also, optical illusions? Really? Can you imagine a scenario where someone learning about some optical illusions is what causes them to have their probability for "souls exist" drop below 1/2 ?

Nope. But that misses the point. Souls are irrelevant (and probably fictional). They are not part of the issue.

The point is that thinking about illusions separates input from processing. But inside the brain, processing provides input for more processing, leading to some output (action) occurring. Illusions can occur anywhere along the path. Chase them and you have to open the box. Do that and you've lost the point source.

Jose

[tomandlu]

"Unask the question."

Or ask a different one - "Is it right that I should feel responsible for the choices I make?" Assuming your answer is 'yes', then it would appear that we 'know' free-will exists without having to establish it empirically. We are aware that we have choices, we are aware that those choices have consequences, and we make our choices based on that knowledge.

Given that it is generally agreed that consciousness is a requirement for free-will, then it seems to me that it is more usefully defined as mind-dependent, just as beauty is. We can debate whether beautiful things have some physical property or attribute that makes them beautiful, but we don't need to debate whether beauty exists, per se. By definition, beauty exists if we perceive it to exist, and any other discussion is just arguing over the small-print - Cogito ergo sum.

[Copper Bezel]

Except that I think we can definitively say that neither beauty nor free will can be reduced to a particular physical property. They're both subjectively experienced, emergent things made up of many complex parts that are not themselves. There's no one part that's irreducibly the thing we're talking about.

The rest of what you said is definitely correct, though.

[tomandlu]

Except that I think we can definitively say that neither beauty nor free will can be reduced to a particular physical property. They're both subjectively experienced, emergent things made up of many complex parts that are not themselves. There's no one part that's irreducibly the thing we're talking about.

The rest of what you said is definitely correct, though.

I would agree with you. My son, who's studying philosophy, says there is still debate (cognitivism vs non-cognitivism iirc). On the other hand, he also says that philosophers are now wondering if they can solve some problems by invoking quantum mechanics, so screw those guys.

[Copper Bezel]

"Screw those guys" is definitely my general feeling there. My philosophy minor in undergrad taught me that philosophy is the study of malformed questions. Figure out what you're actually doing, whether that be art theory, history, literary criticism, semiotics, sociology, or theoretical physics, and then contribute to that actual field. With, like, maybe some data when you run into an answerable question for which you do not have an answer.

[tomandlu]

"Screw those guys" is definitely my general feeling there. My philosophy minor in undergrad taught me that philosophy is the study of malformed questions. Figure out what you're actually doing, whether that be art theory, history, literary criticism, semiotics, sociology, or theoretical physics, and then contribute to that actual field. With, like, maybe some data when you run into an answerable question for which you do not have an answer.

I admire its rigour (which is a bit like telling an author that you admire his spelling).

[Copper Bezel]

Yeah, fair point on both counts. Hard work goes in and internally consistent systems come out, but the same could be said for theology.

[ucim]

"Is it right that I should feel responsible for the choices I make?"

Well, first off you are assuming the consequent: the existance of "choices" to be made, and an "I" to make them (and the presumption that the "I" is the same in both instances of the word in the question; not a trivial point). The question is still malformed, and that malformation may (or may not) come into play depending on how closely you need to examine things to get the answer you seek. The more obvious issue however is the idea of "right", which is ambiguous and begs the reader to misinterpret.

Morally right (as opposed to unjust? and in relative or absolute sense?) Logically right (as opposed to incorrect?) Justifiable (as opposed to unsupportable?). The question is more one of the hearer's mindset than of any kind of objective evaluation of an external circumstance.

Assuming your answer is 'yes', then it would appear that we 'know' free-will exists without having to establish it empirically.

What appears to be so, isn't always the case. We don't "know" anything of the sort, we merely unconsciously "believe" it to be true; this is politely called "beef by-products of the highest purity". More on point, not rejecting the question implies that we have some use of the concept of free will; this doesn't show it exists, just that the idea is useful if not closely examined. And I will agree with that. It is useful if not closely examined; this can easily lead to bigoted abuse (of those whose mental fortitude is sufficiently impaired or inferior to our own). That may be one of the reasons to examine the question in the first place: to justify punishment and feelings of superiority.

By definition, beauty exists if we perceive it to exist

(quibble: that's not "by definition") not-so-quibble: Beauty is not a property of the beautiful object; it is a property of the interaction between it and the mind perceiving it so. Similarly [never mind: mu].

[P]hilosophy is the study of malformed questions.

Well put.

Jose

[tomandlu]

Morally right (as opposed to unjust? and in relative or absolute sense?) Logically right (as opposed to incorrect?) Justifiable (as opposed to unsupportable?). The question is more one of the hearer's mindset than of any kind of objective evaluation of an external circumstance.

"Who's 'im, Bill?"
"A philosopher"
"'Eave 'alf a brick at 'im."

You're quite right of course, but also the reason I always end up wanting to punch philosophers*

In the end, my only real point is that asking science whether free will exists doesn't make much more sense than asking it whether beauty exists.

P.S.

[P]hilosophy is the study of malformed questions.

Well put.

I don't think that was meant as a compliment (although it's nice to have it pointed out that it could be)

* although Wittgenstein gets off with a light slapping

[Copper Bezel]

I think there's a lot of value in making fine distinctions and unpacking intended and unintended meanings. It's possible I simply had a terrible philosophy professor. We spent a lot of time with Kant.

[tomandlu]

I think there's a lot of value in making fine distinctions and unpacking intended and unintended meanings. It's possible I simply had a terrible philosophy professor. We spent a lot of time with Kant.

From my conversations with my son, philosophy raises lots of interesting questions and then answers absolutely none of them, but we've still got the questions, so that's fine. If nothing else, it should teach us to be humble in our assumptions.

[doogly]

Is it fine though? It's not necessarily the case that these questions are valuable just because they can be posed.

[gmalivuk]

I had a professor who described philosophy as (in part) a holding area for questions science couldn't answer yet, where people could formulate interesting questions and speculate about answers and their implications.

The problem is when philosophers want to keep holding onto those questions even after we know how to start addressing them scientifically.

(And the problem on the other side is when people think science is equipped to answer questions that were never in that holding area to begin with, such as core questions about ethics and the like.)

[tomandlu]

I've been musing on this and various related discussions, at least as they related to consciousness, and I thought I'd tack them on here (although it's a bit OT in some ways).

One of the problems of consciousness is that, if we accept that it arises from a purely physical process, then that implies that consciousness must be a mappable state - information that's transferable. However, that seems wrong - is it really conceivable that one could output consciousness like a computer listing and transfer it? This, afaict, is where some philosophers start adding 'quantum' to the mix, in the hope that it can resolve the apparent paradox. This seems very wrong, but it did get me wondering if chaos-theory works as a better model.

We cannot, after all, transfer a complete Mandelbrot from one PC to another - all we can do is transfer the algorithm. This, imho, neatly solves the paradox - consciousness can be created but not transferred. This does, potentially, raise a sticky problem for true AI. We might discover the necessary algorithm, but we might have very little control over the consciousness that it generates.

I'm not convinced this is an original idea - consciousness looks like a classic 3+ body problem, with (at least) memory, sensation and response all influencing each other - but I'd be interested to know whether it should be shot down in flames or not.

[doogly]

Why wouldn't that be conceivable?

[tomandlu]

Why wouldn't that be conceivable?

I find it conceivable that one might be able to transfer the sensation of pain, but not the person's reaction, their conscious response. By all means disagree; I'm not sure I agree with myself either...

[gmalivuk]

We cannot, after all, transfer a complete Mandelbrot from one PC to another - all we can do is transfer the algorithm. This, imho, neatly solves the paradox - consciousness can be created but not transferred.

We can't transfer the complete Mandelbrot set because it's infinite.

Are you suggesting that consciousness is also infinite?

If not, why wouldn't it be transferrable?

[doogly]

And if it's infinite, it can't be stored on one PC either, no?

[gmalivuk]

Also a good point.

[tomandlu]

Mandelbrot was probably a bad choice - I'm more thinking along the lines of "extremely sensitive to initial conditions".

e.g. we can model weather systems, but we can't model our weather system, and there is no inherent contradiction in that.

[gmalivuk]

Sure, it may not be possible to reliably transfer a particular consciousness from one thing (brain, computer, whatever) to another. But that says nothing about whether consciousness is purely physical or whether it could happen on a computer, and thus doesn't really say anything about free will.

[ucim]

Consciousness is not a thing, it is a perspective on a thing. Consciousness is a first-person outlook, and as such is only applicable to the (grammatical) first person.

Jose

[tomandlu]

Sure, it may not be possible to reliably transfer a particular consciousness from one thing (brain, computer, whatever) to another. But that says nothing about whether consciousness is purely physical or whether it could happen on a computer, and thus doesn't really say anything about free will.

I don't think there's anything there that I disagree with (and, yes, I've wandered way OT regarding FW).

I guess what I'm trying to come to terms with is the idea that consciousness could arise in some extended Chinese room scenario, and that to me seems counter-intuitive. At the same time, any other view seems to insinuate some meta-physical quality to consciousness, which I also find unreasonable.

So, either one of those unlikely things is true, or there is some other, more reconcilable, alternative.

On balance, I would favour that, yes, you can create consciousness in a chinese room. It would be really fucking scary, though.

[eSOANEM]

↶
Consciousness is not a thing, it is a perspective on a thing. Consciousness is a first-person outlook, and as such is only applicable to the (grammatical) first person.

Idk, that seems like half a step too far to me. Consciousness is the perspective that it is meaningful for that thing to be first person, not that it is the first person. As such, it's pretty reasonable to say other people who can speak are conscious and then it seems reasonable to extend to all humans but anything beyond that gets blurry.

[ucim]

Consciousness is the perspective that it is meaningful for that thing to be first person, not that it is the first person.

Yes, that is what I said. "Consciousness is a first-person outlook": Consciousness is an outlook, not a (grammatical) "first person". It is an outlook of the "first person" type.

...it's pretty reasonable to say other people entities who can speak are conscious...

FTFY: limiting it to "people" begs the question. However, my car "speaks" to me; is it conscious by dint of this ability? I'd say not.

If you accept consciousness as a "first person perspective on onesself", then it requires:

1: that the entity have a perspective -- a mental model of reality, and

2: that it be able to apply that model to itself.

It's not a binary question but one of degree.

Jose