Are You Conscious?

[Frank Newgent]

Originally Posted by rocketdodger

Many posters here -- who have no education and no capability to understand and no desire to learn -- start to get uncomfortable when real science is discussed.

The problem may lie with your educational tools.

It was just after your Velcro-dot hand puppet Recursion told little orange-haired felt and Styrofoam Hilbert he was considering a simulation of nondeterministic Turing machines as a way to be more popular with the servos that the barrage of slide rules and punch tape began.

[rocketdodger]

Originally Posted by westprog

I think that you'll find that one of the implications of the Special and General theories of relativity is that it's not possible to talk about the state of the universe at any given instant. There is no universal "now" when the state can be recorded.

Wrong.

Now is now, regardless of what reference frame one is in.

[Ichneumonwasp]

Originally Posted by westprog

That the Turing machine is an abstraction does not mean that it doesn't have to explain what goes on. Gravitational theory is an abstraction too. When Newtonian gravity turned out to be inaccurate, it was replaced with RD's beloved General Theory of Relativity - because it didn't model what was going on.

Well, yes, of course. But I don't see how they don't model or explain what is going on. That is their purpose.

What they don't do is "act consciously". They are a model.

Quote:

If Pixy and RD were to accept that a physical implementation of a Turing machine might not have the requisite elements necessary for consciousness - and a "might" is more than they are prepared to give at the moment - then we'd be able to move on.

You don't seem to be so tightly constrained by Turing fundamentalism, so this argument might not be as interminable as it seems to be getting.

Why would a physical implementation of a Turing machine not have the necessary elements for consciousness unless we leave something out? We have brains in front of us as a guide for what seems to be necessary. If we can compute the process abstractly, then we should be able to do so in another medium. We simply can't leave to the side any of the important bits. But that is the same issue with all physical implementations of abstract ideas.

[rocketdodger]

Originally Posted by westprog

And still you have not come up with an example of how a Turing machine can interact with the environment in the way that a human brain does. There's nothing missing except everything.

Except that we told you it would be in the same way the human brain interacts with the environment.

What is so difficult about that concept?

[Belz...]

Originally Posted by westprog

If we want to understand consciousness, then we should look at what it does, and try to find a model that might do the same things.That seems fairly logical.

And aside from behaviour we can observe, what does it do, exactly ?

Quote:

A Turing machine deals with things in order, true - and it doesn't know what is going to arrive. But it instigates each step as it occurs. The TM reads the tape. It doesn't react to the tape moving itself.

I still don't know where you're going with this. Turing machines don't exist.

[Belz...]

Originally Posted by AkuManiMani

"Is science conducted by delusional people reliable?"

A yes or no answer will suffice, but do feel free to explain your answer.

It isn't a simple yes or no answer, Aku. This is the third time that I tell you this.

Quote:

So I guess we should throw out fields like psychology and cognitive neuroscience since they, to a very large degree, depend upon the human capacity to introspect

Ugh. Medecine also often relies on introspection, Aku. But the point is it relies on a large amount of it from various sources to eliminate bias as much as possible. With Psychology it's a lot harder, hence why it's not quite as 'scientific' as, say, chemistry.

Quote:

Once again, you're missing the point entirely. The point I'm making is that sensations themselves are unambiguous objects, distinct from any particular stimulus. If you're shot and, instead of feeling pain, you experience the sensation of sweetness you would immediately know the difference. Likewise, if the sensation were replaced with the sensation of cold, or redness, or what have you -- each of those subjective responses are unambiguous and unmistakable IAOT. Are you following?

Yes, but I disagree. How are you so sure that you can label a sensation appropriately, if it's possible to feel one instead of the other ? I think you place far too much faith in your ability to introspect.

Quote:

First of all, as I've already pointed out to you numerous times [in this discussion alone] that even material objects are in some sense verbs.

I KNOW THIS. I agreed to this already more than once. But saying that nouns are also verbs is only useful to a degree, because at some point you're going to have to agree that you must distinguish nouns from verbs. So, once again: how do you know that consciouness is a "thing" i.e. that it's actually composed of something (particles, I assume, themselves behaviours as we've already ageed) and not just what one of those something does ?

Quote:

Second of all, experience is what consciousness does -- not vis versa. One could still be conscious without any sensory stimulation producing experiences but, absent consciousness, experiences are not possible. Subjective experience depends upon there being a conscious subject.

In fact, experience is so intimately linked to consciouness you'd swear they're one and the same, or at least that consciouness is the sum total of experiences, which would be contrary to your claim that experiences are what consciousness does. I'm not asking you to tell me what you believe, I'm asking you to explain WHY you believe it. And please skip the philosophy. I'm more interested in the evidence.

Quote:

Huh? How the hell do you think that follows from solipsism?

I explained how in the quote you were responding to. Please read my "damn" posts, as you say.

Quote:

I'm actually dead serious. You're probably one of the most unreflective people I've ever conversed with.

That's odd, as I spend most of my free time thinking. Perhaps this is your perception because I disagree with you.

Quote:

I can see why you reject introspection as unreliable; your own skills of introspection are atrocious.

Introspection is unreliable. You shouldn't put so much faith in it.

Quote:

That makes about as much sense as claiming an abacus forms a conclusion because you stop sliding its beads.

No, because it can't compute by itself. Throwing rocks on the ground to count is the same thing: the rocks don't compute.

Quote:

It's the conscious user of the calculator that forms the conclusion.

A conclusion is what follows from premises. I don't understand why you think one needs to be conscious in order to reach it.

Quote:

The calculator itself only produces symbolic results as specified by the functional constraints designed into it.

So does the brain. The brain is just a biological computer, really. The only reason I can fathom as to why you keep saying that consciousness is required is because you think consciousness is something a bit "magical" i.e. different in nature and function from anything else in existence.

Quote:

If solipsism were true then it wouldn't be wrong to treat other people like crap. After all, they would just be characters in the reality my mind has dreamed up

Wrong. If solipsism were true there would be NOTHING to gain, whatsoever, from doing anything about or to these non-existent people. Hence apathy is the only resulting behaviour from that philosophy.

[westprog]

Originally Posted by rocketdodger

Except that we told you it would be in the same way the human brain interacts with the environment.

What is so difficult about that concept?

Yes, you told me that, but that doesn't make it so. There is no way that a Turing machine can respond in real time to the environment. Nobody has even attempted to explain how it might do so, beyond bare assertions and nonsense about general relativity.

What is so difficult in admitting that the Turing machine model is not sufficient to explain how human consciousness interacts with the environment, and that the model needs amending?

[westprog]

Originally Posted by rocketdodger

Wrong.

Now is now, regardless of what reference frame one is in.

If you are going to bring up relativity, you should try to find out a little bit about it. "Now" is local. There is no "now" for the entire universe.

[westprog]

Originally Posted by Belz...

And aside from behaviour we can observe, what does it do, exactly ?

Consciousness reacts in real time to events.

Quote:

I still don't know where you're going with this. Turing machines don't exist.

A Turing machine is a scientific model. The Strong AI contention is that some form of Turing machine implementation is sufficient to achieve consciousness. IOW, if a real life machine has the properties of a Turing Machine, then it can achieve consciousness.

Whether or not you chose to call a real life machine that implements the behaviour of a Turing machine a Turing machine or not is really just a matter of nomenclature. Since the essence of the Strong AI position is that a real life implementation of a Turing machine is doing real computation, denying that Turing machines exist doesn't seem to be particularly helpful to that cause.

If we can observe something about consciousness - the response to events in real time - which is not explained by the Turing machine model - then we have to consider whether that model is, in principle, sufficient. Pixy, to give on example, has claimed that not only is it probable that the Turing model is sufficient, but that it is provably certain that it is sufficient. By demonstrating that human consciousness in its full functionality cannot, after all, be explained by the Turing model, at the very least the certainty of Pixy's position has been somewhat undermined.

[Belz...]

Originally Posted by westprog

Consciousness reacts in real time to events.

Define "real time".

[westprog]

Originally Posted by Belz...

Define "real time".

When something has to happen within a certain specified interval, it's real time. So for someone to catch a ball, the hand has to be directed to move into the path of the ball before it's too late.

We know that consciousness has to keep in sync with the world. Therefore it's functionally necessary for the mind to have some kind of time dependence.

[rocketdodger]

Originally Posted by westprog

If you are going to bring up relativity, you should try to find out a little bit about it. "Now" is local. There is no "now" for the entire universe.

Um, that doesn't make sense.

If I view the rest of the universe from my local reference frame, then this "now" does indeed apply to the rest of the universe.

[Belz...]

Originally Posted by westprog

When something has to happen within a certain specified interval, it's real time. So for someone to catch a ball, the hand has to be directed to move into the path of the ball before it's too late.

Ok, gotcha. That's not what I thought you meant.

Quote:

We know that consciousness has to keep in sync with the world. Therefore it's functionally necessary for the mind to have some kind of time dependence.

Depends what you mean by that. When sleeping, I'm still "conscious" in the sense that I still think about stuff, and yet I'm not much "aware" of the outside world... in fact I could be completely unaware of the outside world, in which case I could be "out of sync" with it without any adverse effect. No ?

[rocketdodger]

Originally Posted by westprog

Yes, you told me that, but that doesn't make it so. There is no way that a Turing machine can respond in real time to the environment. Nobody has even attempted to explain how it might do so, beyond bare assertions and nonsense about general relativity.

What is so difficult in admitting that the Turing machine model is not sufficient to explain how human consciousness interacts with the environment, and that the model needs amending?

What are you talking about?

I can take every event in your life that you have reacted to -- down to the particle interactions on your retinas -- and list them, in an ordered fashion, interleaved with the algorithm to operate on them, on the TM input tape. Then I can take all of the events that you will react to and list them as well.

I am not sure you think otherwise, but you can do whatever you want with the TM input tape.

[westprog]

Originally Posted by Belz...

Ok, gotcha. That's not what I thought you meant.



Depends what you mean by that. When sleeping, I'm still "conscious" in the sense that I still think about stuff, and yet I'm not much "aware" of the outside world... in fact I could be completely unaware of the outside world, in which case I could be "out of sync" with it without any adverse effect. No ?

You still need to breath, and your heart needs to beat - which is part of brain/nervous system function. Is that inextricably linked with consciousness? I don't know - but conscious actions are certainly time dependent.

Whether or not you are conscious when asleep is an interesting point, but you are certainly less conscious, and your brain is less time aware. I don't think it points especially strongly in either direction, but if anything, it demonstrates the time dependence of the conscious mind.

[westprog]

Originally Posted by rocketdodger

Um, that doesn't make sense.

If I view the rest of the universe from my local reference frame, then this "now" does indeed apply to the rest of the universe.

It's a different "now" to what someone a million light years away sees.

[rocketdodger]

Originally Posted by westprog

It's a different "now" to what someone a million light years away sees.

Lol.

You mean like how at the instant I am typing this -- "now" -- the time in New York city is 7:26 instead of 6:26?

Wow, that is enlightening westprog. I hadn't thought of that gem. Everything I have said is rendered invalid by that utterly remarkable piece of logic you just introduced me to.

[Ichneumonwasp]

Originally Posted by westprog

Consciousness reacts in real time to events.




A Turing machine is a scientific model. The Strong AI contention is that some form of Turing machine implementation is sufficient to achieve consciousness. IOW, if a real life machine has the properties of a Turing Machine, then it can achieve consciousness.

Whether or not you chose to call a real life machine that implements the behaviour of a Turing machine a Turing machine or not is really just a matter of nomenclature. Since the essence of the Strong AI position is that a real life implementation of a Turing machine is doing real computation, denying that Turing machines exist doesn't seem to be particularly helpful to that cause.

If we can observe something about consciousness - the response to events in real time - which is not explained by the Turing machine model - then we have to consider whether that model is, in principle, sufficient. Pixy, to give on example, has claimed that not only is it probable that the Turing model is sufficient, but that it is provably certain that it is sufficient. By demonstrating that human consciousness in its full functionality cannot, after all, be explained by the Turing model, at the very least the certainty of Pixy's position has been somewhat undermined.

But you haven't done that. That the Turing machine operates abstractly, independent of time, does not mean that it cannot represent time dependence in its functions. Since it is just a model it is not conscious, so time dependence doesn't matter there. It is only in the implementation of the model that time dependence becomes an issue -- in other words, in the real world.

Unless you can provide some reason as to why a Turing machine cannot represent time dependence in its model of consciousness you have not undermined Pixy's argument.

[westprog]

Originally Posted by rocketdodger

Lol.

You mean like how at the instant I am typing this -- "now" -- the time in New York city is 7:26 instead of 6:26?

No, not like that at all.

[westprog]

Originally Posted by Ichneumonwasp

But you haven't done that. That the Turing machine operates abstractly, independent of time, does not mean that it cannot represent time dependence in its functions. Since it is just a model it is not conscious, so time dependence doesn't matter there. It is only in the implementation of the model that time dependence becomes an issue -- in other words, in the real world.

If when you implement the Turing machine model, you find you need some other quantities not present in the model, you need a new model. So you define a model where time dependence is significant.

Choosing a better model is fairly normal practice in science, when one model doesn't represent what is going on.

Quote:

Unless you can provide some reason as to why a Turing machine cannot represent time dependence in its model of consciousness you have not undermined Pixy's argument.

I know that a TM can model time dependence - but it cannot be time dependent. That is why a TM can simulate what the brain does, but it cannot emulate what the brain does.

[westprog]

Originally Posted by rocketdodger

What are you talking about?

I can take every event in your life that you have reacted to -- down to the particle interactions on your retinas -- and list them, in an ordered fashion, interleaved with the algorithm to operate on them, on the TM input tape. Then I can take all of the events that you will react to and list them as well.

I am not sure you think otherwise, but you can do whatever you want with the TM input tape.

Except catch a ball.

[annnnoid]

I-wasp made this suggestion:
"If we can move past these trivial issues and get to the meat of the matter -- what is 'feeling', 'awareness', 'attention', 'meaning', then we can move forward. Otherwise, we can just keep going round the bend and argue the same points using different words over and over."

Just an interesting observation on where I-wasp is pointing (subjectivity)….don’t know if anyone else has noticed:

What is the only place in the scientific lexicon where the rather unscientific language of subjectivity occurs? Observations that depend on the frame of reference of the observer:

QED.

There is obviously no empirical connection between QED and human subjectivity….except for the simple fact that subjectivity is a profoundly unique phenomenon (and so is QED). Subjectivity really can be described as measurements that depend entirely upon the frame of reference of the observer. It is obviously in no way conclusive that consciousness has some QED component (this has been exhaustively and inconclusively debated on these threads)…..but the existence or the two singularly unique phenomenon in such proximity to each other can hardly be considered coincidental. In a rather eerie way, subjectivity implies some QED relationship….which has rather unpredictable implications.

[westprog]

Originally Posted by annnnoid

I-wasp made this suggestion:
"If we can move past these trivial issues and get to the meat of the matter -- what is 'feeling', 'awareness', 'attention', 'meaning', then we can move forward. Otherwise, we can just keep going round the bend and argue the same points using different words over and over."

Just an interesting observation on where I-wasp is pointing (subjectivity)….don’t know if anyone else has noticed:

What is the only place in the scientific lexicon where the rather unscientific language of subjectivity occurs? Observations that depend on the frame of reference of the observer:

QED.

There is obviously no empirical connection between QED and human subjectivity….except for the simple fact that subjectivity is a profoundly unique phenomenon (and so is QED). Subjectivity really can be described as measurements that depend entirely upon the frame of reference of the observer. It is obviously in no way conclusive that consciousness has some QED component (this has been exhaustively and inconclusively debated on these threads)…..but the existence or the two singularly unique phenomenon in such proximity to each other can hardly be considered coincidental. In a rather eerie way, subjectivity implies some QED relationship….which has rather unpredictable implications.

I don't think that the role of the observer in QED can really be said to depend on consciousness. Nothing in Quantum theory requires a conscious observer. All that is required is some kind of physical interaction.

[annnnoid]

That was not quite what I was getting at Westprog. It is sort of a conceptual thing…that may imply something more than a conceptual thing. That is the point. QED involves self-referential reality, irrespective of whether or not there exists any conscious observer (as you say…though how a conscious observer may influence the issue is another question). This thread is attempting to explore another self-referential, ie: subjective phenomenon: Human subjective experience. Where else in the entire universe does there exist a phenomenon like human subjective experience? Nowhere, as far as we know. What variety of coincidence could it be that the physical phenomenon of subjective experience exists contemporaneously, in some variety of relationship, with QED (because it does, quite obviously)….the only other physical phenomenon with similar characteristics.

Yes….I know it’s wildly speculative (especially given the current level of complexity involved in establishing any QED component to brain architecture, let alone consciousness)….but subjectivity IS a real phenomenon….and it is a singularly unique phenomenon. Looking for a conventional mechanism simply related to brain architecture just seems far too….prosaic.

[Ichneumonwasp]

Originally Posted by westprog

If when you implement the Turing machine model, you find you need some other quantities not present in the model, you need a new model. So you define a model where time dependence is significant.

Choosing a better model is fairly normal practice in science, when one model doesn't represent what is going on.



I know that a TM can model time dependence - but it cannot be time dependent. That is why a TM can simulate what the brain does, but it cannot emulate what the brain does.

Again, that doesn't matter. The Turing machine is not conscious. It is a model. If it can model time dependence, and we have already identified time dependence as important to the process, where is the problem of implementation of the model -- which contains time dependence as one of its important features (even though the Turing machine itself is not time dependent, which doesn't matter because it is a model and not the thing itself)?

No one is saying that a Turing machine, as an abstraction, is conscious. It is the implementation of the model in the real world that can be conscious. So, if that model contains time dependence as part of the model, where is the problem?

[fuelair]

right now, definitely!!

[westprog]

Originally Posted by Ichneumonwasp

Again, that doesn't matter. The Turing machine is not conscious. It is a model. If it can model time dependence, and we have already identified time dependence as important to the process, where is the problem of implementation of the model -- which contains time dependence as one of its important features (even though the Turing machine itself is not time dependent, which doesn't matter because it is a model and not the thing itself)?

No one is saying that a Turing machine, as an abstraction, is conscious. It is the implementation of the model in the real world that can be conscious. So, if that model contains time dependence as part of the model, where is the problem?

This is why I gave a potted history of computing - because this can get a bit confusing.

The claim is that an instantiation of a Turing machine can experience consciousness - and that no additional features are required beyond what is defined in the Turing model. So what the Turing model does when instantiated is the issue.

So does a Turing machine, when instantiated, have time dependence? Well, yes and no. It has to have some kind of way of controlling its operations - but this time dependency has no relationship with the outside world. That's why machines built according to this model (that's hardware and software), cannot be used for control or monitoring purposes.

Similar machines which were designed to a different model were able to be used for control and monitoring purposes. It's important to realise that a different design principle was involved.

Machines built according to the Turing machine principle could model real time activities. However, they could not, in theory or in fact, except by happy accident, perform any kind of control or monitoring functions. The two classes of machine were quite distinct.

[Ichneumonwasp]

Originally Posted by westprog

This is why I gave a potted history of computing - because this can get a bit confusing.

The claim is that an instantiation of a Turing machine can experience consciousness - and that no additional features are required beyond what is defined in the Turing model. So what the Turing model does when instantiated is the issue.

So does a Turing machine, when instantiated, have time dependence? Well, yes and no. It has to have some kind of way of controlling its operations - but this time dependency has no relationship with the outside world. That's why machines built according to this model (that's hardware and software), cannot be used for control or monitoring purposes.

Similar machines which were designed to a different model were able to be used for control and monitoring purposes. It's important to realise that a different design principle was involved.

Machines built according to the Turing machine principle could model real time activities. However, they could not, in theory or in fact, except by happy accident, perform any kind of control or monitoring functions. The two classes of machine were quite distinct.

But none of that matters. The reason for bringing this up in the first place was to argue that if we could model this abstractly by using a Turing machine, then we should theoretically be able to recreate consciousness in another medium. That is all.

The Turing machine model obviously has to include all the salient features that are important to consciousness. If not, what is the point of the model? That a Turing machine works independent of time simply does not have anything to do with the model, the instantiation of the model, or anything else.

The history of computing tells us very little about what is important theoretically in this process. Or are you suggesting that it is impossible to include monitoring functions within a Turing machine model?

[rocketdodger]

Originally Posted by westprog

Except catch a ball.

Yeah well your brain can't catch a ball either, genius -- it is the body that does the catching.

So ... what was your point, again?

[rocketdodger]

Originally Posted by westprog

No, not like that at all.

Oh, well, then you should explain what you are talking about, because nobody has a clue.

If I am on Earth, and you are on a starship traveling near lightspeed, and time dilation causes your timescale to slow relative to mine by a factor of 0.01, then every interval of one second that passes for you corresponds to an inverval of 100 seconds that pass for me.

Agree?

And if, 100 seconds after you leave, I say "now," the current "wall clock" time in my frame will be 99 seconds ahead of the "wall clock" time in your frame.

Agree?

Now, if you agree to all of the above, then what on Earth is your problem with "now?" My "now" is simply 99 seconds ahead of your "now." The values are different for each "now", but who cares? Did anyone ever say the values had to the same?

Now is now. Learn math and physics, westprog.

[westprog]

Originally Posted by rocketdodger

Oh, well, then you should explain what you are talking about, because nobody has a clue.

If I am on Earth, and you are on a starship traveling near lightspeed, and time dilation causes your timescale to slow relative to mine by a factor of 0.01, then every interval of one second that passes for you corresponds to an inverval of 100 seconds that pass for me.

Agree?

And if, 100 seconds after you leave, I say "now," the current "wall clock" time in my frame will be 99 seconds ahead of the "wall clock" time in your frame.

Agree?

Now, if you agree to all of the above, then what on Earth is your problem with "now?" My "now" is simply 99 seconds ahead of your "now." The values are different for each "now", but who cares? Did anyone ever say the values had to the same?

Now is now. Learn math and physics, westprog.

If you are on Earth, and I am on a planet a million light years away, then now for me is a million years in your past. And now for you is a million years in my past.

[westprog]

Originally Posted by rocketdodger

Yeah well your brain can't catch a ball either, genius -- it is the body that does the catching.

So ... what was your point, again?

Very few bodies without brains can manage to catch anything. It's the brain interacting with the body, in real time, that allows the system to catch the ball.

[westprog]

Originally Posted by Ichneumonwasp

The history of computing tells us very little about what is important theoretically in this process. Or are you suggesting that it is impossible to include monitoring functions within a Turing machine model?

I'm saying that modelling monitoring functions is a necessary amendation to the Turing model.

[Ichneumonwasp]

Originally Posted by westprog

I'm saying that modelling monitoring functions is a necessary amendation to the Turing model.

An amendment to how they were used in the past or to how they can work?

Can a Turing machine model monitoring functions theoretically? If not, why not?

[Dancing David]

Originally Posted by annnnoid

I-wasp made this suggestion:
"If we can move past these trivial issues and get to the meat of the matter -- what is 'feeling', 'awareness', 'attention', 'meaning', then we can move forward. Otherwise, we can just keep going round the bend and argue the same points using different words over and over."

Just an interesting observation on where I-wasp is pointing (subjectivity)….don’t know if anyone else has noticed:

What is the only place in the scientific lexicon where the rather unscientific language of subjectivity occurs? Observations that depend on the frame of reference of the observer:

QED.

There is obviously no empirical connection between QED and human subjectivity….except for the simple fact that subjectivity is a profoundly unique phenomenon (and so is QED). Subjectivity really can be described as measurements that depend entirely upon the frame of reference of the observer. It is obviously in no way conclusive that consciousness has some QED component (this has been exhaustively and inconclusively debated on these threads)…..but the existence or the two singularly unique phenomenon in such proximity to each other can hardly be considered coincidental. In a rather eerie way, subjectivity implies some QED relationship….which has rather unpredictable implications.

:raises hand meekly:

Sensation, perception and verbal cognition, sir?

:lowers hand:

[Dancing David]

Originally Posted by westprog

I don't think that the role of the observer in QED can really be said to depend on consciousness. Nothing in Quantum theory requires a conscious observer. All that is required is some kind of physical interaction.

Thank you!

[Dancing David]

Originally Posted by rocketdodger

Yeah well your brain can't catch a ball either, genius -- it is the body that does the catching.

So ... what was your point, again?

How does Steven Hawkings catch a ball?

[Dancing David]

Originally Posted by westprog

Very few bodies without brains can manage to catch anything. It's the brain interacting with the body, in real time, that allows the system to catch the ball.

So is Steven Hawkings conscious?

[AlBell]

Originally Posted by Dancing David

How does Steven Hawkings catch a ball?

Yes, it's possible that damaged humans exist.

I wonder if westprog's basic thrust is to point out the problem with Pixy's formal definition of consciousness? Why isn't the instantiation of a Turing machine using pebbles in the sand conscious?

[Ichneumonwasp]

Originally Posted by AlBell

Yes, it's possible that damaged humans exist.

I wonder if westprog's basic thrust is to point out the problem with Pixy's formal definition of consciousness? Why isn't the instantiation of a Turing machine using pebbles in the sand conscious?

I don't know the example, so I'm making assumptions here, but rocks in the sand are not causal. Whatever solution we arrive at for consciousness, it must have causal properties -- neuron firings cause consciousness in their action.

Rocks in the sand can be seen as providing computation in an observer-dependent manner; in other words, we can define that sort of thing as computation. But that won't work for consciousness. The computations necessary for consciousness must occur through action, in an observer-independent fashion. That's what neurons do.