Machine Consciousness Experiment Conceivability

[psyreporter]

The assignment of meaning always involves valuing and as such, it always results in value.

Well, no, it doesn't. At least, not according to the dictionary definitions of those two words. Perhaps you can spell out your definitions.

What is meaning when it is to be considered as a substance or property of the Universe? Empirical evidence for meaning is not possible and when one attempts to answer the question empirically, one arrives at what is named 'value' (assigned meaning).

The simplest departure from pure randomness implies value (assigned meaning). This is evidence that all that can be seen in the world – from the simplest pattern onward – is value.

[GE Morton]

The assignment of meaning always involves valuing and as such, it always results in value.

Well, no, it doesn't. At least, not according to the dictionary definitions of those two words. Perhaps you can spell out your definitions.

What is meaning when it is to be considered as a substance or property of the Universe? Empirical evidence for meaning is not possible and when one attempts to answer the question empirically, one arrives at what is named 'value' (assigned meaning).

Ah. This "meaning" of yours, then, is a "substance or property of the universe"? But one for which there is no empirical evidence? What, then, IS the evidence for it? Or if there is no evidence, then what is the basis of your claim that it exists? And why give this mysterious property a name --- "meaning" --- which denotes something entirely different in common usage?

Value, of course, is hardly a "property of the universe," either. It is not a property of anything. It is a pseudo-property we assign to things to denote their relative desirability by us, with the value of anything varying from valuer to valuer. Any given thing has as many values as there are people who desire it and are willing to invest some time, effort, resources to acquiring it. How much of those resources they're willing to invest is the measure of its value to them, and, again, varies from valuer to valuer.

The simplest departure from pure randomness implies value (assigned meaning).

Oh, it implies no such thing. Those three concepts --- value, meaning, randomness --- have nothing whatever to do with each other. At least, per the dictionary definitions of those terms. But of course, if you invent new meanings for common terms you can make any logical connection between them you wish.

[psyreporter]

Ah. This "meaning" of yours, then, is a "substance or property of the universe"? But one for which there is no empirical evidence? What, then, IS the evidence for it? Or if there is no evidence, then what is the basis of your claim that it exists? And why give this mysterious property a name --- "meaning" --- which denotes something entirely different in common usage?

Consciousness can be said to be 'meaningful experience' (meaning is the characteristic of consciousness that is unable to be grasped or explained by empirical science). The evidence for the substantial and fundamental nature of 'meaning' therefore would be one's own conscious (meaningful) experience, similar to the argument that consciousness is not an illusion while empirical evidence is lacking and supposedly impossible.

With regard common usage of the term meaning. The reference 'meaning of ...' always refers to value. That which is indicated would have to have been assigned meaning for it to be possible to be referenced in the question. The usage of the term meaning allows re-evaluation of value which allows it to become part of meaningful experience.

As an example: "What is the meaning of the word cat?" would allow one to become aware of the existence of the cat animal and all that is possibly related to it within the scope of meaningful experience (e.g. "Cats are pets" or "My sister loves cats"). The meaning that is indicated would refer specifically to the cat animal however it is obvious that it involves meaningful experience to be able to use the term as a denotation, hence the use of the term meaning.

Value, of course, is hardly a "property of the universe," either. It is not a property of anything. It is a pseudo-property we assign to things to denote their relative desirability by us, with the value of anything varying from valuer to valuer. Any given thing has as many values as there are people who desire it and are willing to invest some time, effort, resources to acquiring it. How much of those resources they're willing to invest is the measure of its value to them, and, again, varies from valuer to valuer.

I would disagree that value has anything to do with subjective desirability.

Most simply said, value is the result of valuing. When one looks at the nature of valuing it is evident that it involves the assignment of meaning. Therefore, in the most basic sense, value is assigned meaning (meaning in a retro-perspective).

The simplest departure from pure randomness implies value (assigned meaning).

Oh, it implies no such thing. Those three concepts --- value, meaning, randomness --- have nothing whatever to do with each other. At least, per the dictionary definitions of those terms. But of course, if you invent new meanings for common terms you can make any logical connection between them you wish.

Pure randomness equals to pure nothingness (zero meaning). Randomness cannot exist, with as an example result that cryptographic algorithms are always able to be broken with sufficient computing power.

(2020) Randomness theory could hold key to internet security
Researchers identified a problem that holds the key to whether all encryption can be broken -- as well as a surprising connection to a mathematical concept that aims to define and measure randomness.
https://www.sciencedaily.com/releases/2 ... 194721.htm

Therefore, one is to consider a factor to be involved that prevents actual randomness to be possible, which is meaning. This is evidence that meaning is fundamental to the Universe.

[SteveKlinko]

Yes but the Magic of the Conscious Visual Experience happens after what the Cortex does.

Of course. Effects always follow causes in time. In this case, the time it takes for the signals from the visual cortex to reach the frontal lobes (a few milliseconds).

I think that Science will discover a new Consciousness Phenomena and will make it an integral part of Science.[/quote]

A "new" conscious phenomenon? Where would one look for such a thing? How would it be recognized as such --- what determining properties must it have to warrant being called "conscious"?
[/quote]


These are the questions that must be answered. It will take a lot of thinking outside the Physicalist Box to get there.

That does not make Conscious Experience a Physical Phenomenon but it will be a begrudging admission by Science that it does not know everything right now. Science will need to change its' Perspective and think outside of the Box it is in.

Well, no physicist (that I know of) claims his science knows everything. But we already know that consciousness is a physical phenomenon, because it is produced by, and only by, physical systems. The decisive test will be whether we can build a machine that can pass the Turing test.

Many Scientists do in fact claim that the Science is closed and no new Phenomenon remains to be discovered. This is why Physicalists must insist that Consciousness is in the Neurons or some effect of the Neurons. They have nowhere else to look.

The Consciousness that you know is Correlated with Physical Systems, but you have no way of knowing what other manifestations of Consciousness exist because you, and everybody else (me too) have no idea what we are talking about. We can only try to push the ball closer to the goal in the best way we can invent. I don't know what the Experience of Redness is, do you?

[jvh]

The generation of machine consciousness will not succeed until a significant barrier is overcome. That barrier is the development of a model of the function of consciousness. There is a natural tendency to want to get on with the experiment and worry about the theory later. However in this case there is a fundamental problem with the lack of a functional model of consciousness. That is without at least a first draft understanding of the purpose of consciousness and what it is trying to achieve, how will it be possible to test whether machine consciousness has been achieved or not ?

The test is the thing, without a test it will not be possible to demonstrate machine consciousness, and a test can only be considered when a functional model of consciousness, even a rough one, is developed.

John

But the right type of test could provide a fundamental clue for developing Models.

The Turing test as a measure of consciousness or at least intelligence was mentioned, but it is too subjective, in the sense that two different testers could reach different opinions regarding the same test subject. What is needed is an objective, or at least somewhat more objective, test, and one that is focussed on consciousness. Interestingly there is such a test, which while limited in scope to humans and computers, is, I believe, able to distinguish a conscious system from a non-conscious one. Its called the "Random rule variation chess test"

John

[SteveKlinko]

The generation of machine consciousness will not succeed until a significant barrier is overcome. That barrier is the development of a model of the function of consciousness. There is a natural tendency to want to get on with the experiment and worry about the theory later. However in this case there is a fundamental problem with the lack of a functional model of consciousness. That is without at least a first draft understanding of the purpose of consciousness and what it is trying to achieve, how will it be possible to test whether machine consciousness has been achieved or not ?

The test is the thing, without a test it will not be possible to demonstrate machine consciousness, and a test can only be considered when a functional model of consciousness, even a rough one, is developed.

John

But the right type of test could provide a fundamental clue for developing Models.

The Turing test as a measure of consciousness or at least intelligence was mentioned, but it is too subjective, in the sense that two different testers could reach different opinions regarding the same test subject. What is needed is an objective, or at least somewhat more objective, test, and one that is focussed on consciousness. Interestingly there is such a test, which while limited in scope to humans and computers, is, I believe, able to distinguish a conscious system from a non-conscious one. Its called the "Random rule variation chess test"

John

Googled "Random rule variation chess test" and got a bunch of possibilities for what this is. I couldn't see how any of the possibilities has anything to do with distinguishing Conscious from Non Conscious. How would you explain how it works to do this?

[jvh]

It's mentioned in my book "Mind Consciousness and the Nature of Living Things" published on Amazon. I quote from the book below:

You play chess with your nephew aged nine. He has lost a pawn from his chess set, so to be fair you also discard a pawn, and play a game. The boy suggests that in future some random variation of the chess rules will be agreed at the start of each game to make it more interesting. You both enjoy this, so you try it with your advanced computer chess programme. However, there is a problem, the software does not permit rule changes. You contact the software designer, who says changing the algorithm to omit a pawn is no problem but handling random rule changes, that are unknown in advance of the game, is impossible even for a supercomputer or AI, despite being child’s play for a conscious mind.

Is the problem with the designer, would a more brilliant one be able to devise an algorithm that can accept rule changes that are unknown in advance, or is the problem inherently unsolvable by conventional computer? It would be an interesting challenge to develop software that can play “random rule variation chess” . If it is impossible to develop such software, because the problem is non-algorithmic, a rule variation trial might be a test of black box consciousness.

Other examples of random chess rule changes, would be:
• identify a square in the middle of the board where no piece can land;
• Add a new square to the side of the board;
• Drop the castling rule;
• A specific pawn on both sides is changed to a rook.

The rule change is secret until the start of the game, the computers software cannot be changed during the game. Each game is given a unique rule change and they are not reused.

Further tests to determine whether decision making is painful and instantaneous would be needed to fully determine consciousness.

A test of whether a creature’s consciousness is active or inactive at a specific moment is a separate issue.

A couple of extra points, firstly the computer may not have its software changed by an external programmer, it is permitted of course for the computer to modify its own software during the game. A second point is that the test is currently only formulated for humans and computers as both can play chess, its extension to non human organisms would be an interesting line of research.

[SteveKlinko]

It's mentioned in my book "Mind Consciousness and the Nature of Living Things" published on Amazon. I quote from the book below:

You play chess with your nephew aged nine. He has lost a pawn from his chess set, so to be fair you also discard a pawn, and play a game. The boy suggests that in future some random variation of the chess rules will be agreed at the start of each game to make it more interesting. You both enjoy this, so you try it with your advanced computer chess programme. However, there is a problem, the software does not permit rule changes. You contact the software designer, who says changing the algorithm to omit a pawn is no problem but handling random rule changes, that are unknown in advance of the game, is impossible even for a supercomputer or AI, despite being child’s play for a conscious mind.

Is the problem with the designer, would a more brilliant one be able to devise an algorithm that can accept rule changes that are unknown in advance, or is the problem inherently unsolvable by conventional computer? It would be an interesting challenge to develop software that can play “random rule variation chess” . If it is impossible to develop such software, because the problem is non-algorithmic, a rule variation trial might be a test of black box consciousness.

Other examples of random chess rule changes, would be:
• identify a square in the middle of the board where no piece can land;
• Add a new square to the side of the board;
• Drop the castling rule;
• A specific pawn on both sides is changed to a rook.

The rule change is secret until the start of the game, the computers software cannot be changed during the game. Each game is given a unique rule change and they are not reused.

Further tests to determine whether decision making is painful and instantaneous would be needed to fully determine consciousness.

A test of whether a creature’s consciousness is active or inactive at a specific moment is a separate issue.

A couple of extra points, firstly the computer may not have its software changed by an external programmer, it is permitted of course for the computer to modify its own software during the game. A second point is that the test is currently only formulated for humans and computers as both can play chess, its extension to non human organisms would be an interesting line of research.

How is the Computer going to be able to be told that there is a rule change if it cannot be reprogrammed? Or do you just type in: Don't use middle square? Ahhh, ok the Computer then must be programmed ahead of time to be able to adapt to rule changes. I think there are algorithms that can adapt to changes in rules. I'm no expert in programming Chess playing Computers but I'll bet it can be done.

But I still don't see how this has anything to do with Conscious or Non Conscious detection. If a Human can beat the Computer with random rule changes, it only means that the Human Brain Machine is better at dealing with randomness than the Silicon Machine. A better ability to deal with randomness does not necessarily correlate to Consciousness.

[SteveKlinko]

It's mentioned in my book "Mind Consciousness and the Nature of Living Things" published on Amazon.

By the way, I searched for your title on Amazon and your book did not come up.

[GE Morton]

A "new" conscious phenomenon? Where would one look for such a thing? How would it be recognized as such --- what determining properties must it have to warrant being called "conscious"?

These are the questions that must be answered. It will take a lot of thinking outside the Physicalist Box to get there.

You need to answer those questions before you can launch any productive search.

Many Scientists do in fact claim that the Science is closed and no new Phenomenon remains to be discovered.

Well, there are thousands of scientists in the world, and no doubt some of them hold some pretty bizarre views. But I know of none with any prominence who believes that nothing remains to be discovered.

This is why Physicalists must insist that Consciousness is in the Neurons or some effect of the Neurons. They have nowhere else to look.

That consciousness is an effect of (certain configurations of) neurons is a certainty, easily verified empirically and beyond dispute. There is no need to look elsewhere. The term "consciousness" denotes two things: the phenomena of direct experience, which is subjective and wholly dependent upon certain physical systems, and a pseudo-property we impute to creatures who display certain patterns of behavior. Considering it to be a "non-physical substance" or an "additional plane or realm of existence" or a "transcendental aspect of reality," etc., is just a flight of fancy with no empirical anchors; an unverifiable, unfalsifiable conjecture of the same ilk as postulates of "spirit worlds" and supernatural beings. It is a speculation which ignores the actual usage of that term in common language.

The Consciousness that you know is Correlated with Physical Systems, but you have no way of knowing what other manifestations of Consciousness exist because you, and everybody else (me too) have no idea what we are talking about.

You're correct there --- I have no idea what you're talking about because you not only have not produced any manifestations of consciousness that are not products of physical systems, but cannot tell us where to look for them or how we might identify them if we found one. You're asking us to look for a needle in a haystack, without explaining what either a needle or a haystack looks like, or where the haystacks are to be found.

I don't know what the Experience of Redness is, do you?

Yes, I do. It is a particular, distinctive, subjective impression I experience when beholding a ripe tomato or a stop sign. It is the form in which neural signals of a certain type are represented in (my) consciousness.

[GE Morton]

You play chess with your nephew aged nine. He has lost a pawn from his chess set, so to be fair you also discard a pawn, and play a game. The boy suggests that in future some random variation of the chess rules will be agreed at the start of each game to make it more interesting. You both enjoy this, so you try it with your advanced computer chess programme. However, there is a problem, the software does not permit rule changes. You contact the software designer, who says changing the algorithm to omit a pawn is no problem but handling random rule changes, that are unknown in advance of the game, is impossible even for a supercomputer or AI, despite being child’s play for a conscious mind.

First, a chess-playing program is not an AI, at least not one which would pass the Turing test. The chess program would merely be a subroutine within that AI, the rules of which could be changed at will (or per the agreement you suggest) by other routines within the AI.

[SteveKlinko]

A "new" conscious phenomenon? Where would one look for such a thing? How would it be recognized as such --- what determining properties must it have to warrant being called "conscious"?

These are the questions that must be answered. It will take a lot of thinking outside the Physicalist Box to get there.

You need to answer those questions before you can launch any productive search.

I am trying to answer these questions with the Machine Consciousness Experiments. I'm sorry that you have missed the whole point of the Experiments.

Many Scientists do in fact claim that the Science is closed and no new Phenomenon remains to be discovered.

Well, there are thousands of scientists in the world, and no doubt some of them hold some pretty bizarre views. But I know of none with any prominence who believes that nothing remains to be discovered.

Not just that nothing remains to be discovered, but that completely New and Paradigm shifting things need to be discovered.

This is why Physicalists must insist that Consciousness is in the Neurons or some effect of the Neurons. They have nowhere else to look.

That consciousness is an effect of (certain configurations of) neurons is a certainty, easily verified empirically and beyond dispute. There is no need to look elsewhere. The term "consciousness" denotes two things: the phenomena of direct experience, which is subjective and wholly dependent upon certain physical systems, and a pseudo-property we impute to creatures who display certain patterns of behavior. Considering it to be a "non-physical substance" or an "additional plane or realm of existence" or a "transcendental aspect of reality," etc., is just a flight of fancy with no empirical anchors; an unverifiable, unfalsifiable conjecture of the same ilk as postulates of "spirit worlds" and supernatural beings. It is a speculation which ignores the actual usage of that term in common language.

All Science knows for sure is that Conscious Experience seems to be Correlated with Neural Activity. There is exactly Zero explanation for how this happens. So the statement that Consciousness is Certainly in the Neurons is Logically and Evidently not true. The theory of Connectism is just as verifiable as any of the "In The Neurons Theories". Connectism says that Consciousness is Connected to Neurons but is yet a separate Phenomenon from the Neurons and is not created by the Neurons. So I can say that Connectism is easily verified empirically and beyond dispute just as validly as with the "In The Neurons theories".

The Consciousness that you know is Correlated with Physical Systems, but you have no way of knowing what other manifestations of Consciousness exist because you, and everybody else (me too) have no idea what we are talking about.

You're correct there --- I have no idea what you're talking about because you not only have not produced any manifestations of consciousness that are not products of physical systems, but cannot tell us where to look for them or how we might identify them if we found one. You're asking us to look for a needle in a haystack, without explaining what either a needle or a haystack looks like, or where the haystacks are to be found.

I don't know what the Experience of Redness is, do you?

Yes, I do. It is a particular, distinctive, subjective impression I experience when beholding a ripe tomato or a stop sign. It is the form in which neural signals of a certain type are represented in (my) consciousness.

You have a Huge Explanatory Gap in going from Signals to Representations in (your) Consciousness. You have Explained nothing and have revealed that you don't have an Explanation for what the Experience of Redness is.

[GE Morton]

I am trying to answer these questions with the Machine Consciousness Experiments. I'm sorry that you have missed the whole point of the Experiments.

You've missed the point of the comment, namely, that you need to have answers to those questions before you set up any experiments; if you don't, you will have no means of interpreting the results of those experiments (we covered this).

Not just that nothing remains to be discovered, but that completely New and Paradigm shifting things need to be discovered.

There is nothing new about the assumption that consciousness is a manifestation of some sort of "alternative plane/dimension/realm" of "reality." It is as old as Plato, if not older; also called the "spirit realm," "eternal Forms," etc. That "paradigm" --- substance dualism --- was abandoned by most because claims about that supposed "realm" are unverifiable, unfalsifiable, and theoretically sterile --- they yield no verifiable predictions or any useful information.

All Science knows for sure is that Conscious Experience seems to be Correlated with Neural Activity.

It does not "seem to be" correlated. It is ALWAY so correlated; there are no known exceptions. That relationship is as certain as relationship between mass and gravity.

There is exactly Zero explanation for how this happens. So the statement that Consciousness is Certainly in the Neurons is Logically and Evidently not true.

Consciousness is not "in" neurons. No neuron is conscious. It is an emergent effect produced by certain complex networks of neurons. And there can be no explanation for how that happens, because explanations require that the explicanda be empirically accessible, and conscious phenomena are not; they are private, subjective, ineffable, and unanalyzable.

You have a Huge Explanatory Gap in going from Signals to Representations in (your) Consciousness. You have Explained nothing and have revealed that you don't have an Explanation for what the Experience of Redness is.

See above.

[SteveKlinko]

I am trying to answer these questions with the Machine Consciousness Experiments. I'm sorry that you have missed the whole point of the Experiments.

You've missed the point of the comment, namely, that you need to have answers to those questions before you set up any experiments; if you don't, you will have no means of interpreting the results of those experiments (we covered this).

Yes, and I disagree with your premise that questions need to be answered before the Experiment. The Experiment itself will answer many questions.

Not just that nothing remains to be discovered, but that completely New and Paradigm shifting things need to be discovered.

There is nothing new about the assumption that consciousness is a manifestation of some sort of "alternative plane/dimension/realm" of "reality." It is as old as Plato, if not older; also called the "spirit realm," "eternal Forms," etc. That "paradigm" --- substance dualism --- was abandoned by most because claims about that supposed "realm" are unverifiable, unfalsifiable, and theoretically sterile --- they yield no verifiable predictions or any useful information.

But none of those things have actually been discovered and confirmed.

All Science knows for sure is that Conscious Experience seems to be Correlated with Neural Activity.

It does not "seem to be" correlated. It is ALWAY so correlated; there are no known exceptions. That relationship is as certain as relationship between mass and gravity.

Science cannot know the other things Consciousness might be doing because Science only measures Neural Activity and does not Measure the actual Conscious Experience. Science would have to be able to independently Measure Conscious Experience in some way to come to the conclusions you have stated. Let me repeat: All Science knows is that there seems to be a Correlation of Neural Activity with Conscious Experience as reported by a test subject. That is not a good Scientific Measurement of Conscious Experience, so we should be more careful in jumping to conclusions

There is exactly Zero explanation for how this happens. So the statement that Consciousness is Certainly in the Neurons is Logically and Evidently not true.

Consciousness is not "in" neurons. No neuron is conscious. It is an emergent effect produced by certain complex networks of neurons. And there can be no explanation for how that happens, because explanations require that the explicanda be empirically accessible, and conscious phenomena are not; they are private, subjective, ineffable, and unanalyzable.

True, so any conclusions you have made about Conscious Experiences are Speculations and nothing more. Your talk of certainties is founded on Ineffable and Unanalyzable phenomena.

You have a Huge Explanatory Gap in going from Signals to Representations in (your) Consciousness. You have Explained nothing and have revealed that you don't have an Explanation for what the Experience of Redness is.

See above.

At least you admit there is a Huge Explanatory Gap with your statement. You do not have an Explanation for what the Experience of Redness is. But that's ok, nobody does.

[GE Morton]

I am trying to answer these questions with the Machine Consciousness Experiments. I'm sorry that you have missed the whole point of the Experiments.

You've missed the point of the comment, namely, that you need to have answers to those questions before you set up any experiments; if you don't, you will have no means of interpreting the results of those experiments (we covered this).

Yes, and I disagree with your premise that questions need to be answered before the Experiment. The Experiment itself will answer many questions.

No, it won't. It can't, because you have no means of linking the effects you observe with the (pseudo-)cause you postulate, or ruling out all of the other conceivable causes for the effect. Someone else could claim the effect you observe was caused by invisible leprechauns, or N-rays beamed from Betelgeuse. You need to understand and be able to characterize your postulated cause well enough, precisely enough, to be allow specific predictions of the effect from specific conditions in the cause. As I said, we've covered this before. Your explanation of the effect --- "some interference from something in 'conscious space'" --- would be no more informative than "goddidit." The effect would no more support your "conscious space" than it would those leprechauns.

There is nothing new about the assumption that consciousness is a manifestation of some sort of "alternative plane/dimension/realm" of "reality." It is as old as Plato, if not older; also called the "spirit realm," "eternal Forms," etc. That "paradigm" --- substance dualism --- was abandoned by most because claims about that supposed "realm" are unverifiable, unfalsifiable, and theoretically sterile --- they yield no verifiable predictions or any useful information.

But none of those things have actually been discovered and confirmed.

Neither has your "conscious space." Nor would the experiments you describe confirm it, for the reasons above.

Science cannot know the other things Consciousness might be doing because Science only measures Neural Activity and does not Measure the actual Conscious Experience.

That's correct, because "conscious experience" is not measurable. As I said, it is not available for empirical analysis. Neither would your experiments measure it.

Science would have to be able to independently Measure Conscious Experience in some way to come to the conclusions you have stated.

No, it doesn't need to measure it; it only needs to observe it, and consistently relate it to neural phenomena. You would have to do the same, and correlate it with events or changing variables in your "conscious space," which you could independently verify and measure.

True, so any conclusions you have made about Conscious Experiences are Speculations and nothing more. Your talk of certainties is founded on Ineffable and Unanalyzable phenomena.

Oh, certainly not. That conscious experiences are caused by physical processes is beyond dispute. All conscious experiences that we know of, at least. And, of course, we have no grounds for postulating conscious experiences anywhere other than those we ourselves experience --- which we know to be produced by neural processes --- and those we impute to other creatures who have nervous systems similar to our own --- on the basis of the behaviors they exhibit. If you wish to claim that conscious experiences may exist without a neural substrate, then the burden is upon you to produce an example.

"Consciousness" is not a "thing." Not a substance, not a plane or aspect or dimension of "reality." It is a property of physical systems of a certain type --- nothing more.