Quote:

So if that doesn't change by the time we have machines that can pass the Turing test, then you agree that we will categorize them as conscious, treat them as conscious, and for all intents and purposes they will be conscious.

I think we're in sync here. I can imagine a possible alternative is that if we find that there are people who treat them badly just because they are machines, as happened in the Spielberg AI, and we find that we can't stop people from doing so, we may have to outlaw the creation of such machines until we know whether they are really conscious or not.


Quote:

I will give you one very good reason. Consciousness is necessarily subjective which means it is not amenable to scientific inquiry. Have you read anything by David Chalmers on the subject? I would recommend Facing Up to the Problem of Consciousness.

I haven't read anything by Chalmers yet, and hope to find the time at some point, but in the quote you give below he is merely stating something very poetically that I have tried to state a number of times in this thread. It is the key point about the difference between consciousness and turing-test-passing behavior. When I talk to people about it, I often use a visual example, as he does, such as experiencing a fireplace (as opposed to "redness" in the text below, although I have used the color example as well). He's saying the exact same thing, but better, because he's a professional writer and I'm not.

Quote:

The really hard problem of consciousness is the problem of experience. When we think and perceive, there is a whir of information-processing, but there is also a subjective aspect. As Nagel (1974) has put it, there is something it is like to be a conscious organism. This subjective aspect is experience. When we see, for example, we experience visual sensations: the felt quality of redness, the experience of dark and light, the quality of depth in a visual field. Other experiences go along with perception in different modalities: the sound of a clarinet, the smell of mothballs. Then there are bodily sensations, from pains to orgasms; mental images that are conjured up internally; the felt quality of emotion, and the experience of a stream of conscious thought. What unites all of these states is that there is something it is like to be in them. All of them are states of experience.

I think you are dead wrong in the jump you are making. Just because consciousness is experience, and we know virtually nothing about it, and therefore, today, have no reliable way to determine whether a turning-test-passing-machine is conscious, does not mean that that will always be the case.

[the following is an edited version of my initial post, which I changed because I thought I could address it better].

Let's see if I understand your point of view.

* You seem to agree that consciousness and passing the Turing test are different matters. In other words, you seem to agree with me that the Turing test isn't valid as a determinant of what is and isn't conscious.

* You also seem to be saying that since consciousness is subjective experience, it will be forever out of reach of our understanding; therefore we will never have a better way than the Turing test for knowing whether something is conscious.

* You therefore draw the conclusion that if something passes the Turing test we have a moral imperative to treat it as conscious. You conclude that it doesn't matter, in a sense, whether it is or isn't conscious, since there is no way for us to know for sure. For all practical intents and purposes, the Turing test is remains the standard.

Let me know if I am misinterpreting you.

I think you are making the following error.

There was once a time when we could look up at the sky and, realizing that we didn't have the means to go up there or do anything more than gaze at it with our naked eyes, conclude that the nature of the motions of the objects we viewed would be forever unknowable. We could posit this or that about it, but we couldn't PROVE that the sun was not rolled along by a giant equivalent of a dung beetle (as the ancient egyptians thought), and therefore we should act as if it were. Even in principle, there was no way to prove it because there is no way to gather any information that would help us decide; it wascompletely  out of our reach. And we didn't want to anger the giant dung beetle-like thing, so therefore to hedge our bets we had to worship him, etc.


One error in that reasoning was that although according to all known principles and science of the time, there was absolutely no way of imagining getting a better understanding of the sky, it turned out not to be the case. It was a mere assumption, ultimatedly based on our old favorite "what can't be done now can never be done".

You say, similarly, that we can never know whether a turing-test-passing-object is really conscious, that that information is simply out of our reach due to its having to do with subjective experience, and therefore we should hedge our bets and assume that it is so that we don't cause suffering to a conscious entity.

But in reality, just because the sky was out of our reach many years ago, and an understanding of consciousness is out of our reach now, does not mean that we will never be able to understand, measure,  etc. It is a mere groundless assumption.

You might make the point, "well, even if we do think we have an understanding at this point, we can STILL never really KNOW, because it is subjective etc."

I saw a book recently, written in the last 50 years sometime, that attempted to reconcile our basioc knowledge of astronomy with a religious assumption that the heavens revolved around the Earth. It is possible to do. That is it is possible, if one desires, to construct a logical system that makes it possible that the heavens revolve around the earth, even knowing what we know.

But it just isn't REASONABLE, knowing what we know, to think that. (I hope you will agree.) While it MIGHT be the case, we still make the choice to risk people's lives and send them into space based on the assumption that we DO really understand what is happening, and what is happening is not that everything revolves around the earth.

Similarly, while we may never be able to absolutely prove that we know so much about consciousness that we can say with absolute certainty that a particular object that passes the Turing test isn't conscious, just as you, right now, don't know enough to be sure you are not a brain in a vat being stimulated to imagine you are sitting in a chair reading something on a computer screen, it is certainly conceivable that we may have enough understanding of consciousness that we can say that such a possibility is so unreasonable as to not deserve to be acted upon.

That is, you think it's completely impossible because subjective experience is not a class of thing we can know far more about than we do. I am saying that that's exactly logically equivalent to someone saying long ago that  the nature of the objects in the sky was not knowable because the sky is not a class of thing that we could know far more about than we did then. In one case, the supposed reason is that it is subjective experience. In the other, it's that it's too high. But the two assumptions are logically equivalent as far as I can see.

I doubt if you could make really concerete case to the contrary; I think any argument you make to the contrary would ultimately be mere intuition.

But I could be wrong -- feel free to try if you want.

OK, you say it isn't necessarily forever out of our understanding, for instance it may be an emergent phenomenon.

Suppose that we come to understand that it is an emergent phenomenon based on large numbers of information processing nodes operating in parallel within a volume of not more than 2 square feet.

From what you said in your reply, you seem to be admitting that the above is something that could happen.

Then, suppose we have something that acts like a human, but uses a single extremely fast processor. Then, by our understanding of consciousness, something that you state we may have at some point, the thing that acts like a human using a single processor wouldn't be conscious.

How do you resolve that?

"The two cases are not equivalent because you are equivocating physical impossibility with logical impossibility. If detecting consciousness was just physically impossible like travelling faster than light, of course I would be foolish to say that we will never be able to do it. However I say that is is logically impossible, so unless you can come up with an error in my logic I think it is perfectly reasonable to say that we will never be able to do it."

I understand what you are saying above and am glad I asked the question that prompted your statement above because it helps me understand your point better.

Answering VERY quickly, I don't agree that it is logically impossible to detect consciousness.

If  one were to assume, as you say you do, that we may understand consciousness in the future, we may understand it to have certain signs. For instance, a very compact body of a  lot of massive particles NECESSARIly has the associated "sign" of gravity. There may be associated signs of consciousness. There is no reason whatsoever to assume that there won't be, if one believes we may be able to understand consciousness.

But even without detection of signs, our understanding may enable us to say: we understand the conditions under which consciousness arises, such as the emergent phenomenon mentioned above; and something that doesn't have those conditions therefore isn't conscious, whether or not it passes the turning test.

Quote from: David Lucifer on 2003-01-25 10:16:05   

[ You have already stated quite clearly that any possible signs of consciousness don't tell you whether or not it is conscious.

I can't recall exactly what I said that you are interpreting that way, but I certainly do not believe what you are saying I have "stated quite clearly."

If I did say something to that effect, I meant it in the way that I also say I can't really know whether or not I am a brain in a vat being proded to think I am reading a message from you.

That is, my belief is that it is impossible to know anything of consequence as an absolute fact, but for practical purposes, that doesn't matter. What I am interested in is acquiring knowledge that reaches a practical point of certainty.


Quote:

Whether or not the entity in question displays the conditions under which consciousness arises is just more signs and behavior that you have already discarded as inadmissable evidence.

Hopefully we can take "Gary has discarded signs as inadmissable evidence," a thesis explicitly, directly, and obviously contradicted in the message you're responding to here,  and contradicted again in the present message, out of the picture.

Frankly, it's a waste of time to respond to a statement of X by saying "you have already admitted not-X" because obviously the person believes X if he is saying it.

So let's not waste our time, OK?



Quote:

The problem is you can't detect consciousness because the very word "detect" means to observe objective phenomena. You can certainly correlate objective phenomena with things you believe to be conscious (which is my position), but you can't use that to determine if it is "really" conscious (which is your position).

OK, correct me if I'm wrong, but you seem to admit above that there may be objective things that we are correlated with consciousness, other than the Turing test. You don't say so explicitly, but you seem to imply it.

I assume you will count certain brain activity picked up by various medical sensors that exist today as another such objective phenomenon.

If I am misunderstanding you in the above, let me know.

If I am am understanding you correctly in the above, let's go on to another step:

Do you admit that there MAY be objective phenomena that always go along with consciousness, in the same way that a curvature of space always goes along with the presence of a massive object?

Or, a lesser bar, that there may be some objective phenomena that always accompanies consciousness that we can measure to the degree of certainty that we can measure the presence or absence of radio waves?

Or do you assume such a thing is impossible?

Or I guess to put it in clearer terms:

Do you or do you not think it is POSSIBLE that we may one day understand the necessary (or necessary and sufficient) conditions for the arising of consciousness, in the sense that is usually meant by the word "understand"?

ISOLATED NEUROSCIENTIST in a black-and-white room knows everything about how the brain processes colors but does not know what it is like to see them. This scenario suggests that knowledge of the brain does not yield complete knowledge of conscious experience.

Quote from: David Lucifer on 2003-02-03 19:00:21   

Quote from: garyrob on 2003-02-03 10:09:52    I can't recall exactly what I said that you are interpreting that way, but I certainly do not believe what you are saying I have "stated quite clearly." You said it when you rejected the Turing test as a test for consciousness. Your reason was that it could pass the test and still not be conscious. I took that to mean that you don't count objective phenomena as a test for consciousness. Have you changed your mind?

That simply isn't logical.

To reject one objective phenomenon as a test while accepting the possibility of others is not equivalent to changing one's mind, or to asserting that both X and not-X are true.

If I must be more explicit, I reject the objective phenomenon of being  "white" as a test for whether an item is a snowflake. Such an item could be a piece of paper, or a sheep. That whiteness test for being a snowflake would, in my view, be about as valid as the Turing test is as a test for consciousness.

But there are other objective phenomena that I would accept as a test for being a snowflake, and I think it's quite possible that equally good tests may come to existence in the future that would be a test for consciousness. In this thread, I have repeatedly given examples of objective phenomena that I hold as valid tests of various qualities; the sole purpose of those examples was to say that by analogy, such valid tests might emerge for consciousness. But a belief that it is possible that such tests will one day emerge does not contradict the assertion that there is no such test today.

If I have to argue on such basic and obvious points, there isn't time left over to make arguments about things that are more difficult and are of actual interest.

With regard to the Chalmers extract, I remember well when that article appeared in Scientific American and I strongly applauded it, since I agree with everything it says, and I think that if you were making a real effort to understand what I've said you would know that. (I didn't, however, recall that the author was the Chalmers you mentioned earlier.)

Chalmers discusses Jackson's example: "But there is still something crucial about color vision that Man does not know: what it is like to experience a color such as red. It follows that there are facts about conscious experience that cannot be deduced from physical facts about the functioning of the brain." I fully agree with him that this cannot be arrived at through the kind of intellectual understanding he describes.

Being able to understand certain facts about consciousness such as the minimal requirements for its coming into being does not imply being able to fully grok redness if one has never experienced color. Those are different types of "facts". There is no reason why it wouldn't be possible to grok one and not the other.

Chalmers also says: " I am not denying that consciousness arises from the brain. We know, for example, that the subjective experience of vision is closely linked to processes in the visual cortex. It is the link itself that perplexes, however. Remarkably, subjective experience seems to emerge from a physical process. But we have no idea how or why this is. "

This is a major part of what I have been saying, except that I have been pointing out that just because TODAY we have no idea how or why this is, does not mean we NEVER will. Chalmers is not saying, in the extract you quote, that we never will understand the connection or in principle cannot; he is saying that we don't understand it, pure and simple.

Finally, in your last message you say "To summarize, if you now concede that objective tests are good enough to reasonably ascertain whether something is conscious, then your previous criticisms of the Turing test are no longer valid". I hope I don't need to restate the total illogic of that assertion, which I hope I have made clear above. But just in case I do, I will, for the last time. There MAY ONE DAY be objective tests that are good enough. That does not mean that every conceivable objective test is good enough; in particular it does not mean that the Turing test is good enough.  I don't think it is.

Why? Because I think that creating a non-conscious entity that passes the Turing test is quite possibly a pure engineering problem, and one that there is plenty of motivation to solve. Wouldn't it be great to fill Al Quaeda with totally expendable non-conscious entities that fool Bin Laden into thinking they are his followers but are in fact spies? The motivation is there, and it may be an engineering problem, albeit an extremely difficult one considering where we are at TODAY technologically. Where there's a motivation for solving an engineering problem, there is a good chance it will be solved. And if it is, the Turing test will be obviously not enough. Pure and simple. However, a test based on an understanding of the necessary and sufficient physical conditions for consciousness, which involves seeing if those condistions are met, would still be conceivable.

I'm going to have to bow out of this discussion. It's requiring me to spend too much time responding to "arguments" involving what I personally regard as too little rational substance, such as the assertion that "if you now concede that objective tests are good enough to reasonably ascertain whether something is conscious, then your previous criticisms of the Turing test are no longer valid".

Much of this discussion has been a pleasure for which I thank you, really. But I think it is losing its purpose. Let's move on.