Has consciousness been fully explained?

[AlBell]

Yup. A computer is just another inanimate object.

 

[Piggy]

rocketdodger, I see you've been on this thread since my last reply to you. Would you mind responding to my post about what is happening when I wake up?

Can you explain what's going on in terms of information processing, if that's not just shorthand for a physical process (as "fuel processing" would be for what's happening in my truck's engine)?

I'm quite serious about that. Can we just focus on that example for a bit?

What is your explanation for the phenomenon of waking up and becoming conscious, and how does your explanation allow for the existence of conscious computer programs?

Thanks.

ETA: QuixoteCoyote, you can chime in, too, since I asked you essentially the same question.

 

[Frank Newgent]

Assigning a higher value to a statement which has been verified because it's been verified (as corresponding with reality) would be a normative choice... ie that it ought to be true.

There is nothing that ought to be true (or false) about liking the poblano peppers stuffed with goat cheese, bacon, wild rice, fresh basil, and garlic cloves that I made my wife and I for dinner a little while ago. That is a personal preference.

I stress the normative aspect as descriptive forms of theorizing depend on making normative choices; for example finding value in what is capable of being verified (or falsified) by observation/experiment (definition of empirical) is an act expressive of belief in a norm.

I hope that makes sense.

It doesn't really make sense -- can you just give me an example?

The establishment of a goal of to develop, for example, a simple method of chatting with artificial intelligence.

Establishing such a goal would be a normative act expressive of the belief that human intelligence can be so precisely described that it can be simulated by a machine.

Do you agree that it ought to be true this can be done?

I believe that we ought to agree a simpler goal would be superior to the unnecessarily complicated goal concerning AI I originally posted. As the objective here ought to be to better understand one another I realize that I should change, for the purposes of our discussion, what ought to be understood as unnecessarily complicated (dare I say irrelevant) wording.

Because it has been verified such things count I have now done so.

In believing I ought have latitude I chose a less relevant goal than I should have. I ought to apologize. I shouldn't appear casual concerning goals which ought to be understood as valuable.

Should you accept this apology I ought promise, as a matter of principle and in the interest of relevance and parsimony, to confine my attention to only what we ought to agree is important.

 

[AkuManiMani]

I believe that we ought to agree a simpler goal would be superior to the unnecessarily complicated goal concerning AI I originally posted. As the objective here ought to be to better understand one another I realize that I should change, for the purposes of our discussion, what ought to be understood as unnecessarily complicated (dare I say irrelevant) wording.

Because it has been verified such things count I have now done so.

In believing I ought have latitude I chose a less relevant goal than I should have. I ought to apologize. I shouldn't appear casual concerning goals which ought to be understood as valuable.

Should you accept this apology I ought promise, as a matter of principle and in the interest of relevance and parsimony, to confine my attention to only what we ought to agree is important.

Oh-ho! I see wut you didz thar! xD

 

[Ichneumonwasp]

Actually, a computer doesn't work through code. Not on a physical level.

Sure, when you program, you're defining variables, assigning values, writing functions that change those values, and so forth.

But you do understand, I trust, that this only exists on a symbolic level because we've built the interface to allow us to interact with it that way.

The actual physical machine has no variables in it. Just machine parts.

You're discussing the workings of the machine on an abstract informational level, not a physical level. I'm discussing the operation of the machine on a physical level, which is all that is objectively real.

Our brains are built to engage in entification -- that's what allows us to do much of what we do. And if you're accustomed to regularly discussing the actions of certain machines in that way, it's easy to get into the mindset that this is what these machines objectively do. But that is an error.



If you look at what the machine is physically doing, and only that... just at the physical processes which form the dog's-body between you punching numbers on the keypad and a set of pixels lighting up on the screen... you'll find that there's nothing involved except matter, energy, and the laws of physics -- the same stuff that governs the behavior of everything else in the universe.

It's just that we've designed this machine so that we can "tell it what to do" by using an interface that we have designed so that it involves apparent actions such as defining variables, assigning values, and writing functions.

If the function of the machine actually did require the objective presence of symbols -- well, first of all, you'd be out of luck because that in itself is an impossibility, but if it did you'd be able to find those symbols in the physical machine. And of course they're not there.




Actually, it's not. It's designed to have predictable electro-physical responses (such as lighting up certain arrays of pixels, or spraying patterns of ink on paper, or vibrating a speaker cone, or moving a mechanical arm) to certain electro-physical actions that are done to it, such as pressing keys.

There's nothing more than that happening. And as I understand the basic laws of physics, nothing more that possibly could be happening.

There are machines at my place of work which literally add. They take a certain number of things from one place and a certain number of things from another place and they put them all in the same place. That's what they physically do.

Calculators, however, do not add, except as far as we imagine them to.

The example of someone telling you to take a left turn is more complex, but in any case, all we can say is that their body has made noise. Your brain has evolved to have physical responses to that noise, and those responses depend on many variables, including your brain's history of exposure to language.

Even if you don't understand their language, then from their (subjective) point of view or from an abstract linguistic point of view, yes, they've said "Take a left turn" or some equivalent.

But if there's a calculator performing an addition routine repeatedly, and our time-bomb virus goes off and all life on earth dies, then it's impossible to say that this machine is somehow "adding" anything at all. It's just doing what it does physically, which does not involve adding anything to anything.



You can't stop an abacus from adding because you can't start an abacus adding, at least when you're dealing with large numbers. You can only start it moving beads, and if no one understand the symbolism of that act, then addition has not even occurred on a symbolic level, which is the only level on which it can occur, unless you're physically taking things and aggregating them.

I am not looking at this only from the perspective of the program (the abstraction) but from the interaction of the program and the machine workings.

It makes no sense to speak only of what happens at the machine level and leave out the program. The whole point of computer function is that the machine is controllable and not randomly opening gates. The machine does what it does because of the program. The program directs the machine to open certain gates at one time and close others. While all the machine does is open and close gates, those actions are already defined by the machine code/assembly code/whatever we call these things to have meaning. We discuss the program because it is easier to discuss, but even it is not a pure abstraction. It is able to guide the machine to do certain things by interfacing with the other levels of computer language.

If we think in terms of a program as a pure abstraction and the internal workings of a machine as electrons and gates, how does a computer do anything?

It works because a program is not just a pure abstraction. It controls from the top down how the machine works. A program that codes for the machine to perform the actions that are addition (which gates open when and where) will continue to work and perform addition whether or not anyone looks at it. The gates don't open and close randomly. They open and close intentionally in order to perform that algorithm that we set it to do.

We could certainly start an abacus moving beads following an algorithm that anyone who knows the symbol system would identify as adding. It wouldn't do this randomly, but according to the algorithm. So it adds. Whether or not a person who happens by understands that it is following that algorithm and adding is not important to the issue of it adding. It still adds whether or not anyone looks at it. You are still describing the understanding of adding; not adding itself.

 

[rocketdodger]

I believe that we ought to agree a simpler goal would be superior to the unnecessarily complicated goal concerning AI I originally posted. As the objective here ought to be to better understand one another I realize that I should change, for the purposes of our discussion, what ought to be understood as unnecessarily complicated (dare I say irrelevant) wording.

Because it has been verified such things count I have now done so.

In believing I ought have latitude I chose a less relevant goal than I should have. I ought to apologize. I shouldn't appear casual concerning goals which ought to be understood as valuable.

Should you accept this apology I ought promise, as a matter of principle and in the interest of relevance and parsimony, to confine my attention to only what we ought to agree is important.

This might help:

http://www.merriam-webster.com/dictionary/ought

As usual, you are simply using a word and giving it some kind of arbitrary root definition that in your mind ( and your mind alone ) is irreducible.

But, you are wrong -- "ought" has a definition, it has many definitions, and all the ways you are using it here are certainly describable by mathematics.

For example, the definitions given by merriam-webster:

1) used to express obligation -- why is this not mathematically describable? Is the concept of obligation somehow magical and beyond description? No, of course not.

2) advisability -- again, is this somehow magical and beyond description? No.

3) expectation -- again, is this somehow magical and beyond description? No.

4) logical consequence -- obviously this is describable, it isn't even a question.

 

[rocketdodger]

rocketdodger, I see you've been on this thread since my last reply to you. Would you mind responding to my post about what is happening when I wake up?

Can you explain what's going on in terms of information processing, if that's not just shorthand for a physical process (as "fuel processing" would be for what's happening in my truck's engine)?

I'm quite serious about that. Can we just focus on that example for a bit?

What is your explanation for the phenomenon of waking up and becoming conscious, and how does your explanation allow for the existence of conscious computer programs?

Thanks.

ETA: QuixoteCoyote, you can chime in, too, since I asked you essentially the same question.

After I am done with work, Piggy, that kind of post takes too much time for me to do it during a compile or build.

 

[AkuManiMani]

This might help:

http://www.merriam-webster.com/dictionary/ought

As usual, you are simply using a word and giving it some kind of arbitrary root definition that in your mind ( and your mind alone ) is irreducible.

But, you are wrong -- "ought" has a definition, it has many definitions, and all the ways you are using it here are certainly describable by mathematics.

For example, the definitions given by merriam-webster:

1) used to express obligation -- why is this not mathematically describable? Is the concept of obligation somehow magical and beyond description? No, of course not.

2) advisability -- again, is this somehow magical and beyond description? No.

3) expectation -- again, is this somehow magical and beyond description? No.

4) logical consequence -- obviously this is describable, it isn't even a question.

Erm...RD, to put it mildly I think Frank was just mocking your lack of -- uhm -- concision :-X

 

[Piggy]

I am not looking at this only from the perspective of the program (the abstraction) but from the interaction of the program and the machine workings.

Well, I would agree with you there.

Does a digital calculator add?

Of course. That's what we build them to do. And if they didn't, we wouldn't buy them and use them.

Do they process information? Absolutely.

It's after midnight, I'm just home from work, and I need to get some sleep, but hopefully I can pick this up again sometime this weekend.

 

[Piggy]

After I am done with work, Piggy, that kind of post takes too much time for me to do it during a compile or build.

That's ok.

It's funny, today I was thinking that you and I probably agree on much more than is apparent from these threads. (And my often combative stance is probably much to blame for that circumstance.)

In fact, I'd like to explore that territory some more, but I don't have time to get into that at the moment (for the same reasons you cite).

What I'll say for now is this.... When it comes to the majority of brain processes, I absolutely agree with you that it makes perfect sense to speak of the brain as an organic computer. Or, more accurately, to speak of computers as artificial brains.

It's also undeniable that the brain processes information, in a way that a liver or a calf muscle does not.

"Information processing" is an extremely useful and enlightening framework with which to view and model the operations of both computers and brains. And experimentation has definitively shown the cinture to be accurate.

I doubt I'll post tomorrow, but I'll try to chime back in over the weekend, and maybe we can explore some common ground, and perhaps that will help illuminate what our real differences are.

 

[Piggy]

If we think in terms of a program as a pure abstraction and the internal workings of a machine as electrons and gates, how does a computer do anything?

It works because a program is not just a pure abstraction. It controls from the top down how the machine works. A program that codes for the machine to perform the actions that are addition (which gates open when and where) will continue to work and perform addition whether or not anyone looks at it. The gates don't open and close randomly. They open and close intentionally in order to perform that algorithm that we set it to do.

Well, that's an interesting puzzle, because if you talk about logic gates, what are you actually talking about?

To what degree is that an abstraction, and to what degree does it correlate with physical reality?

If you say it correlates perfectly, then it would make sense to ask what the part numbers are for the logic gates, and if you can purchase them by the dozen, or if you open the calculator up then where are they?

But if you deny that the term refers to anything objective, you're also in trouble, because the behavior of the machine can indeed be described accurately in those terms.

Clearly, the calculator does operate according to logical rules, and clearly those rules have been designed into the machine.

On the other hand, the machine itself and all of its behavior can be completely described in terms of matter and energy with no reference whatsoever to logic.

Touchy stuff, this.

More later, when it's not in the wee hours and I've gotta work in the morning.

 

[Ichneumonwasp]

Well, that's an interesting puzzle, because if you talk about logic gates, what are you actually talking about?

To what degree is that an abstraction, and to what degree does it correlate with physical reality?

If you say it correlates perfectly, then it would make sense to ask what the part numbers are for the logic gates, and if you can purchase them by the dozen, or if you open the calculator up then where are they?

But if you deny that the term refers to anything objective, you're also in trouble, because the behavior of the machine can indeed be described accurately in those terms.

Clearly, the calculator does operate according to logical rules, and clearly those rules have been designed into the machine.

On the other hand, the machine itself and all of its behavior can be completely described in terms of matter and energy with no reference whatsoever to logic.

Touchy stuff, this.

More later, when it's not in the wee hours and I've gotta work in the morning.

The logic gates are a component part of the chip. It makes sense to ask where we can buy chips -- Intel (sorry AMD).

 

[westprog]

Well, that's an interesting puzzle, because if you talk about logic gates, what are you actually talking about?

To what degree is that an abstraction, and to what degree does it correlate with physical reality?

If you say it correlates perfectly, then it would make sense to ask what the part numbers are for the logic gates, and if you can purchase them by the dozen, or if you open the calculator up then where are they?

But if you deny that the term refers to anything objective, you're also in trouble, because the behavior of the machine can indeed be described accurately in those terms.

Clearly, the calculator does operate according to logical rules, and clearly those rules have been designed into the machine.

On the other hand, the machine itself and all of its behavior can be completely described in terms of matter and energy with no reference whatsoever to logic.

Touchy stuff, this.

More later, when it's not in the wee hours and I've gotta work in the morning.

It's certainly possible to describe a computer and logic gates. However, is it possible to provide an unambiguous physical description, so that we can objectively recognise a logic gate as defined by our operational description so that the two concepts match up.

This is the key problem for computationalism. It's the basis for the argument which has been going on for the last few thousand posts.

 

[Myriad]

Well, that's an interesting puzzle, because if you talk about logic gates, what are you actually talking about?

To what degree is that an abstraction, and to what degree does it correlate with physical reality?

It's certainly possible to describe a computer and logic gates. However, is it possible to provide an unambiguous physical description, so that we can objectively recognise a logic gate as defined by our operational description so that the two concepts match up.

This is the key problem for computationalism. It's the basis for the argument which has been going on for the last few thousand posts.

Of course logic gates are physically real. Look up TTL chips. Four logic gates per 14-pin DIP chip (about the size of 16 gigabytes of flash RAM today). 7400 was the quad NAND. I used to have a lot of the part numbers (and some of the pinouts) memorized -- though since you can build anything out of enough quad NANDs, I suppose I still remember enough to get by. High technology, in the 1970s.

Logic gates are just as real today, just a lot smaller and a lot more of them in the package.

Of course, there are a lot of different ways to make them. So what? If homogeneity were a requirement for real existence, then bridges and cookies wouldn't be real either. (Can you specify how to recognize a bridge without resorting to any abstract, conceptual, operational concepts such as "spanning" or "crossing over" something?)

Respectfully,
Myriad

 

[Piggy]

Of course logic gates are physically real. Look up TTL chips. Four logic gates per 14-pin DIP chip (about the size of 16 gigabytes of flash RAM today). 7400 was the quad NAND. I used to have a lot of the part numbers (and some of the pinouts) memorized -- though since you can build anything out of enough quad NANDs, I suppose I still remember enough to get by. High technology, in the 1970s.

Logic gates are just as real today, just a lot smaller and a lot more of them in the package.

Of course, there are a lot of different ways to make them. So what? If homogeneity were a requirement for real existence, then bridges and cookies wouldn't be real either. (Can you specify how to recognize a bridge without resorting to any abstract, conceptual, operational concepts such as "spanning" or "crossing over" something?)

Respectfully,
Myriad

In this case, my questions to Wasp actually weren't purely rhetorical, so I do appreciate the input.

For any given chip, can you specify what percentage of the chip's weight constitutes the logic gate? Can you cut the logic gate out of it in the same way you can cut the gate out of a fence?

If so, then can we call it some other sort of gate without resorting to the term "logic" which necessarily denotes abstraction?

As for a bridge, we get into the same problem when we put ourselves into a situation in which it's important to distinguish between objective physical reality and abstractions.

A bridge is a functional thing, requiring a functional definition. Anything can be a bridge as long as it performs that function.

On the other hand, if you take an old wooden bridge, take it apart, and re-construct it in a museum, it's still a "bridge" even though it doesn't perform that function anymore.

The whole point of this part of the discussion is to distinguish between what's objectively real and what is an abstraction.

But that doesn't mean that abstractions are not anchored in OPR.

Our problem comes when we confuse the two and come to think, for example, that a STOP sign actually contains information. (It doesn't, of course.)

Do our brains contain information? Well, sure. As long as we keep in mind that "information" is a kind of short-hand.

But if we want to discuss brains as physical objects, then no, that drops off the radar. Ultimately, everything a brain does must be describable in terms of matter and energy without any reference to logic or information.

That doesn't mean that we're somehow wrong or inaccurate when we use those terms to discuss what the brain is doing.

I don't think anyone has a problem with talking about the brain in computational terms, as long as we stick to demonstrated areas of similarity between brains and computers.

Similarly, we can discuss functional similarities between the legs we're born with and artificial legs that amputees use.

We design artificial legs precisely so they can perform certain functions of real legs.

But we would not then presume to assert that artificial legs of the type we have now must be able to perform every function performed by organic legs, such as mytosis, because we have not designed them to do that.

Similarly, we have designed computers to perform certain tasks performed by our brains -- and they often do a wonderful job of this -- as well as tasks that our brains do not perform such as load balancing at power plants or running simulations of weather systems.

This does not mean that they can also perform any and every task performed by organic brains, especially considering that we do not know how some of those tasks are performed.

Added into the mix, we find that the kinds of things we can pull off with computers -- perception, memory, response, learning -- can be done by the brain without involving those processes responsible for conscious awareness.

So if we do design a conscious machine -- which should be theoretically possible -- there's no reason to believe that it could not incorporate a computer.

On the other hand, this is a far cry from asserting that it's possible to produce a "conscious program". That's an entirely different kettle of fish.

And it's that claim specifically which physicalists object to, and with good reason.

 

[PixyMisa]

Of course logic gates are physically real. Look up TTL chips. Four logic gates per 14-pin DIP chip (about the size of 16 gigabytes of flash RAM today). 7400 was the quad NAND. I used to have a lot of the part numbers (and some of the pinouts) memorized -- though since you can build anything out of enough quad NANDs, I suppose I still remember enough to get by. High technology, in the 1970s.

Logic gates are just as real today, just a lot smaller and a lot more of them in the package.

Of course, there are a lot of different ways to make them. So what? If homogeneity were a requirement for real existence, then bridges and cookies wouldn't be real either. (Can you specify how to recognize a bridge without resorting to any abstract, conceptual, operational concepts such as "spanning" or "crossing over" something?)

Right. Logic gates are defined by their behaviour. So, ultimately, is everything else. People need to learn to stop asking what things are and ask instead what they do.

 

[PixyMisa]

Actually, a computer doesn't work through code. Not on a physical level.

Of course it does. Code is physical. If it weren't, it wouldn't do anything.

 

[westprog]

Of course logic gates are physically real. Look up TTL chips. Four logic gates per 14-pin DIP chip (about the size of 16 gigabytes of flash RAM today). 7400 was the quad NAND. I used to have a lot of the part numbers (and some of the pinouts) memorized -- though since you can build anything out of enough quad NANDs, I suppose I still remember enough to get by. High technology, in the 1970s.

Logic gates are just as real today, just a lot smaller and a lot more of them in the package.

Of course, there are a lot of different ways to make them. So what? If homogeneity were a requirement for real existence, then bridges and cookies wouldn't be real either. (Can you specify how to recognize a bridge without resorting to any abstract, conceptual, operational concepts such as "spanning" or "crossing over" something?)

Respectfully,
Myriad

I said that they were real. The question is not whether a logic gate is physically real or not. The question is whether it is possible to come up with a physical description of a logic gate that encompasses what it does, rather than a functional description which describes the use we make of it, or a manufacturing description that describes what they are made of.

Yes, bridges are real, material things. The issue is whether we can produce an unambiguous physical description of a bridge which includes all the things which we consider to be a bridge, and excludes all the things we don't. I suspect that such a description could be made, but it would be extremely cumbersome. Physical concepts are tricky things to define. Generally, the more complex the concept, the more difficult it is to produce a physical definition.

So why does this matter? We don't need a physical definition in order to make computers. We need an engineering specification, and a mathematical analysis. We understand the physics of what's going on, so why is this even an issue?

It's only an issue if there's a claim that logic gates, acting as logic gates, produce a specific physical effect. It's not possible to claim that something produces a physical effect, and not produce a physical definition of what is producing the effect. In particular, when the claim is that logic gates and neurons are in some way the same thing, and if it's claimed that they are producing exactly the same experience, then a precise physical definition is essential.

This has been gone over again and again. There have been attempts to produce physical definitions, but none of them have been remotely precise enough.

 

[Ichneumonwasp]

I said that they were real. The question is not whether a logic gate is physically real or not. The question is whether it is possible to come up with a physical description of a logic gate that encompasses what it does, rather than a functional description which describes the use we make of it, or a manufacturing description that describes what they are made of.

Yes, bridges are real, material things. The issue is whether we can produce an unambiguous physical description of a bridge which includes all the things which we consider to be a bridge, and excludes all the things we don't. I suspect that such a description could be made, but it would be extremely cumbersome. Physical concepts are tricky things to define. Generally, the more complex the concept, the more difficult it is to produce a physical definition.

So why does this matter? We don't need a physical definition in order to make computers. We need an engineering specification, and a mathematical analysis. We understand the physics of what's going on, so why is this even an issue?

It's only an issue if there's a claim that logic gates, acting as logic gates, produce a specific physical effect. It's not possible to claim that something produces a physical effect, and not produce a physical definition of what is producing the effect. In particular, when the claim is that logic gates and neurons are in some way the same thing, and if it's claimed that they are producing exactly the same experience, then a precise physical definition is essential.

This has been gone over again and again. There have been attempts to produce physical definitions, but none of them have been remotely precise enough.

I'm sorry, but that is not even a worthwhile issue to bring up in such a discussion. We all know Wittgenstein. Definitions are word games; all definitions share family resemblances. That tells us nothing about the nature of the world but only about how we use language and define words. It will not help you.


The proper physical 'definition' of a logic gate is a logic gate. Look at what it does. That is its definition. Words are words.

 

[AkuManiMani]

I'm sorry, but that is not even a worthwhile issue to bring up in such a discussion. We all know Wittgenstein. Definitions are word games; all definitions share family resemblances. That tells us nothing about the nature of the world but only about how we use language and define words. It will not help you.


The proper physical 'definition' of a logic gate is a logic gate. Look at what it does. That is its definition. Words are words.

That being said, consciousness is not a word, symbol, or abstraction. Its a concrete reality for which we don't have a scientific understanding. Therefore, merely abstracting operational byproducts and phenomena associated with consciousness is not sufficient to produce consciousness. A proper physical 'definition' of consciousness would not be a "logic gate".