Refuting the notion that we have any evidence an android would be conscious.

[Interesting Ian]

It seems to me that everything that we perceptually experience always has a degree of subconscious interpretation associated with it. In other words we do not passively perceive, but rather our minds operate on the information received from sensors in order to interpret that information in a readily accessible form. For example, from a certain suitably arranged array of lines drawn on a sheet of paper, we subconsciously directly perceive it as being a cube, although strictly speaking the cube is inferred. But notice there does not appear to be any inference. It seems for all the world that we directly see a cube.

Likewise it might seem that when we observe people, we directly observe their inner mental states. Thus when you see someone laugh with joy, or scream in pain, it may seem that you do not infer this person's mental states from the outward signs, but rather you witness these emotions directly. But strictly speaking your expression of joy does not constitute the joy itself, rather it is the outward sign of a particular inner mental state of joyousness. Thus we are not in a position of certainty in ascribing mental states to others.

But what makes the inference that other people are conscious legitimate, is the premise that the experienced emotion at least plays some role in the causal loop that produces the outward words and gestures. If, instead, we have established that the observed words and gestures are wholly explained in some other way, without involving these emotions, then the inference collapses. The emotive behaviour could not then constitute any evidence towards an interior emotional state. We might point to the example of actors, who although may display emotions, are in reality just following a certain predefined routine.

Now if an android were ever to be created, then the totality of its behaviour is simply a consequence of millions of lines of programmed software. The software advances from instruction to instruction in a deterministic manner. Hence everything the android says and does is fully accounted for by software, and there's no explanatory gap left over for machine consciousness to fill. Hence if an android were to say for example "I am conscious and experience emotions just as you do", we know these words are produced by deterministic lines of software that functions perfectly well without any involvement of consciousness. It is clear then that whatever an android might say this constitutes no evidence whatsoever for any inner mental states. Note that this does not constitute a proof that androids are not conscious, but it does show that we have no evidence that they are conscious.

[Nyarlathotep]

Originally posted by Interesting Ian
Likewise it might seem that when we observe people, we directly observe their inner mental states. Thus when you see someone laugh with joy, or scream in pain, it may seem that you do not infer this person's mental states from the outward signs, but rather you witness these emotions directly. But strictly speaking your expression of joy does not constitute the joy itself, rather it is the outward sign of a particular inner mental state of joyousness. Thus we are not in a position of certainty in ascribing mental states to others.

It would seem that much of your argument hinges on this one paragraph and I am not sure what you mean by the distinction that you are drawing. What is the difference between infering someones mental state by their outward actions and experiencing it directly. To me that sounds like you are saying that if someone is happy or sad I feel that myself too, something that is obviously untrue (I can tell if someone is happy without actually feeling happy myself and I assume everyone else can too). I do not THINK that is what you mean but that is what it sounds like you mean.

[Wile E. Coyote]

What?

Do we have any evidence that androids actually exist? Oh, hypothetically speaking. I get it.

But to address your theory, I would have to say that a sufficiently complex android would indeed be conscious in the same way that humans are. Humans are also governed by a set a rules that have been programmed into us by genetics, society, and experience. We make all of our decisions based on what we calculate to be the most likely to produce a desired outcome. We do most of this subconsciously (like I think you said), but that does not mean the processes are not running.

And android would work in exactly the same way, except it would be made of different material and would probably make decisions based purely on logic or probabilty as opposed to emotion.

I think you have some kind of underlying motive to this question, but I am unable to get a hold of your writing style sometimes. Have I hit upon the gist of it?

[Interesting Ian]

Originally posted by Nyarlathotep

It would seem that much of your argument hinges on this one paragraph and I am not sure what you mean by the distinction that you are drawing.



{Shrugs}

I don't think I am able to express it in a more concise and clear way.




What is the difference between infering someones mental state by their outward actions and experiencing it directly.



You honestly don't know?


To me that sounds like you are saying that if someone is happy or sad I feel that myself too, something that is obviously untrue (I can tell if someone is happy without actually feeling happy myself and I assume everyone else can too).


Yes you can tell. But strictly speaking it is an inference. Yes? In other words their happiness manifests itself by a smile on their face and appropriate behaviour. We see the latter and infer the former. But the behaviour itself doesn't constitute happiness. Rather it is a manifestation of happiness.

[Interesting Ian]

Originally posted by Wile E. Coyote
[B]What?



. . I would have to say that a sufficiently complex android would indeed be conscious in the same way that humans are.



Then I suggest you believe so without any evidence or reasons.




Humans are also governed by a set a rules that have been programmed into us by genetics, society, and experience. We make all of our decisions based on what we calculate to be the most likely to produce a desired outcome. We do most of this subconsciously (like I think you said), but that does not mean the processes are not running.



The question of whether we whether we have evidence that other human beings are conscious can be addressed after. At this juncture I simply want people to concede that we have no evidence that an android would be conscious.



And android would work in exactly the same way, except it would be made of different material and would probably make decisions based purely on logic or probabilty as opposed to emotion.



The fact that an android may or may not operate in a similar way to what you propose human beings operate, gives no evidence that an android would be conscious.



I think you have some kind of underlying motive to this question, but I am unable to get a hold of your writing style sometimes. Have I hit upon the gist of it?

I don't know. As far as I'm concerned you have provided no reasons or evidence whatsoever that an android would be conscious.

[Agammamon]

II you are also implying that we aren't composed of millions of lines of code-that the difference between machine and organic intelligence is that for MI we know all of the code. For organic intelligence we don't know enough about the inner workings of the brain to tell for sure if we are coded or there is some sort of mystical adjunct that allows intelligence but only for meat machines.
On the other hand we do see from AI experiments that at the very least lower level behaviors (such as exhibited by insects) can be duplicated by MI. Even what are considered very complex behaviors such as pattern recognition, parsing, etc can be duplicated.
Couple that with no evidence of the extra entity that supposedly seperates us from the non-sentient and observation of animal behaviors that show that intelligence/sentience is not an all or nothing proposition (self-awareness and abstract thinking are properties that many animals show in varying degrees) and we can accept as a reasonable proposition that there is no inherent divide between OI and MI.
And for those who decrie a mechanistic explanation of the Human mind, saying that it eliminates free will, I respond with;
Free will exists or not independent of our belief in it. If free will doesn't exist it is better to know than to pretend otherwise.


"The question of whether we whether we have evidence that other human beings are conscious can be addressed after. At this juncture I simply want people to concede that we have no evidence that an android would be conscious."

As you state that outward behavior is not an indicator of conciousness (even if it mimics Human behavior) and I can't (regardless of your statement otherwise) directly detect your inner state (I can only see your behavior which you've already eliminated as a measure of conciousness) then we have no evidence that you are concious, or even that I am.

[Wile E. Coyote]

Originally posted by Interesting Ian

The question of whether we whether we have evidence that other human beings are conscious can be addressed after. At this juncture I simply want people to concede that we have no evidence that an android would be conscious.


We have no evidence that an android would be conscious. I concede that point. However, since androids are fictional at this time, all conjecture about their properties is just that ... conjecture.

We could similarly say that there is no evidence that visiting extraterrestrials would be conscious.

I think you need to define what you consider "conscious".

[Interesting Ian]

Originally posted by Agammamon
II you are also implying that we aren't composed of millions of lines of code-that the difference between machine and organic intelligence is that for MI we know all of the code. For organinc intelligence we don't know enough about the inner workings of the brain to tell for sure if we are coded or there is some sort of mystical adjunct that allows intelligence but only for meat machines.
On the other hand we do see from AI experiments that at the very least lower level behaviors (such as exhibited by insects) can be duplicated by MI. Even what are considered very complex behaviors such as pattern recognition, parsing, etc can be duplicated.
Couple that with no evidence of the extra entity that supposedly seperates us from the non-sentient and observation of animal behaviors that show that intelligence/sentience is not a all or nothing proposition (self-awareness and abstract thinking are properties that many animals show in varying degrees) and we can accept as a reasonable proposition that there is no inherent divide between OI and MI.
And for those who decrie a mechanistic explanation of the Human mind, saying that it eliminates free will, I respond with;
Free will exists or not independent of our belief in it. If free will doesn't exist it is better to know than to pretend otherwise. [/B]

All this is irrelevant. I ask you again.

Do you or do you not agree we have no evidence that an android would be conscious??

[Agammamon]

Hmm, I thought my post made both my position and the reasoning behind it clear.

Yes, I do agree that we have no evidence that an android would be concious.

And I also state that we have no more evidence that a person is concious than we have for an android. If you're not willing to accept one as concious then you can't accept the other as.

[Nyarlathotep]

Originally posted by Interesting Ian


Yes you can tell. But strictly speaking it is an inference. Yes? In other words their happiness manifests itself by a smile on their face and appropriate behaviour. We see the latter and infer the former. But the behaviour itself doesn't constitute happiness. Rather it is a manifestation of happiness.

Ah. I think I see what you are saying. Since the android is programmed and we can program it however we want even if it were conscious we could just as easily program it to frown when happy and smile when sad. As a result we cant really know if any apparent emotional state is the result of our programming or the result of true consciousness. Am I close?

If this is what you are saying I agree with you this far. Though I disagree that we can ever eally tell what someone elses mental state in any way beside inferring from their outside actions, because people can fake or hide emotions and if they are good enough you'd never know what hey are actually thinking or feeling. Instead I think people know other people to be conscious because they themselves are conscious so they just assum that other humans are too.

[Dragonrock]

I agree that we wouldn't be able to tell directly if the artificial intelligence were conscious, but we might be able to infer it. It's impossible right now to read a person's mind. But, if we built a mind then we would understand the language that mind speaks. We could record the information directly from the mind and read what the mind was thinking. If we saw the logical leaps, inferences, and other signs that are part of being conscious and self aware then we could assume that the mind is conscious.

As far as expressing vs feeling emotions, good actors are feeling the emotions they express. Method acting teaches them to get in character and live the life of who they are pretending to be. Some directors help the actors by telling jokes if the character should be happy, pissing them off if they should be mad, or doing whatever they must to make the actors feel what they need to feel.

[Agammamon]

There are two things that bother me about AI discussions in general, one is what exactly are we talking about when we say conciousness? Mostly when I use it I mean self-awareness/self control of behavior. The other is emotions. Emotions in animals are evolved responses and expressed in behaviors appropriate to the environment they evolved in, so why do we tend to assume that an AI will have the same emotions/emotional behaviors that we do, or that similar emotional responses won't be triggered at a vastly different threshold. Or even any emotion at all?
It may be hard to imagine since it is completely foreign to our experience (and emotions are very useful for us in directing action), but a fully self-aware and independent agent could conceivably exist without any emotional reponses at all. To be honest it sounds a bit like Nietzche's "New Man". Unhindered by hate/love, guilt whatever. It is difficult for me to picture how this entity determines it's own goals without the goad of emotion driving it on, but when it does have a goal it would move towards achieving it without most of the considerations that can make us so hesitant. As a practical matter, others would be considered only in so far as they can imped or help it.

[Interesting Ian]

Originally posted by Agammamon
Hmm, I thought my post made both my position and the reasoning behind it clear.

Yes, I do agree that we have no evidence that an android would be concious.



Good! That was all I was arguing for! :D

Now, 3 people disagree with this and do think there is evidence that an android would be conscious. Taking note of my objections could you 3 explain to me what this evidence consists in?

If you don't respond I might assume you're unable to justify your vote.

[Dancing David]

Well Ian I agree with the basic premise, if we assume that there is no objective way to test to see if a creature contains consiousness, then I agree there is no evidence that the android is consious.

But if the argument is that the android can't be consious becuase it's a causal system I have to disagree, respectfully.

You said this tread will not determine the test for being consious, so I say no more.

Peace

[Nyarlathotep]

Originally posted by Interesting Ian


Good! That was all I was arguing for! :D

Now, 3 people disagree with this and do think there is evidence that an android would be conscious. Taking note of my objections could you 3 explain to me what this evidence consists in?

If you don't respond I might assume you're unable to justify your vote.

Nope, I said I agree with you IF I understand you correctly.

[CWL]

Define "conscious".

[bjornart]

I define 'conscious' as fairly equivalent to 'self-aware'. The only one who can actually know if a being is self-aware or not is that self, you can't test it from the outside.
Your argument presumes something put together from deterministic components cannot be self-aware, which I disagree with, so no, you haven't proven anything.

[Mercutio]

Great question, Ian! I believe that, in your example, we would have no evidence at all that an android would be conscious. But that's just the easy part. I think (and I think you would agree with this, but am not sure) that by the same reasoning, we have no evidence that another person is conscious. You have said that could wait for another discussion, and I believe you, so I'll just let that sit for now. More importantly, I think your hypothetical situation is the best doorway I've seen yet to refute the notion that we have any evidence for consciousness in ourselves! I am totally serious here, but as with your caveat against discussing consciousness in others, I will avoid going into detail on this thread unless you (Ian) specifically ask for it.

[Interesting Ian]

Originally posted by Dancing David
Well Ian I agree with the basic premise, if we assume that there is no objective way to test to see if a creature contains consiousness, then I agree there is no evidence that the android is consious.

But if the argument is that the android can't be consious becuase it's a causal system I have to disagree, respectfully.


No the argument doesn't establish that an android cannot be conscious, anymore than it establishes a rock cannot be conscious. We just have absolutely no reasons to suppose it would be conscious, that's all.

[Interesting Ian]

Originally posted by bjornart
I define 'conscious' as fairly equivalent to 'self-aware'. The only one who can actually know if a being is self-aware or not is that self, you can't test it from the outside.
Your argument presumes something put together from deterministic components cannot be self-aware, which I disagree with, so no, you haven't proven anything.

No, my argument doesn't presume that. See my post above.

[hammegk]

Originally posted by CWL
Define "conscious".

I'd say that anything that is "alive" has some type of "consciousness".

I'd even be willing to state that "life" exists for any specific thing-in-itself that might -- or might not -- react to a specific stimulus.

Will a Turing Machine ever demonstrate "consciousness"; NO.

[Win]

Just to put my two cents in.

If materialism is true, then the answer is definitely no.

If materialism is false, then the answer is possibly no.

And what the android says has nothing to do with answering the question.

[Agammamon]

Even if you just consider consciousness as a property life has you still run into the problem of defining life. I am alive and your dog is alive, my desk and my grandmother are not. These are easy, but when you move closer to the boundary of life/not life it gets much fuzzier. Are virii alive? Most say yes. How about prions?
How do we tell if something is alive? Ability to reproduce? I haven't reproduced, am I alive? How about someone who is sterile? They can't reproduce. Or crystals which can but are not considered alive. Local reduction of entropy? Well my refrigerator can do that. As for stimulus reaction, we have machines that do that and they aren't considered alive.

A Turing Machine is by definintion indistinguishable from a Human Being. If a TM can't demonstrate consciousness then neither can a Human.

And until we have a solid definition of consciousness and devise a test for it, even you cannot know if you're conscious. You may only be programmed to think you are.

You know, if this thing has a built in spell-checker I would be greatful if someone could let me know how to find it.

[LW]

Originally posted by Agammamon
A Turing Machine is by definintion indistinguishable from a Human Being

I don't know what definitions you use for Turing Machines, but I use the following :

A Turing machine is a quadruple (K, A, T, s) where
- K is a finite set of states not containing the halt state h
- A is a finite alphabet containing the blank symbol # but not containing the symbols L and R
- s that belongs to K is the initial state
- T is a transition function T: K x A -> (K U { h }) x (A U { L, R }) (where U denotes the set union)

It may be just me, but I see quite little similarities with that definition and human beings.

OK, I lied here, I usually use a little more complicated definition but my browser crashed when I was writing it the first time and for the rewrite I decided to use the simplest Turing machine definition that I know.

[Michael Redman]

Originally posted by Interesting Ian
Now, 3 people disagree with this and do think there is evidence that an android would be conscious. The poll question does not ask if there is evidence that an android would be conscious. Your question is:

Have I demonstrated that we could have no evidence that an android is conscious?

The answers is “no”.

We do, of course, currently have no evidence of conscious androids, because there are no androids. What androids may end up being, if they do end up being, is likely not what we might now speculate, even if speculation were sufficient to claim something "proven". There is no way for us, sitting here today, to know if there will, in the future, be evidence of android consciousness.

It seems likely, for example, that an android brain could be built that did not rely on lines of human-produced code, but was able to grow, adapt, and learn, just as a flesh brain does. There is no reason to think that we can’t someday replicate with synthetics all the structures and processes of a living brain. Why couldn’t an artificial brain be built as complex as ours, with all the dynamic connections and processes, chemical influences, and other assorted mechanisms. It is, of course, ludicrous to assume that it could not be done. Would such a brain provide any evidence of consciousness? Presuming that you don’t invent a special definition for “evidence of consciousness” tailored to be unfulfilable*, there is no way for us to know.



*That is, as long as your definition of evidence for consciousness is something which humans can produce.

[Interesting Ian]

Originally posted by Agammamon


A Turing Machine is by definintion indistinguishable from a Human Being. If a TM can't demonstrate consciousness then neither can a Human.



No, that's not true. It's only indistinguishable in terms of its responses.



And until we have a solid definition of consciousness and devise a test for it,



We do not need a definition nor would it be possible to supply one. We not need a definition because we have unimpeded direct access to it. And clearly consciousness and conscious experiences cannot be ostensively defined but can only be known through the experience itself eg the raw experience of redness cannot be conveyed by a definition.



even you cannot know if you're conscious. You may only be programmed to think you are.


No, I could be programmed to say I am or to have the appropriate neural processes in the brain. But what's wrong with Descartes argument in this respect? To "think" something seems to me to implicity imply a conscious experience, and if it doesn't imply a conscious experience then it contradicts my immediate experience of "thinking".

[bjornart]

Originally posted by Interesting Ian


No, my argument doesn't presume that. See my post above.

My bad, however, what you do say is:

Hence if an android were to say for example "I am conscious and experience emotions just as you do", we know these words are produced by deterministic lines of software that functions perfectly well without any involvement of consciousness. It is clear then that whatever an android might say this constitutes no evidence whatsoever for any inner mental states.

So my objection remains. If consciousness is, as I define it, self-awareness it is clear that whatever a being might say this constitutes no evidence whatsoever for any inner mental states. You argument implies, rather than presumes, that a deterministic mind cannot be self-aware. Right conclusion, wrong argument, my vote remains.

[Interesting Ian]

Originally posted by Michael Redman
The poll question does not ask if there is evidence that an android would be conscious. Your question is:

Have I demonstrated that we could have no evidence that an android is conscious?

The answers is “no”.

We do, of course, currently have no evidence of conscious androids, because there are no androids. What androids may end up being, if they do end up being, is likely not what we might now speculate, even if speculation were sufficient to claim something "proven". There is no way for us, sitting here today, to know if there will, in the future, be evidence of android consciousness.



I want to know in principle how we could have evidence that an android is conscious if it just follows physical rules? What explanatory role could consciousness play in any android?



It seems likely, for example, that an android brain could be built that did not rely on lines of human-produced code, but was able to grow, adapt, and learn, just as a flesh brain does.



How does this imply consciousness in any shape or form?




There is no reason to think that we can’t someday replicate with synthetics all the structures and processes of a living brain. Why couldn’t an artificial brain be built as complex as ours, with all the dynamic connections and processes, chemical influences, and other assorted mechanisms.



But still, the processes in such a artificial brain would be governed by physical laws. How are we to infer any conscious experiences when they are not required to account for the behaviour of the android?

[Interesting Ian]

Originally posted by bjornart
If consciousness is, as I define it, self-awareness it is clear that whatever a being might say this constitutes no evidence whatsoever for any inner mental states.



If everything it says is a consequence of an algorithm or physical laws unfolding deterministically, then yes. Because then inner mental states are not required to account for its behaviour.




You argument implies, rather than presumes, that a deterministic mind cannot be self-aware. Right conclusion, wrong argument, my vote remains.

How does my argument imply this? I merely maintain we cannot have any evidence. Simply because we don't have any evidence for something doesn't mean to say it's not true!

[Michael Redman]

I am saying that we can potentially have exactly the same evidence for consciousness in androids that we have in other people.

You are assuming that the inner mental state is something other than the mechanical operation of the brain, operating under physical laws. I would agree that there is no way to have evidence that artificial minds have inner mental states apart from the operation of the physical brain. I would further agree that, if human minds did, in fact, possess a characteristic that was not simply due to the physical operation of their brains, that we could likely never have evidence than any artificial mind possesses the same.

But I don't believe there is evidence that humans have inner mental states apart from the operation of the physical brain, either. Because I believe that the mind is nothing more than the operation of the physical brain, I see no reason why androids can not have minds exactly like our own, mental state and all. So, what I consider consiousness to be, an android could potentially have, and provide evidence of.

[roger]

Originally posted by LW


I don't know what definitions you use for Turing Machines, but I use the following

I assume that he was confusing Turing test with Turing Machine,
as something that passes the Turing test would be "indistinguishable from a Human Being"

[CWL]

Originally posted by hammegk


I'd say that anything that is "alive" has some type of "consciousness".

Define "alive".

(Sorry Hammegk, but you knew that it was coming. :D)

[hammegk]

Originally posted by CWL


Define "alive".

(Sorry Hammegk, but you knew that it was coming. :D)

As previously provided, "I'd even be willing to state that 'life' exists for any specific thing-in-itself that might -- or might not -- react to a specific stimulus."

What do you bring to the table? ;)

[Underemployed]

Dear Ian,

I voted yes. Within the terms of your thought experiment, we can have no evidence of consciousness in the android, whether it has a pre-programmed mechanical brain or a biological one that has developed its own rules.

This raises the troublesome issue of determining if you or I 'have' consciousness, as we would be no different from the android.

But this is only because we have no satisfactory definition of consciousness. Saying it is 'self awareness' is just substituting one unknowable for another.

But then, if we had a good ol' fashioned empirical definition of consciousness, we'd also have a Nobel Prize to show at the next Amazing Meeting.

[Peskanov]

Ian;

----
quote:
No, that's not true. It's only indistinguishable in terms of its responses.
----

The proposition "a computer (turing machine type) can be indistinguishable from a human still in terms of its responses" is still a hypothesis; and I think we are a lot of years from seeing it confirmed.
Although I am happy to see you are getting a realistic position and admitting that this will be probably the case.

Another point: I have read two times your text, but I am not sure about the line of reasoning. Are you suggesting that empathy happens with a human, but would not with an android?

[Dancing David]

Originally posted by hammegk


As previously provided, "I'd even be willing to state that 'life' exists for any specific thing-in-itself that might -- or might not -- react to a specific stimulus."

What do you bring to the table? ;)

Cool quote hammegk , I will have to ponder it! So by this definition viruses are biological automata while bacteria are alive , cool. So it is the abilty to repond or not respond which makes life.

IAN you specifiaclly stated that we can not discuss the definition of or attempts to define consiousness. So those of us who voted no can't give our explanation. I think that there are ways to set deterministic rules into a 'fuzzy' logic much like our brain's use and there fore given enough tome there could be mecahnical consiousness, would it have a soul is another questio.(Would it wear mismatched clothes and watch TV is another) Being able to give eveidence, can't until they make the android.

Peace

[BobK]

Oh, sorry.

I thought this discussion was about asteroids.:D

[synaesthesia]

Ian,

Your 'argument' is based upon the presupposition that direct apprehension is an element of consciousness. You make no effort to reconcile the epistemological difficulties inherent in such a view, nor to rule out the operation of this mysterious force in an android.

[justsaygnosis]

Braniac V and DATA live......just ask Superman and Jean Luc Picard.

[Interesting Ian]

Originally posted by synaesthesia
Ian,

Your 'argument' is based upon the presupposition that direct apprehension is an element of consciousness. You make no effort to reconcile the epistemological difficulties inherent in such a view, nor to rule out the operation of this mysterious force in an android.

Direct apprehension of what?? I am quite unable to see my argument rests upon any presuppositions.

Are you saying then there is evidence that an android would be conscious? If so then what is the nature of this evidence?

[Interesting Ian]

Originally posted by Dancing David


IAN you specifiaclly stated that we can not discuss the definition of or attempts to define consiousness.



Where on earth did I mention this???

Why do materialists always accuse me of saying things and when I challenge them they can never substantiate their allegation??

{SIGHS}


So those of us who voted no can't give our explanation. I think that there are ways to set deterministic rules into a 'fuzzy' logic much like our brain's use and there fore given enough tome there could be mecahnical consiousness,



With what reason would you suppose there is consciousness if you cannot infer it from the androids behaviour?

[Zombified]

A couple of questions.

1. Do you believe there is evidence that humans have something androids cannot? If so, what?

2. Do you believe that there can be life and/or consciousness not based on organic chemistry? If so, can you distinguish between a conscious inorganic life form and an android? If not, what is special about organic chemistry?

3. What if the android was not programmed with millions of lines of code, but was rather given a simple learning program, and allowed to learn from its environment the way humans do?

4. What if the android's logic was nondeterministic, for example, a quantum computer instead of a classical one?

5. Going back to your previous thread, where we discussed the brain-in-a-jar scenario, are you now able to address the issue of when consciousness ends as more and more neurons are replaced by a computer simulation of each neuron?

6. Do you have an operational definition of "consciousness"?

Thanks.

[Jesse]

Terminator 2: Judgement Day

[CWL]

Originally posted by hammegk


As previously provided, "I'd even be willing to state that 'life' exists for any specific thing-in-itself that might -- or might not -- react to a specific stimulus."

What do you bring to the table? ;)

I read your suggestion the first time. ;)

Not sure what to say about that (or what to bring to the table) - although I do find your suggestion interesting. Could you perhaps elaborate on "react to a specific stimulus"?

---

[2 cents]

Anyway, here are my SEK 0.2 about the topic of this thread as such (for anyone who might be remotely interested).

As much as I normally like to disagree with Intersting Ian when he is refuting notions regarding the metaphysical, in this particluar instance he does have a point. The problem is of course that we can never conclusively establish whether anyone - or anything - besides ourselves actually is conscious. There simply is no way to tell. You know, the usual old P-zombie stuff...

Is it reasonable to accept the indications which may be offered (e.g. in the form of "reactions to stimuli" as suggested by Hammegk) as evidence of consciousness? Maybe so. If we do not, it appears we are approaching the dead end of solipsism. Even skepticism needs to be based on certain axioms, otherwise any attempt to obtain any form of knowledge becomes utterly pointless.

Notwithstanding, it does all boil down to defining "consciousness" I(very)HMO.

[/2 cents]

[Kevin_Lowe]

If we assume, as Ian does, that an android will be a Turing machine (in the proper sense, as opposed to a machine that can pass the Turing test) then there's good reason to think it will never be conscious.

Committing to any other position is fairly close to committing to calling a sufficiently sophisticated difference engine, abacus, or arrangement of pebbles in holes "conscious".

But this begs the real question:

Ian, why are you wasting time with this when there are so many unanswered posts for you in the "evidence for God" poll thread?

Do I need to start a "Refuting the notion that we have any evidence that Ian is not running scared" poll?

[Peskanov]

Kevin;

----
quote:
If we assume, as Ian does, that an android will be a Turing machine (in the proper sense, as opposed to a machine that can pass the Turing test) then there's good reason to think it will never be conscious.
----

And which one is it? In this board you will find arguments which range from old, tired ones ("computers can't have free will") to bizarre syllogisms involving quantum mechanics or gödel theorems. And all of them seem really weak until now.

----
quote:
Committing to any other position is fairly close to committing to calling a sufficiently sophisticated difference engine, abacus, or arrangement of pebbles in holes "conscious".
----

I don't see the relation. From the viewpoint of consciousness being a function of the brain, nobody can make that affirmation. It would be like saying that a potato can boot "windows xp" because it has enough computation components inside (it certainly has).

[ntech]

I'm still not sure that Interesting Ian is conscious.

[Interesting Ian]

Originally posted by Kevin_Lowe
If we assume, as Ian does, that an android will be a Turing machine (in the proper sense, as opposed to a machine that can pass the Turing test) then there's good reason to think it will never be conscious.

Committing to any other position is fairly close to committing to calling a sufficiently sophisticated difference engine, abacus, or arrangement of pebbles in holes "conscious".

But this begs the real question:

Ian, why are you wasting time with this when there are so many unanswered posts for you in the "evidence for God" poll thread?

Do I need to start a "Refuting the notion that we have any evidence that Ian is not running scared" poll?

{shrugs}

You can if you wish. I simply get bored with addressing the same subject all the time. You do not like this thread?

[Interesting Ian]

Originally posted by Peskanov


----
Kevin
If we assume, as Ian does, that an android will be a Turing machine (in the proper sense, as opposed to a machine that can pass the Turing test) then there's good reason to think it will never be conscious.
----
Pesk
And which one is it? In this board you will find arguments which range from old, tired ones ("computers can't have free will") to bizarre syllogisms involving quantum mechanics or gödel theorems. And all of them seem really weak until now.

Kevin
Committing to any other position is fairly close to committing to calling a sufficiently sophisticated difference engine, abacus, or arrangement of pebbles in holes "conscious".
----
Pesk
I don't see the relation. From the viewpoint of consciousness being a function of the brain, nobody can make that affirmation. It would be like saying that a potato can boot "windows xp" because it has enough computation components inside (it certainly has). [/B]

I agree with Kevin with the feelings that he expresses. On the other hand, I think it is difficult for the materialist to affirm that the brain is conscious but an android (as in a Turing machine) wouldn't be.

Any comments on this particular subject?

[Peskanov]

Ian,

----
quote:
On the other hand, I think it is difficult for the materialist to affirm that the brain is conscious but an android (as in a Turing machine) wouldn't be.
----

Most of your reasonings seems to deal with the views of eliminative materialism. The problem is that I don't share this view.
I prefer the aproach of neuroscientists, who use the classic concepts related to the mind (emotions for example) and redefine these to match their findings.
You can see that procedure as outrageous but this already happened in the past with physics and chemistry redefining tons of common concepts like force, cristal or lightning. I think this is the correct aproach fot materialism to deal with the mind: acknowledge the old, intuitive concepts, and match if possible.

----
quote:
If, instead, we have established that the observed words and gestures are wholly explained in some other way, without involving these emotions, then the inference collapses.
----

For me this phrase is the key point of your text.
From a materialist pov (not eliminative materialism) emotions, qualia, consciousness are reducible and material.
If the android is built emulating the components that allegedly created emotions first, then emotions must be infered from it.

[PixyMisa]

Originally posted by Interesting Ian
It seems to me that everything that we perceptually experience always has a degree of subconscious interpretation associated with it. In other words we do not passively perceive, but rather our minds operate on the information received from sensors in order to interpret that information in a readily accessible form. For example, from a certain suitably arranged array of lines drawn on a sheet of paper, we subconsciously directly perceive it as being a cube, although strictly speaking the cube is inferred. But notice there does not appear to be any inference. It seems for all the world that we directly see a cube. Baloney. It's obviously a drawing that represents a cube. We're just good enough at interpreting such things that we can say "it's a cube".Likewise it might seem that when we observe people, we directly observe their inner mental states.Baloney again. We are obviously observing the outward representation of their mental state. Everyone except small children can understand this.Thus when you see someone laugh with joy, or scream in pain, it may seem that you do not infer this person's mental states from the outward signs, but rather you witness these emotions directly. But strictly speaking your expression of joy does not constitute the joy itself, rather it is the outward sign of a particular inner mental state of joyousness. Thus we are not in a position of certainty in ascribing mental states to others.True. All we have is the evidence.But what makes the inference that other people are conscious legitimate, is the premise that the experienced emotion at least plays some role in the causal loop that produces the outward words and gestures.Baloney. What makes it legitimate to infer that people are conscious is that they act as if they were conscious.If, instead, we have established that the observed words and gestures are wholly explained in some other way, without involving these emotions, then the inference collapses.Baloney. We already know that the observed words and gestures can be entirely explained in terms of biochemistry or physics (depending on which level you prefer). Doesn't matter. People are still conscious.The emotive behaviour could not then constitute any evidence towards an interior emotional state.Baloney. If people consistently express emotions, the mechanism is irrelevant. They have emotions.We might point to the example of actors, who although ay display emotions, are in reality just following a certain predefined routine.Baloney. The actor has emotions, the character that the actor portrays has emotions. The two are not the same.Now if an android were ever to be created, then the totality of its behaviour is simply a consequence of millions of lines of programmed software.Baloney. If an android is ever created, it will have to be self-programming and self-learning. It will be able to modify its own mind, just as we do.The software advances from instruction to instruction in a deterministic manner.Baloney. First, computers are not purely deterministic, because they are built of atoms and atoms are not deterministic. Second, the system will need to be massively parallel, with lots of feedback mechanisms. The slightest difference in any part of the android brain will slowly ripple outwards to befome a tiny difference in the brain as a whole. Even if you wipe the brain clean and start again, you won't ever get the exact same run twice. You can't even do this with a large computer today.Hence everything the android says and does is fully accounted for by software, and there's no explanatory gap left over for machine consciousness to fill.Baloney. There's exactly the same explanatory gap as there is with the human brain. Do you have any concept of the sort of complexity you're talking about here?Hence if an android were to say for example "I am conscious and experience emotions just as you do", we know these words are produced by deterministic lines of software that functions perfectly well without any involvement of consciousness.Baloney. The evidence for consicousness is exactly the same as it is for a meat-human.It is clear then that whatever an android might say this constitutes no evidence whatsoever for any inner mental states.Baloney. A claim of a mental state represents an inner mental state. This is unavoidable. The inner mental state might be much simpler than that of a human, or might be equally complex. To say that it does not exist is false-to-fact.Note that this does not constitute a proof that androids are not conscious, but it does show that we have no evidence that they are conscious. Excuse me, Ian, I have this sandwich, and it's too big for me too eat the whole thing. Do you like baloney?

[Interesting Ian]

75% of people to date have voted that I haven't shown there is no evidence to suppose an android could be conscious. From the feedback I've been getting I would say this is a surprisingly high percentage. I'm just wondering if people are deliberately voting against me because they dislike me. Might be a good idea for me to create another sock puppet. Hmmmm . . .

Anyway, people can correct me, but I suspect that people here are thinking that our brains are bioelectrochemical machines, our minds are created by our brains, so therefore it must be the case that it would be possible in principle to create some sort of artifact or some sort of machine, which would actually be a mind.

Thus if we can emulate the overall function of the brain to get the same output, as might conceivably happen with a turing maching android, then necessarily the android would be conscious.

Are people basically in agreement with this? For the 75% of people who voted against me do you have any other reasons to suppose the android would be conscious?

These final questions are especially addressed to Kevin Lowe. Do you think it is impossible in principle to create a turing machine android which would exhibit indistinguishable behaviour from human beings? Or even if such an android could be created, would you still conclude that it would not be conscious? If so what is the crucial distinction whereby a bioelectrochemical machine is a mind, but the androids electronic brain wouldn't be?

[PixyMisa]

Originally posted by Kevin_Lowe
If we assume, as Ian does, that an android will be a Turing machine (in the proper sense, as opposed to a machine that can pass the Turing test) then there's good reason to think it will never be conscious.I find this a far more interesting question than Ian's, and I'd like to hear your reasoning.Committing to any other position is fairly close to committing to calling a sufficiently sophisticated difference engine, abacus, or arrangement of pebbles in holes "conscious".Yes, but that in itself is no evidence either way.

Of course, the Turing machine is a theoretical construct, and no perfect Turing machine can be built, so the question is not relevant to practical matters.But this begs the real question:

Ian, why are you wasting time with this when there are so many unanswered posts for you in the "evidence for God" poll thread?

Do I need to start a "Refuting the notion that we have any evidence that Ian is not running scared" poll? Ian appears to be incapable of constructing or following a logical argument. He appears to think that any statement that supports his position is logical, and any that disagrees is illogical. Because of this, he is unable to understand that the questions raised in the other thread are in fact relevant, and indeed highlight crucial flaws in his argument. That, as far as I can tell, is why he flies into fits of abuse: he cannot understand why people are disagreeing with him, when his argument is (to him) so obviously right, and others are so obviously wrong.

[PixyMisa]

Originally posted by Interesting Ian
75% of people to date have voted that I haven't shown there is no evidence to suppose an android could be conscious. From the feedback I've been getting I would say this is a surprisingly high percentage. I'm just wondering if people are deliberately voting against me because they dislike me. Might be a good idea for me to create another sock puppet. Hmmmm . . .Ian, people are voting against you because you are wrong.Anyway, people can correct me, but I suspect that people here are thinking that our brains are bioelectrochemical machines, our minds are created by our brains, so therefore it must be the case that it would be possible in principle to create some sort of artifact or some sort of machine, which would actually be a mind.Yup.Thus if we can emulate the overall function of the brain to get the same output, as might conceivably happen with a turing maching android, then necessarily the android would be conscious.That's not quite it. The key point is that we can only infer consciousness from observable behaviour, not that consciousness is required for such behaviour.Are people basically in agreement with this?Yup.For the 75% of people who voted against me do you have any other reasons to suppose the android would be conscious?Only that we are androids too.These final questions are especially addressed to Kevin Lowe. Do you think it is impossible in principle to create a turing machine android which would exhibit indistinguishable behaviour from human beings? Or even if such an android could be created, would you still conclude that it would not be conscious? If so what is the crucial distinction whereby a bioelectrochemical machine is a mind, but the androids electronic brain wouldn't be? As I noted, it's impossible to build a perfect Turing machine. However, it is possible to approach such a goal arbitrarily closely, and a mind running on such a machine would also exhibit all the observable characteristics of consciousness. If Kevin has an argument as to why a pure Turing machine could not be conscious, I'd be interested to hear it too.

[synaesthesia]

Originally posted by Interesting Ian (Emphasis added)
Thus when you see someone laugh with joy, or scream in pain, it may seem that you do not infer this person's mental states from the outward signs, but rather you witness these emotions directly. But strictly speaking your expression of joy does not constitute the joy itself, rather it is the outward sign of a particular inner mental state of joyousness. Thus we are not in a position of certainty in ascribing mental states to others.

You mention the fact that it is possible to mistake our knowledge of somebody else's emotion as being the result of direct, ineffable, infallible access. Intersubjective qualia as it were.

The most intuitive and generally true account is that there is space between what affects the behaviour (inside and outside the skin) of people. There is a process by which the information comes to us and is made availible. There is NO fundamental epistemic divide between the behaviour outside the human skin and what is inside of it.

We are not in a position of certainty for all the normal epistemological reasons, there is no reason to believe in an internal state independent of complete human behavior.

The revelation of modern neuroscience is that the same applies even when you examine humans. There are always processes, and processes within processes that give us knoweldge of ourselves. Breakdowns, electrical, chemical interference can change our relationship to ourselves just as it can with the world.

Now if an android were ever to be created, then the totality of its behaviour is simply a consequence of millions of lines of programmed software. The software advances from instruction to instruction in a deterministic manner. Hence everything the android says and does is fully accounted for by software, and there's no explanatory gap left over for machine consciousness to fill.

There are huge explanatory gaps in our understanding of the physical composition of the world, of it's structure, of the structure of the robots that we program. Nobody totally understands it.

That being said, there is indeed no need to invoke an undetectable - independent of the merely physical behavior - consciousness in the case of an anderoid. This is for precisely the same reason that there is no need for it in the case of man: the internal organs are not magically insulated from scientists' prying eyes.

Regards,
Synaesthesia

"We gladly accept a lucid, well-constructed statistical theory of the atomic nucleaus, if experimentation supports it. Becoming aquainted with such a theory, we do not ask, "Fine, but how are the atoms actually behavionr?" -because we know the foolishness of such a question. But similar revelation in the realm of anthropology we will fight with our last breath."
-Stanislaw Lem

[Interesting Ian]

Originally posted by synaesthesia


You mention the fact that it is possible to mistake our knowledge of somebody else's emotion as being the result of direct, ineffable, infallible access. Intersubjective qualia as it were.



What??? This "sentence" doesn't make sense. We are implicitly aware of certain phenomenological states, yes. But what has this got to do with judging whether an android is conscious?



The most intuitive and generally true account is that there is space between what affects the behaviour (inside and outside the skin) of people.


This "sentence" again doesn't make sense. If you mean that ones overt behaviour doesn't constitute the very emotion or whatever itself, then yes. I explained this in my opening post.


There is a process by which the information comes to us and is made availible. There is NO fundamental epistemic divide between the behaviour outside the human skin and what is inside of it.


You're advocating eliminative materialism? If so then address my original arguments against this position. If you're dissatisfied with those arguments then . . .{shrugs}. . there's nothing else I can say.



We are not in a position of certainty for all the normal epistemological reasons, there is no reason to believe in an internal state independent of complete human behavior.


And if so then an android would be conscious by definition.


The revelation of modern neuroscience is that the same applies even when you examine humans. There are always processes, and processes within processes that give us knoweldge of ourselves. Breakdowns, electrical, chemical interference can change our relationship to ourselves just as it can with the world.


But only objective knowledge. It is the subjective knowledge which is of all pervasive importance in understanding human beings.



There are huge explanatory gaps in our understanding of the physical composition of the world, of it's structure, of the structure of the robots that we program. Nobody totally understands it.

That being said, there is indeed no need to invoke an undetectable - independent of the merely physical behavior -



Apart that is from our direct experience of our own mentality ;)

[Peskanov]

Ian;

----
quote:
Are people basically in agreement with this? For the 75% of people who voted against me do you have any other reasons to suppose the android would be conscious?
----

I agree...mostly. I would suppose the android is conscious if:
1- The physical brain functions are succesfully formalised.
2- The artificial brain maps correclty these functions.
An output similar to the human one should be the result from these two points.
Point 1 is still far from done, so there is still time to find meaningful incomputable functions in the brain...

[hammegk]

Originally posted by PixyMisa
....However, it is possible to approach such a goal arbitrarily closely, and a mind running on such a machine would also exhibit all the observable characteristics of consciousness. If Kevin has an argument as to why a pure Turing machine could not be conscious, I'd be interested to hear it too.

The only observable characteristic of consciousness I have is that *I* think. You postulate that a very very fast Turing Machine will require an *I* algorithm; why would that be needed to pass a Turing Test?

[synaesthesia]

Originally posted by hammegk


The only observable characteristic of consciousness I have is that *I* think. You postulate that a very very fast Turing Machine will require an *I* algorithm; why would that be needed to pass a Turing Test?

Well, in some sense it's structure could inevitably be heuristically characterized in the same ways that human structure is. There's no need for an I algorithm is such a machine, just as there is no I algorithm in humans.

[hammegk]

Originally posted by synaesthesia


Well, in some sense it's structure could inevitably be heuristically characterized in the same ways that human structure is. There's no need for an I algorithm is such a machine, just as there is no I algorithm in humans.

I'd sure agree that my *I* that thinks isn't an algorithm. I'll even give you the benefit of the doubt & agree you are most likely not a p-zombie.

That is what you meant to say, right? ;)

[PixyMisa]

Originally posted by hammegk
The only observable characteristic of consciousness I have is that *I* think.If that's the only observable characteristic you will accept, then nothing else is conscious, and the question is answereed. Not answered usefully, mind you, but answered. You postulate that a very very fast Turing Machine will require an *I* algorithmNo, I don't. That's complete nonsense and bears no relation to what I said. Are you confusing the Turing machine with the (totally unrelated) Turing test?why would that be needed to pass a Turing Test?The "I algorithm", as you call it, is really just introspection; what is commonly referred to in programming as reflection. It is a necessary ability for any self-modifying system. There's no specific piece of code that handles the "I"; reflection is required by every part of the system so that it can learn and adapt. That's what gives rise to consciousness.

Now, it is clear that it is impossible to build a system capable of passing the Turing test using only hard-coded rules; the number of rules required would be so large that it would be physically impossible to implement. Does this answer the question?

[Walter Wayne]

www.dictionary.com
ev·i·dence n.
1. A thing or things helpful in forming a conclusion or judgment: The broken window was evidence that a burglary had taken place. Scientists weigh the evidence for and against a hypothesis.

2. Something indicative; an outward sign: evidence of grief on a mourner's face.
Not evidence is not proof. One piece of evidence need not be at all convincing, and it can very well lead to the wrong conclusion.

Since we don't no the nature of consciousness even in other people, an android acting in a "conscious" manner would be evidence however week.

I think at least 4 people who voted don't understand what evidence is, and probably more.

[Hand Bent Spoon]

I'm with Turing on this one.

Let the android take a Turing test, and it is conscious if it passes.

[synaesthesia]

Originally posted by hammegk
I'll even give you the benefit of the doubt & agree you are most likely not a p-zombie.

But it would not be ME who recieves the benefit of the doubt when we believe it likely that I am not a p-zombie. It would be the idea that we have access to the distinguishing features between conscious and unconscious. In other words, you have already rejected the presuppositions of the p-zombie thought experiment and argument.

[Kevin_Lowe]

Originally posted by Interesting Ian
These final questions are especially addressed to Kevin Lowe. Do you think it is impossible in principle to create a turing machine android which would exhibit indistinguishable behaviour from human beings? Or even if such an android could be created, would you still conclude that it would not be conscious? If so what is the crucial distinction whereby a bioelectrochemical machine is a mind, but the androids electronic brain wouldn't be?

In answer to the first question, it appears to me at the moment that we'll never be able to do this until we have a computer that can access relevant data as fast as a human can. We have only the vaguest idea how we are as good as we are at searching out relevant data.

This is an engineering problem and a psychology problem, and I'm sure it's not absolutely insoluble. Just incredibly difficult.

So my answer is that I'm not absolutely certain it cannot be done, but it's astoundingly difficult and we are not even close yet.

Okay, so suppose we build such an android anyway (to address the second question). I'm still unsure that it would be conscious.

The problem is, of course, that we only have the vaguest idea what consciousness is. We know it's a thing that arises from and depends on brains, or rather certain bits of brains, but that's about it. So I could be wrong.

But the problem is that a Turing machine doesn't understand anything. It just crunches numbers in an entirely deterministic way.

Searle's "Chinese Room" intuition pump is the classic philosophical argument against Turing machine consciousness. Searle proposed that you imagine yourself in a sealed room with two slots and some books that consist of rules for manipulating Chinese characters. Someone pushes a string of Chinese characters in one slot, you generate a new string of Chinese characters by applying the rules in the books to the original string, and you poke it out the other slot.

Now if we further suppose that by applying these rules you were actually passing a Turing Test in Chinese (successfully faking a human conversation), would you understand Chinese? The intuition most people get is that you aren't understanding Chinese, and that even if you got so good at applying the rules that you could do so without reference to the books you still wouldn't understand Chinese.

The argument goes that all a computer is doing is manipulating symbols in just this way, so it isn't doing whatever it is we humans do.

Objections to this argument are fairly widespread, and I won't try to present them all, but I find them a bit weak and ad hoc.

The second problem I have, which I alluded to earlier, is that anything a Turing machine can do you can also do with a toilet roll and a pile of pebbles.

You just unroll a suitable length of toilet roll, put rocks on sheets for 1s and leave them empty for 0s, and apply computational rules as you walk up and down the roll adjusting stones.

Now call me a bigot, but I just don't believe such a toilet roll arrangement can be conscious, even though it might be doing exactly the same processing tasks as our android (albeit at a vastly slower pace). I think that information processing has to be a vital part of the package, but I don't think that information processing as such is necessarily consciousness.

There's also the problem that, as I understand it, there are some things humans can do that Turing machines are just plain categorically incapable of. Like the Halting Problem: computers cannot tell if a program will ever end, for certain.

None of this means that something other than a Turing machine might not be conscious. In fact, it's reasonably likely that there are conscious aliens out there with brains kind of unlike ours. But if we limit ourselves to Turing machines, I don't think they'll ever understand anything or be conscious.

But as I said, I could be wrong.

Edited to add the word "not" in a vital spot

[PixyMisa]

Originally posted by Kevin_Lowe
This is an engineering problem and a psychology problem, and I'm sure it's not absolutely insoluble. Just incredibly difficult.Agreed.But the problem is that a Turing machine doesn't understand anything. It just crunches numbers in an entirely deterministic way.The brain doesn't understand anything either. It just sends electrical signals back and forth, influenced by the chemicals in the bloodstream.Searle's "Chinese Room" intuition pump is the classic philosophical argument against Turing machine consciousness. Searle proposed that you imagine yourself in a sealed room with two slots and some books that consist of rules for manipulating Chinese characters. Someone pushes a string of Chinese characters in one slot, you generate a new string of Chinese characters by applying the rules in the books to the original string, and you poke it out the other slot.

Now if we further suppose that by applying these rules you were actually passing a Turing Test in Chinese (successfully faking a human conversation), would you understand Chinese? The intuition most people get is that you aren't understanding Chinese, and that even if you got so good at applying the rules that you could do so without reference to the books you still wouldn't understand Chinese.Searle's Chinese Room is baloney. You don't understand Chinese, the books don't understand Chinese, but if the system responds appropriately to the questions it is obvious that the system does understand Chinese. Looking at the books or the human to try to find an understanding of Chinese is like looking for a biochemical theory of language while completely ignoring the electrical element of brain function. You're not only looking at the wrong level of function, you're excluding critical elements of the system.Objections to this argument are fairly widespread, and I won't try to present them all, but I find them a bit weak and ad hoc.Well, I don't know what arguments you've seen, but that's mine right there.The second problem I have, which I alluded to earlier, is that anything a Turing machine can do you can also do with a toilet roll and a pile of pebbles.Yes and no. A toilet roll and a pile of pebbles and a human can make a Turing machine. The rocks and roll by themselves do nothing.Now call me a bigot, but I just don't believe such a toilet roll arrangement can be conscious, even though it might be doing exactly the same processing tasks as our android (albeit at a vastly slower pace).OK, you're a bigot.I think that information processing has to be a vital part of the package, but I don't think that information processing as such is necessarily consciousness.Maybe not, but you need to define consciousness. Is a system that can learn and answer queries regarding its internal state conscious?There's also the problem that, as I understand it, there are some things humans can do that Turing machines are just plain categorically incapable of.Your understanding is flawed.Like the Halting Problem: computers cannot tell if a program will ever end, for certain.Humans cannot solve the halting problem either. The halting problem cannot be solved. (In the general case. Specific cases are often easily solved either by human or machine.)None of this means that something other than a Turing machine might not be conscious. In fact, it's reasonably likely that there are conscious aliens out there with brains kind of unlike ours. But if we limit ourselves to Turing machines, I don't think they'll ever understand anything or be conscious.

But as I said, I could be wrong.Well, as far as your argument here goes, you are wrong. It's really simple: brains are computers. Any computable function can be performed by a Turing machine. If brains can give rise to consciousness, so can a Turing machine.

The one caveat is that brains are physical things and the Turing machine is an ideal. Acausal quantum events can occur in the brain, but cannot occur in an ideal Turing machine. (They can, of course, occur in any physical implementation of a Turing machine.) So it may be that consciousness depends of quantum phenomena and a theoretical, perfect Turing machine cannot become conscious. I don't think this is so, and it is irrelevant to any discussion of real computers anyway since such are subject to quantum events.

[PixyMisa]

To throw another newt on the fire: a sufficiently powerful computer can simulate every atom in your brain and all of the interactions between them. The response of the simulation to stimuli would be exactly the same as your response. How is the simulated brain not conscious?

[Peskanov]

PixyMisa;

----
quote:
Now, it is clear that it is impossible to build a system capable of passing the Turing test using only hard-coded rules; the number of rules required would be so large that it would be physically impossible to implement. Does this answer the question?
----

You are wrong on this one. You seem to have drawn a too severe frontier between data and code.
I could explain myself more, but this would be too off-topic. Instead, I would like to point you to neural net algorithms. These does not modify it's own code, still, it's rules and behaviour change and evolve over time.

Kevin;
----
quote:
In answer to the first question, it appears to me at the moment that we'll never be able to do this until we have a computer that can access relevant data as fast as a human can. We have only the vaguest idea how we are as good as we are at searching out relevant data.
----

Accordingly to some calcutations, such computers already exist; what we lack right now is the knowledge about the human brain. We can't do the software.
About being close or not to emulating the brain, I am not so sure. I used to think like you, but the last news about the hippocampus prothesis surprised me greatly. I didn't know we were so advanced in brain mapping! Also, a project to emulate the full brain of a mouse has been started.

[ntech]

There is absolutely no reason that a computer brain could not eventually be built that is capable of learning and changing itself. They are working on a primitive project like that now. I see no reason why a computer mind could not eventually be as conscious as the rest of us fleshlings. It would just be a hell of a lot smarter.

[PixyMisa]

Originally posted by Peskanov
You are wrong on this one. You seem to have drawn a too severe frontier between data and code.
I could explain myself more, but this would be too off-topic. Instead, I would like to point you to neural net algorithms. These does not modify it's own code, still, it's rules and behaviour change and evolve over time.Actually, in a discussion of android consciousness, neural nets are bang on topic. But if you look at a neural net in terms of the operation of a classical computer architecture, then effectively it is changing its code. It's just that the code looks different. But yeah, the boundary between code and data can get fuzzy even in conventional systems. My statement holds if you do draw my rigid line, though.

[Peskanov]

Pixy;
Ok, I was misreading your reply. I mentally substituted "hard coded rules" for "code", hence the misunderstanding.
I agree, it's nearly impossible to to get human behaviour with a finite set of rules.

ntech:
----
quote:
There is absolutely no reason that a computer brain could not eventually be built that is capable of learning and changing itself. They are working on a primitive project like that now. I see no reason why a computer mind could not eventually be as conscious as the rest of us fleshlings. It would just be a hell of a lot smarter.
----

What frigthens me is no the possible smarteness of an AI. From the projects running now, the only ones showing advance are those based on brain emulation.
But we are emulating without understanding, so it will be difficult to amplify it's posibilities...I hope.
In the other hand there are few problems to speed up the emulation respect the original. In fact, the tech. perspectives show that speed up a AI thousands of times over a human brain will be possible.
Isn't it frightening?

[Kevin_Lowe]

Originally posted by PixyMisa Agreed.The brain doesn't understand anything either. It just sends electrical signals back and forth, influenced by the chemicals in the bloodstream.

The problem with that particular argument by analogy is that I am as sure that I understand English as I am sure of anything. But I am nearly equally sure that the brain is just sending electrical and chemical signals back and forth.

Anyone who feels as I do is pretty much forced into the position that my understanding must somehow be a characteristic of that brain business, in a murky way we do not fully understand.

But that doesn't force me to believe that I understand Hawking's physics, just because I could conceivably operate a toilet roll computer that crunched the numbers. I can imagine walking up and down that toilet roll, and the thing I call "understanding" is not taking place.


Searle's Chinese Room is baloney. You don't understand Chinese, the books don't understand Chinese, but if the system responds appropriately to the questions it is obvious that the system does understand Chinese.

That's one of the responses I consider weak and ad hoc.

I associate the word "understanding" with certain kind of subjective mental goings-on in my own mind.

Saying that "the system understands Chinese" sounds ridiculous to me. It would be like saying that "those holes in the ground, plus these pebbles, plus some rules for moving the pebbles around, plus a human, is in love with Sean Connery".

I just don't see how "understanding" is a claim you can meaningfully make for a human plus a pile of inanimate matter, if the human doesn't understand.

The system might act like it loves Sean Connery or understands Chinese. But that doesn't mean it actually does so, and we can actually verify this by taking the sucker apart and seeing how it ticks.


Looking at the books or the human to try to find an understanding of Chinese is like looking for a biochemical theory of language while completely ignoring the electrical element of brain function. You're not only looking at the wrong level of function, you're excluding critical elements of the system.Well, I don't know what arguments you've seen, but that's mine right there.

I guess it's an argument grounded in a couple of assumptions I take issue with. Firstly that a sufficiently good imitation of X is X, and secondly that you can ascribe subjective states of consciousness to arbitrary amounts and kinds of inanimate stuff on the basis that they form part of an information processing system.


OK, you're a bigot.Maybe not, but you need to define consciousness. Is a system that can learn and answer queries regarding its internal state conscious?

Well, no. Because computer programs a few lines long can do those things, and there's no sign they are anything other than unconscious.

I guess I define consciousness as that experience I have of perceiving the world, and the evidence available to me indicates that other humans have meaningfully similar experiences.


Your understanding is flawed.Humans cannot solve the halting problem either. The halting problem cannot be solved. (In the general case. Specific cases are often easily solved either by human or machine.)

Okay. My understanding was that computers could spot some halting situations by use of heuristics cooked up by humans but that was it. I though that the vast majority of halting questions were computationally insoluble but susceptible to human analysis. Where are I wrong?

Well, as far as your argument here goes, you are wrong. It's really simple: brains are computers. Any computable function can be performed by a Turing machine. If brains can give rise to consciousness, so can a Turing machine.

The one caveat is that brains are physical things and the Turing machine is an ideal. Acausal quantum events can occur in the brain, but cannot occur in an ideal Turing machine. (They can, of course, occur in any physical implementation of a Turing machine.) So it may be that consciousness depends of quantum phenomena and a theoretical, perfect Turing machine cannot become conscious. I don't think this is so, and it is irrelevant to any discussion of real computers anyway since such are subject to quantum events. [/B]

Hang on a sec.

Isn't it the case that Turing machines are discrete state machines, that click from state to state, whereas we humans are a constantly active kind of processing mechanism without discrete states?

Sure, we have on/off events like individual neurons firing, but that's not even close to being the whole story of human brain activity by itself.

So isn't that at least one fairly serious difference of kind, rather than of degree, between a biological brain and a Turing machine?

To be a bit mean, to my mind it's as ridiculous to attribute understanding to a moron and a toilet roll as it is for Escaping Ian to attribute understanding to spooky immaterial mind-stuff. Understanding is something that we only have reason to believe takes place in brains. I think claiming that anything else can understand stuff on the basis of an undersupported analogy like "brains are Turing machines" is silly.

Something mechanical other than a Turing machine might well be a candidate for consciousness, though.

[Michael Redman]

Originally posted by PixyMisa
To throw another newt on the fire: a sufficiently powerful computer can simulate every atom in your brain and all of the interactions between them. The response of the simulation to stimuli would be exactly the same as your response. How is the simulated brain not conscious? Well this is really the question, isn't it? The Immaterialist simply doesn't believe that consciousness is explained by the operation of the brain, so a simulated brain wouldn't be conscious, just because it recreates the workings of the brain. You and I think it would, because we think consciousness is explained by the working of our brains. There really isn’t any need to argue proof, if your assumption about the nature of consciousness is that it is supernatural, (or whatever the immaterialist thinks) and given only to people, through a process outside of the laws of physics. The android doesn’t have a change. Unless God want to have a little fun . . .

[Interesting Ian]

Originally posted by Hand Bent Spoon
I'm with Turing on this one.

Let the android take a Turing test, and it is conscious if it passes.

But I've explained why we couldn't have any evidence to suppose so! :mad:

[Upchurch]

Originally posted by Interesting Ian


But I've explained why we couldn't have any evidence to suppose so! :mad: LOL!

I saw that, Ian! Caught ya!

This exact post was posted by "The One called Neo" just a few minutes ago and I thought, "Neo? Since when was he in this discussion?" I went out and checked the stats and it was the only post by Neo in the whole thread. When I went back to read it again, it was gone only to be re-posted by you just now.

GOT to be careful when switching sock puppets like that, Ian. It can cause you problems. ;)

[blackpriester]

Originally posted by Michael Redman
I am saying that we can potentially have exactly the same evidence for consciousness in androids that we have in other people.

You are assuming that the inner mental state is something other than the mechanical operation of the brain, operating under physical laws. I would agree that there is no way to have evidence that artificial minds have inner mental states apart from the operation of the physical brain. I would further agree that, if human minds did, in fact, possess a characteristic that was not simply due to the physical operation of their brains, that we could likely never have evidence than any artificial mind possesses the same.

But I don't believe there is evidence that humans have inner mental states apart from the operation of the physical brain, either. Because I believe that the mind is nothing more than the operation of the physical brain, I see no reason why androids can not have minds exactly like our own, mental state and all. So, what I consider consiousness to be, an android could potentially have, and provide evidence of.

CLAP-CLAP...
Yet, I'm always so sad to see people posting things almost EXACLTY the way I would have put them... leaves me only the crumbs at this giant table of intellectual gorging ;).

[Interesting Ian]

Originally posted by PixyMisa
It's really simple: brains are computers. Any computable function can be performed by a Turing machine. If brains can give rise to consciousness, so can a Turing machine.



Hmmmm . . .I feel that you're correct here. On the other hand I agree with everything Kevin Lowe has said on this issue.

[Interesting Ian]

Originally posted by ntech
There is absolutely no reason that a computer brain could not eventually be built that is capable of learning and changing itself. They are working on a primitive project like that now. I see no reason why a computer mind could not eventually be as conscious as the rest of us fleshlings. It would just be a hell of a lot smarter.

Apart from the idea that the brain produces consciousness therefore why shouldn't a computer, do you have any independent reasons to suppose a computer would ever be conscious?

[PixyMisa]

Originally posted by Kevin_Lowe
The problem with that particular argument by analogy is that I am as sure that I understand English as I am sure of anything. But I am nearly equally sure that the brain is just sending electrical and chemical signals back and forth.Yes.Anyone who feels as I do is pretty much forced into the position that my understanding must somehow be a characteristic of that brain business, in a murky way we do not fully understand.Understanding is a property of the system as a whole. It's a process. You can't point at anything in the brain and say "there, that's understanding". It's the computation performed by the brain that produces understanding.But that doesn't force me to believe that I understand Hawking's physics, just because I could conceivably operate a toilet roll computer that crunched the numbers. I can imagine walking up and down that toilet roll, and the thing I call "understanding" is not taking place.I'm not sure how Hawking's physics come into it. But with the toilet roll example, you're missing the point: all you are doing is manipulating pebbles and toilet paper. You are not the one who understands. You are merely a mechanical device. It's the system that is giving rise to the understanding: you and the toilet paper and the pebbles. Understanding doesn't lie in any of them; it lies in the processing.That's one of the responses I consider weak and ad hoc.That's because you don't understand it, I'm afraide. I'll try again.I associate the word "understanding" with certain kind of subjective mental goings-on in my own mind.Fine.Saying that "the system understands Chinese" sounds ridiculous to me. It would be like saying that "those holes in the ground, plus these pebbles, plus some rules for moving the pebbles around, plus a human, is in love with Sean Connery".No. You're confusing the mechanism with the process. My brain cells don't think that Uma Thurman is cute. I think that Uma Thurman is cute.I just don't see how "understanding" is a claim you can meaningfully make for a human plus a pile of inanimate matter, if the human doesn't understand.All the human is here is a processor for manipulating data. The human is acting as a computer and executing a program which is giving rise to a new mind. That new mind is where the understanding lies. If you look at the human, you won't find the understanding, because you are looking in the wrong place.The system might act like it loves Sean Connery or understands Chinese.Yes.But that doesn't mean it actually does so, and we can actually verify this by taking the sucker apart and seeing how it ticks.That is exactly what it means. If we ask the system questions in Chinese and get back meaningful answers in Chinese, then the system understands Chinese. There is no way to avoid that fact.

If you take the system apart looking for the "understanding", you won't find it, because understanding is not a thing. It's like me taking your brain apart looking for the bit that understands English. I won't find it for two reasons: first, you don't have a little sealed-off understanding-English bit of the brain, and second, the brain is now dead and is not about to understand anything. If you take the books or the human out of Searle's room, you've killed it too. Except in that case you can put them back and it will come to life again.I guess it's an argument grounded in a couple of assumptions I take issue with. Firstly that a sufficiently good imitation of X is XIf you can't tell the difference, you can't claim a difference exists.and secondly that you can ascribe subjective states of consciousness to arbitrary amounts and kinds of inanimate stuff on the basis that they form part of an information processing system.The brain is an information processing system, remember? I never ascribed anything to inanimate stuff, I ascribed consciousness to the system. If you pull out any part of the system, and observe it when it is not functioning, it's not conscious.Well, no. Because computer programs a few lines long can do those things, and there's no sign they are anything other than unconscious.Except that they can learn and answer questions about their internal state.I guess I define consciousness as that experience I have of perceiving the world, and the evidence available to me indicates that other humans have meaningfully similar experiences.So you only accept consciousness at human level? Or are chimpanzees and dolphins conscious? Monkeys? Cats and dogs? Birds? Snakes? Fish? Lobsters? Worms?Okay. My understanding was that computers could spot some halting situations by use of heuristics cooked up by humans but that was it.Yes.I though that the vast majority of halting questions were computationally insoluble but susceptible to human analysis. Where are I wrong?The exact same analysis performed by humans can be performed by computers, in every case. The halting problem is unsolvable in the general case, whether it's a computer or a human trying to address it.Hang on a sec.

Isn't it the case that Turing machines are discrete state machines, that click from state to state, whereas we humans are a constantly active kind of processing mechanism without discrete states?Yeah. Doesn't matter. A Turing machine can execute a different function with every click, and time-slice its way to an identical process as a human brain.Sure, we have on/off events like individual neurons firing, but that's not even close to being the whole story of human brain activity by itself.No. So to emulate all of the on-off events and the interactions, you need lots of clicks. That's all.So isn't that at least one fairly serious difference of kind, rather than of degree, between a biological brain and a Turing machine?Nope. This has been mathematically proven: The Turing machine is a completely general computing engine. Anything that any computer can do can also be done by a Turing machine. It doesn't matter how strange the architecture of the computer is. It doesn't matter whether it's made of living cells or toilet paper and pebbles. It's still possible to map the function on to a Turing machine. Always. To be a bit mean, to my mind it's as ridiculous to attribute understanding to a moron and a toilet roll as it is for Escaping Ian to attribute understanding to spooky immaterial mind-stuff. Understanding is something that we only have reason to believe takes place in brains.But as I've pointed out, it's not the brain that understands, it's the mind.I think claiming that anything else can understand stuff on the basis of an undersupported analogy like "brains are Turing machines" is silly.It's not an unsupported analogy, it's an indisputable mathematical proof.Something mechanical other than a Turing machine might well be a candidate for consciousness, though.Any computer other than a Turing machine can be emulated by a Turing machine.

That's the whole point of the Turing machine: it's the simplest completely general computer. Note that the term "Turing complete" is used to refer to a system that can perform completely general computations. A non-programmable pocket calculator is not Turing complete. All general-purpose computers are, and they are precisely as powerful as each other in this regard. It doesn't matter at all whether you have a single 4-bit processor or a huge array of 64-bit superscalar monster chips, they have exactly the same possible range of computations.

So does the human brain.

[PixyMisa]

Kevin, have you read Hofstadter's Godel, Escher, Bach?

It's not perfect, but it's an entertaining read and it does cover a lot of stuff regarding Turing (both his machine and his test) and Godel's Incompleteness Theorem and how they relate to both human and machine minds.

[PixyMisa]

Originally posted by Interesting Ian
Hmmmm . . .I feel that you're correct here. On the other hand I agree with everything Kevin Lowe has said on this issue. Stop that, Ian! You're confusing me!

[Interesting Ian]

Originally posted by Kevin_Lowe
Saying that "the system understands Chinese" sounds ridiculous to me. It would be like saying that "those holes in the ground, plus these pebbles, plus some rules for moving the pebbles around, plus a human, is in love with Sean Connery".

I just don't see how "understanding" is a claim you can meaningfully make for a human plus a pile of inanimate matter, if the human doesn't understand.

The system might act like it loves Sean Connery or understands Chinese. But that doesn't mean it actually does so, and we can actually verify this by taking the sucker apart and seeing how it ticks.


I agree entirely with your sentiments here, but in principle we could do this with brains also! This is one of my major problems with the notion that the world is physically closed. And if the world is physically closed, what justification do we have to infer that other people are conscious?

[synaesthesia]

Originally posted by ntech
There is absolutely no reason that a computer brain could not eventually be built that is capable of learning and changing itself. They are working on a primitive project like that now. I see no reason why a computer mind could not eventually be as conscious as the rest of us fleshlings. It would just be a hell of a lot smarter.

No, I disagree with you there. If we made machines think, they would be as silly and prone to delusion as we are.

Once they start spouting off about Qualia and Spirit, even Interesting Ian will think they're conscious.

[synaesthesia]

Originally posted by Interesting Ian
But I've explained why we couldn't have any evidence to suppose so! :mad:

You haven't made any argument, every key issue was just assumed. Obviously you can't expect to convince anyone with that.

[synaesthesia]

Originally posted by PixyMisa
Kevin, have you read Hofstadter's Godel, Escher, Bach?


That's an awwwwsome book. P-Zombies the world over should read it.

[synaesthesia]

Originally posted by Interesting Ian


I agree entirely with your sentiments here, but in principle we could do this with brains also! This is one of my major problems with the notion that the world is physically closed. And if the world is physically closed, what justification do we have to infer that other people are conscious?

Same as we do now. Our evidence of consciousness is merely physical. :rolleyes:

[PixyMisa]

Originally posted by synaesthesia
That's an awwwwsome book. P-Zombies the world over should read it. My mother gave it to me for Christmas when I was 16. I blinked a couple of times and thanked her politely. Then I started to read it... Didn't even want to put it down for Christmas dinner :)

[Michael Redman]

Originally posted by blackpriester


CLAP-CLAP...
Yet, I'm always so sad to see people posting things almost EXACLTY the way I would have put them... leaves me only the crumbs at this giant table of intellectual gorging ;). I just beat you to the crumb.

[hammegk]

Originally posted by synaesthesia

...Our evidence of consciousness is merely physical. :rolleyes:

When will you publish the proof underlying that statement?

*Me* is perceived & perceives at least part of what-is as "physical". *I* is another matter -- one I actually have an incontrovertible (and conscious) data point for.

Regarding "the physical", at heart what you have is smoke & mirrors, and contingent on the claim "an objective physical world exists". What is a quark, and Field strengths that increase with distance? a string? the Higgs Field?

BTW, *I* also know that *I.Me* is NOT a p-zombie. I'm not so sure about you, but again give you the benefit of the doubt. ;)

[hgc]

Originally posted by hammegk


When will you publish the proof underlying that statement?

*Me* is perceived & perceives at least part of what-is as "physical". *I* is another matter -- one I actually have an incontrovertible (and conscious) data point for.

Regarding "the physical", at heart what you have is smoke & mirrors, and contingent on the claim "an objective physical world exists". What is a quark, and Field strengths that increase with distance? a string? the Higgs Field?

BTW, *I* also know that *I.Me* is NOT a p-zombie. I'm not so sure about you, but again give you the benefit of the doubt. ;)
Ah, here comes hammy, with his magical *I*. How about just a scintilla of evidence for that? And no, its self-evident nature will not suffice.

[hammegk]

Originally posted by hgc

Ah, here comes hammy, with his magical *I*. How about just a scintilla of evidence for that? And no, its self-evident nature will not suffice.

Oh, you believe you ARE a p-zombie. Kool. *I* think is my evidence, thanks.

No comment on the physicality of Higgs Field & other "physical" odds & ends?

[hgc]

Originally posted by hammegk


Oh, you believe you ARE a p-zombie. Kool. *I* think is my evidence, thanks.

No comment on the physicality of Higgs Field & other "physical" odds & ends?
Since a p-zombie is indistinguishable from a concious human anyway, I'll take no offense.

[hammegk]

Originally posted by hgc

Since a p-zombie is indistinguishable from a concious human anyway, I'll take no offense.

Only by accepting from your perceptions the truth of materialism's first axiom. I don't thanks to *I*.

You also must believe that a roll of toilet paper & some pebbles becomes capable of seeing the redness of red when the Turing Machine made of it is manipulated by a "higher being" -- a human "higher being" should do fine, ok?

[hgc]

Originally posted by hammegk


Only by accepting from your perceptions the truth of materialism's first axiom. I don't thanks to *I*.

You also must believe that a roll of toilet paper & some pebbles becomes capable of seeing the redness of red when the Turing Machine made of it is manipulated by a "higher being" -- a human "higher being" should do fine, ok?
Raising gibberish to a high art.

You can perfume the pig with all the Turning machines you want, but as long as *I* is its own proof, it's still magical thinking.

[PixyMisa]

Originally posted by hammegk
You also must believe that a roll of toilet paper & some pebbles becomes capable of seeing the redness of red when the Turing Machine made of it is manipulated by a "higher being" -- a human "higher being" should do fine, ok? No "higher being" necessary. You could use a clockwork monkey at the result would be the same.

[hammegk]

Originally posted by PixyMisa
No "higher being" necessary. You could use a clockwork monkey at the result would be the same.

LOL. Add all the Turing Machines you wish to -- some higher being operates all of them.


Originally posted by hgc
Raising gibberish to a high art.

Alright, you finally read what materialists actually say: magical indeed.

BTW, is the toilet paper & pebble machine "alive" when it sees the redness of red, or will it have to experience love first?

[PixyMisa]

Originally posted by hammegk
LOL. Add all the Turing Machines you wish to -- some higher being operates all of them.False.

You do not need a Turing Machine to operate a Turing Machine.

You really don't know what you're talking about, do you? This is mathematics. There's no room for differences of opinion. You're just plain wrong.

[hgc]

Originally posted by PixyMisa
False.

You do not need a Turing Machine to operate a Turing Machine.

You really don't know what you're talking about, do you? This is mathematics. There's no room for differences of opinion. You're just plain wrong.
hammy's *I* perceives its own mathematics.

[blackpriester]

Originally posted by synaesthesia


That's an awwwwsome book. P-Zombies the world over should read it.

I might make myself look silly: What is a P-Zombie?

[PixyMisa]

Originally posted by hgc
hammy's *I* perceives its own mathematics. That only allows for three possibilities, by Godel's incompleteness theorem:

1. I am right and he is wrong.
2. He does not possess a sufficiently powerful formal system to understand my argument.
3. He is not internally consistent.

(Actually, there are 7 possibilities in all, since any two or all three of these could also apply.)

[PixyMisa]

Originally posted by blackpriester
I might make myself look silly: What is a P-Zombie? Something that possesses all the observable characteristics of a conscious mind but is not actually conscious. A fictional concept even less useful than a spherical cow of uniform density.

[hammegk]

Originally posted by PixyMisa
False.

You do not need a Turing Machine to operate a Turing Machine.

You really don't know what you're talking about, do you? This is mathematics. There's no room for differences of opinion. You're just plain wrong.
You are right; you continue to demonstrate that you are a f*cking idiot in this venue.

BTW, your "clockwork monkey" is just another Turing Machine.

That only allows for three possibilities, by Godel's incompleteness theorem:

1. I am right and he is wrong.
2. He does not possess a sufficiently powerful formal system to understand my argument.
3. He is not internally consistent.

(Actually, there are 7 possibilities in all, since any two or all three of these could also apply.)

Feel free to continue to contend non-life can demonstrate consciousness, although I am beginning to think you are just a programmed machine.

[Interesting Ian]

Originally posted by Michael Redman
[B]Well this is really the question, isn't it? The Immaterialist simply doesn't believe that consciousness is explained by the operation of the brain, so a simulated brain wouldn't be conscious, just because it recreates the workings of the brain. You and I think it would, because we think consciousness is explained by the working of our brains.



No, consciousness is not explained. Even materialists have to concede consciousness is not explained. Maybe they would say that consciousness sometime in the future will be explained, but it isn't currently explained. We have simply no idea how brain stuff creates mind stuff.

Incidentally I have argued in my opening post of the "refuting the notion there is no evidence for a God" thread that consciousness cannot in principle be scientifically explained.




There really isn’t any need to argue proof, if your assumption about the nature of consciousness is that it is supernatural, (or whatever the immaterialist thinks)



Consciousness would only be "supernatural" if you believe in libertarian free will I think. But one can be a immaterialist and also a naturalist. To be a naturalist you simply have to believe that reality as a whole can be described by physical laws. It needn't imply materialism at all.

[PixyMisa]

Originally posted by hammegk
BTW, your "clockwork monkey" is just another Turing Machine.Wrong again. Do you know what a Turing Machine is or not?

[hgc]

Originally posted by PixyMisa
Wrong again. Do you know what a Turing Machine is or not?
LOL. Given enough monkeys with enough hammers, maybe they'll build one.

[PixyMisa]

Recent research indicates that they are more likely to hit each other with the hammers and urinate on the researchers. Hmm... the research report is here (http://www.cnn.com/2003/WORLD/europe/05/09/offbeat.britain.monkey.ap/).

[Interesting Ian]

Originally posted by Upchurch
LOL!

I saw that, Ian! Caught ya!

This exact post was posted by "The One called Neo" just a few minutes ago and I thought, "Neo? Since when was he in this discussion?" I went out and checked the stats and it was the only post by Neo in the whole thread. When I went back to read it again, it was gone only to be re-posted by you just now.

GOT to be careful when switching sock puppets like that, Ian. It can cause you problems. ;)

Yeah, I'll have to be careful when I use a sock puppet that no-one knows about :)

[hgc]

Originally posted by PixyMisa
Recent research indicates that they are more likely to hit each other with the hammers and urinate on the researchers. Hmm... the research report is here (http://www.cnn.com/2003/WORLD/europe/05/09/offbeat.britain.monkey.ap/).
Damn those p-monkeys! There they go acting like the real thing again.

[Zombified]

Originally posted by PixyMisa
This has been mathematically proven: The Turing machine is a completely general computing engine. Anything that any computer can do can also be done by a Turing machine. It doesn't matter how strange the architecture of the computer is. It doesn't matter whether it's made of living cells or toilet paper and pebbles. It's still possible to map the function on to a Turing machine. Always. Any computer other than a Turing machine can be emulated by a Turing machine.

That's the whole point of the Turing machine: it's the simplest completely general computer. Note that the term "Turing complete" is used to refer to a system that can perform completely general computations. A non-programmable pocket calculator is not Turing complete.To be pedantic, there are a couple of things Turing machines can't do that a neuron might be able to.

1. Turing machines have a finite number of states and a finite number of symbols. Turing machines cannot represent a continuum. However, a neuron might be an analog device.

2. Turing machines cannot make random transitions, and cannot model random processes.

These objections might not be especially serious. First, Turing machines can approximate analog-ness to any desired precision, so long as its finite. Networks of neurons happen to be rather resilient to noise, so its not clear that any error in the precision of a neural simulation would be a problem. It's not even clear that neurons really are truly analog: they deal in pulses, and those pulses are "represented" by individual atoms crossing synaptic boundaries.

The randomness issue can be similarly addressed; an approximation can be made, in fact a very good approximation if your random number generator has more states than the rest of your turing machine program. In addition, there is such a thing as a "probabalistic" turing machine, which would involve state transition chosen by random (the state rule might require you to roll a die when choosing your next state). (These are used to compare the strength of conventional computers to quantum computers).

So (possibly probabilist) Turing machines can almost certainly provide an excellent simulation of a neural network. (You don't necessarily need semantically interpretable rules; successfully simulating each neuron and their interconnections would give you the equivalent of a brain.)

If you consider those differences insurmountable, however, Turing machines are not your only option for building an artificial brain. Is there any reason to believe that a "brain" built out of (analog) transistors, resistors, and capacitors would be unconscious? Imagine building a single chip that simulates a single neuron. It's an analog device, and it may be subject to noise effects, so the objections to Turing machines do not apply. Now connect a few billion together in the same arrangement a human brain is wired. Why would we think that's not conscious?

There were a few related questions I posed to Ian earlier, though he apparently has not had the opportunity to address them. No doubt he will do so at his convenience, but to restate them so that I don't forget:

- Is there anything special about organic chemistry that makes it the only possible channel for immaterial consciousness? If so, why?

- If not, and we can imagine inorganic, conscious life, how would we distinguish between that creature and an artificial one?

- (This is new, but along the same lines) What if we took the stupidest flatworm we could find, extracted its DNA, and genetically engineered the h*ll out of it so that we could grow an artificially engineered nervous system similar in capability to a human? Would you attribute consciousness to it? Is it possible to engineer consciousness by any means, not necessarily with a computer?

- Going back to the brain-in-a-jar, if we replaced a single neuron with an artificial computer or electronic simulation, what changed? If we do two neurons, what changes? If we do N? At one point in the process of replacing organic neurons with simulated ones does the sum total brain lose its consciousness?

- How are immaterial minds affected by drugs or material damage?

[Interesting Ian]

Originally posted by synaesthesia


Same as we do now. Our evidence of consciousness is merely physical. :rolleyes:

So what is this evidence then? Our behaviour can be completely understood with reference to the physical laws of nature, so it is not clear how you can appeal to behaviour. So what else?

[PixyMisa]

Originally posted by Zombified
To be pedantic, there are a couple of things Turing machines can't do that a neuron might be able to.

1. Turing machines have a finite number of states and a finite number of symbols. Turing machines cannot represent a continuum. However, a neuron might be an analog device.Well, the Turing Machine as an ideal has an infinitely long tape. That's why it can compute any computable function.2. Turing machines cannot make random transitions, and cannot model random processes.Yes, and I pointed this out in my response to Kevin Lowe. However, any physical implementation of a Turing Machine will be subject to the laws of QM and will be able to do truly random things. On the other hand, any physical implementation of a Turing Machine will have a finite tape.

(Snipped stuff I fully agree with.)If you consider those differences insurmountable, however, Turing machines are not your only option for building an artificial brain. Is there any reason to believe that a "brain" built out of (analog) transistors, resistors, and capacitors would be unconscious? Imagine building a single chip that simulates a single neuron. It's an analog device, and it may be subject to noise effects, so the objections to Turing machines do not apply. Now connect a few billion together in the same arrangement a human brain is wired. Why would we think that's not conscious?Right. And from there we get the argument about replacing the brain with micro-electronic components one neuron at a time. When does consciousness go away, and why?Going back to the brain-in-a-jar, if we replaced a single neuron with an artificial computer or electronic simulation, what changed? If we do two neurons, what changes? If we do N? At one point in the process of replacing organic neurons with simulated ones does the sum total brain lose its consciousness?Yeah, that one :)How are immaterial minds affected by drugs or material damage?And why are they affected the way they are? It's not just the existence of the effects that indicates that mind is a brain function, but the nature of the effects. Same goes for Yahzi's Bat. If the mind is an immaterial thing and the brain just mediates the senses and control, why do I lose consciousness?

[Interesting Ian]

Originally posted by synaesthesia


Our evidence of consciousness is merely physical. :rolleyes:


Once they (ie androids) start spouting off about Qualia and Spirit, even Interesting Ian will think they're conscious.



There is the possibility that anomalous cognition might play a role in our certainty that other people are conscious. This is surely conceivable.

When gazing into an androids eyes we might feel that the android is strangely empty. That there is nothing inside there.

So it would be insufficient for the android to start spouting off about qualia and spirits. We would need to feel that it is conscious.

So your assertion that "even Interesting Ian will think they're conscious" might not necessarily be true. :p

[Interesting Ian]

Originally posted by PixyMisa
Originally posted by blackpriester
I might make myself look silly: What is a P-Zombie?
--------------------------------------------------------------------------------

Something that possesses all the observable characteristics of a conscious mind but is not actually conscious. A fictional concept even less useful than a spherical cow of uniform density


No. A p-zombie is something that possesses all the observable characteristics of a conscious person but is not actually conscious. A conscious mind does not have any observable characteristics.

And could you tell me why p-Zombies are logically impossible? I accept they are naturally impossible.

[Interesting Ian]

Originally posted by Zombified

If you consider those differences insurmountable, however, Turing machines are not your only option for building an artificial brain. Is there any reason to believe that a "brain" built out of (analog) transistors, resistors, and capacitors would be unconscious? Imagine building a single chip that simulates a single neuron. It's an analog device, and it may be subject to noise effects, so the objections to Turing machines do not apply. Now connect a few billion together in the same arrangement a human brain is wired. Why would we think that's not conscious?



No, the question is why should we think it is conscious?



There were a few related questions I posed to Ian earlier, though he apparently has not had the opportunity to address them. No doubt he will do so at his convenience, but to restate them so that I don't forget:


- Is there anything special about organic chemistry that makes it the only possible channel for immaterial consciousness? If so, why?



My feeling would be not unless someone can persaude me otherwise. If I were a materialist I think I would be a functionalist.



- If not, and we can imagine inorganic, conscious life, how would we distinguish between that creature and an artificial one?



I don't believe it would be possible to create conscious life. I'm not a materialist remember.



- (This is new, but along the same lines) What if we took the stupidest flatworm we could find, extracted its DNA, and genetically engineered the h*ll out of it so that we could grow an artificially engineered nervous system similar in capability to a human? Would you attribute consciousness to it? Is it possible to engineer consciousness by any means, not necessarily with a computer?



No



- Going back to the brain-in-a-jar, if we replaced a single neuron with an artificial computer or electronic simulation, what changed? If we do two neurons, what changes? If we do N? At one point in the process of replacing organic neurons with simulated ones does the sum total brain lose its consciousness?



It wouldn't provided the electronic simulation provided the exact same function. Whether that would be possible I don't know.



- How are immaterial minds affected by drugs or material damage?

The self operates through the brain resulting in ones mind. Thus changes in brain states will affect mind states.

[ntech]

Originally posted by Interesting Ian


No, the question is why should we think it is conscious?

Because, currently we are infants in technology and someday they will be able to emulate the human brain exactly. A brain that learns and includes memories of emotion etc will be possible.

Do you think our brains will remain a mystery forever? If we survive long enough as a race, we will learn all we need to know to achieve consciousness in an artificial brain.

I understand that this flies in the face of theism but hey!

[Underemployed]

Is it possible to engineer consciousness by any means, not necessarily with a computer?(Zombified)

------

No.(Ian)

I'd have to take issue with you here. Be it engineering a flatworm or assembling the toilet-roll-pebble array (which I would dearly love to see), it is no different to growing a baby inside a womb. The baby has (or perhaps, will have - another thread topic) consciousness, yet it was assembled unconsciously by biological processes.

Hence, though the idea is deeply unsettling, any system capable of mimicking brain processes must be conscious based on any understanding of the word. Proving it is something else altogether.

But the toilet-roll-pebble-array (hereinafter referred to as the TRPA) is not alive in any sense of the word either. Thus consciousness is independent of life!

Move along now, nothing to see here...

(Edited for speeling)

[hammegk]

Originally posted by PixyMisa
Wrong again. Do you know what a Turing Machine is or not?

Yes I agree you are wrong as usual. Yup, I know what a Turing Machine is.

Are you so dense you would like to unequivocally state that "I, PixyMisa, believe that a Turing Machine can reach a state of "consciousness", but could not emulate a clockwork monkey?"

That's the position you appear to be taking.



Originally posted by hgc
hammy's *I* perceives its own mathematics.

Do you suppose math (in a material sense) to be more, less, or equivalent to you as a p-zombie? How about to a living, conscious, human?

Originally posted by Zombified
To be pedantic, there are a couple of things Turing machines can't do that a neuron might be able to.

1. Turing machines have a finite number of states and a finite number of symbols. Turing machines cannot represent a continuum. However, a neuron might be an analog device.

Umm. And is what-is infinitely dividible? Let us know when you intend to publish your monograph to clear this up.


2. Turing machines cannot make random transitions, and cannot model random processes.
Now if we just knew what "random" actually means ... ;)

Are you writing a two-part monograph?

Continuing from elsewhere, could you actually prove a neutrino had "no, nil, zero" effect on the beer in the fridge?

Originally posted by Underemployed

..... Thus consciousness is independent of life!

Move along now, nothing to see here...

Actually there is something to see; another logical endpoint of materialism.

[Zombified]

Originally posted by Interesting Ian
No, the question is why should we think it is conscious?We can't prove consciousness of other people either, we infer it from behavior. (In a different context you made this very point.) If we can't distinguish a simulation from the real thing, is there really a difference?
I don't believe it would be possible to create conscious life. I'm not a materialist remember.That wasn't really the question. Imagine we build a rocket and go to another planet and discover that there are humanoids made out of metal and silicon walking around, with a culture and philosophy, the ability to create more of themselves, etc. How do we decide whether they are inorganic but conscious life or programmed robots?
It wouldn't provided the electronic simulation provided the exact same function. Whether that would be possible I don't know.Whoa. Let's make sure we both understand question and answer precisely, because your answer would seem to imply that the end result, an artificial simulation of billions of neurons, is conscious. Is that what you're saying? I did not think that would be your position, but I suppose its possible, since you aren't requiring an organic brain.

Given that, what's the difference between my brain-in-jar-uploaded-to-computer, and Pixy's built-from-scratch computer? I mean, in terms of what we can actually observe.

I do believe it is possible in principle to simulate a single neuron artificially; it's just a matter of puzzling out all the chemistry.The self operates through the brain resulting in ones mind. Thus changes in brain states will affect mind states. Are you saying there are three things, self, mind, and brain, where a self cannot have a mind without a brain to help it? Which ones die when the brain does? If memory is lost due to a brain/mind injury, does that imply that the "self" has no memory?


Edit to add: I forgot the engineered flatworm question. What's the difference between that and an organic brain? Of course, the answer is the same difference as between my uploaded brain and Pixy's computer program. Which is...?

[Zombified]

Originally posted by hammegk
Umm. And is what-is infinitely dividible? Let us know when you intend to publish your monograph to clear this up.The field of real numbers. The field of complex numbers. Euclidean space. Etc. This is already covered, there is no reason for me to publish anything. Actual space is another matter, but that is not relevant to the question of whether Turing machines can perform calculations involving continua.

Are you attempting to argue that Turing machines can in fact emulate minds, or are you just being contrary?
Now if we just knew what "random" actually means ... ;)Unpredictable in principle. As in outcome-of-measurement random.
Continuing from elsewhere, could you actually prove a neutrino had "no, nil, zero" effect on the beer in the fridge?What the h*ll are you talking about? You must be grovelling through some thread I've forgotten. The effect of a neutrino on beer in fridges is not zero, but it is extremely small. Very extremely. So much so as to be as good as zero to anyone but a particle astrophysicist. Regardless, it has nothing to do with the question of consciousness.

[EdipisReks]

for those of you interested in an AI contstuct that is thinks and is self improving, here is a good (though somewhat complicated article) about seed AI (http://singinst.org/GISAI/)

[Interesting Ian]

Originally posted by Zombified



--------------------------------------------------------------------------------
I don't believe it would be possible to create conscious life. I'm not a materialist remember.
--------------------------------------------------------------------------------

That wasn't really the question.



You introduced the idea of artificial creatures!



Imagine we build a rocket and go to another planet and discover that there are humanoids made out of metal and silicon walking around, with a culture and philosophy, the ability to create more of themselves, etc. How do we decide whether they are inorganic but conscious life or programmed robots?



For a person such as myself? I think it would be incredibly difficult to create such a programmed robot which would be convincing given enough time. Maybe impossible. But if it were possible to create such robots, we might be able to distinguish conscious life from robots by means of anomalous cognition.


--------------------------------------------------------------------------------
It wouldn't provided the electronic simulation provided the exact same function. Whether that would be possible I don't know.
--------------------------------------------------------------------------------

Whoa. Let's make sure we both understand question and answer precisely, because your answer would seem to imply that the end result, an artificial simulation of billions of neurons, is conscious.


Well, I suspect it might be impossible. But it is not clear to me that there is anything wrong in principle with the self operating through an artificial simulation of the brain, rather than the biological brain itself. Replace a TV sets internal components with other components performing the same function and you will still get the same TV programme. The TV programme still isn't created ex nihilo from those internal components though.



Given that, what's the difference between my brain-in-jar-uploaded-to-computer, and Pixy's built-from-scratch computer? I mean, in terms of what we can actually observe.


A conscious built from scratch computer would be creating a self ex nihilo.


--------------------------------------------------------------------------------
The self operates through the brain resulting in ones mind. Thus changes in brain states will affect mind states.
--------------------------------------------------------------------------------

Are you saying there are three things, self, mind, and brain, where a self cannot have a mind without a brain to help it?


The mind is a result of the brain acting as a kind of "filter" to the self.


Which ones die when the brain does?


The mind.


If memory is lost due to a brain/mind injury, does that imply that the "self" has no memory?



No, the disembodied self I would speculate has access to everything that has ever happened to it. It is the brain which impedes those memories.

[PixyMisa]

Originally posted by Interesting Ian
No. A p-zombie is something that possesses all the observable characteristics of a conscious person but is not actually conscious. A conscious mind does not have any observable characteristics.True that. I should have said "possesses all the characteristics of having a conscious mind."And could you tell me why p-Zombies are logically impossible? I accept they are naturally impossible. I'm not sure that I can refute p-Zombies on a purely logical basis, but I agree that they are physically impossible, the same way that a hard-coded AI is physically impossible. The universe isn't big enough to build all the rules required to run a p-Zombie.

[PixyMisa]

Originally posted by hammegk
Yup, I know what a Turing Machine is.Your prior statements fly in the face of this claim.Are you so dense you would like to unequivocally state that "I, PixyMisa, believe that a Turing Machine can reach a state of "consciousness", but could not emulate a clockwork monkey?"

That's the position you appear to be taking.You are clearly unable to read your own posts, much less mine. You stated unequivocally that a clockwork monkey is a Turing Machine.BTW, your "clockwork monkey" is just another Turing Machine.This is mathematically false. Simple as that.Do you suppose math (in a material sense) to be more, less, or equivalent to you as a p-zombie? How about to a living, conscious, human?What do you mean by "math in a material sense". Are you referring to computation? If so, then all general purpose computers can compute the same range of functions. p-Zombies, conscious humans and desktop computers are identical in this respect. Turing proved this.Umm. And is what-is infinitely dividible? Let us know when you intend to publish your monograph to clear this up.Space and time may be infinitely divisible, but when you get down to Planck scales, further subdivisions not only do not, but cannot matter.

[Zombified]

Originally posted by Interesting Ian
For a person such as myself? I think it would be incredibly difficult to create such a programmed robot which would be convincing given enough time. Maybe impossible.I think we are speaking about idealizations here. Obviously it is very difficult to build a Matrix, too.But if it were possible to create such robots, we might be able to distinguish conscious life from robots by means of anomalous cognition.I take it by "anamalous cognition" you are referring to what many people would call paranormal abilities such as ESP or telepathy?
A conscious built from scratch computer would be creating a self ex nihilo.Obviously. However, once I have created my neural simulation which "imports" a mind/self from a living brain, and Pixy creates his "ex nihilo" artificial brain, how would we distinguish them? Could you come along afterwards and decide which brain was which, assuming we ran both on equivalent hardware?

Would there be any difference, besides the possibility of what you're calling anamolous cognition?

If that is in fact the only detectable difference, in both the artificial brain and the is-this-alien-a-robot cases, does that imply that anamolous cognition must be the only convincing evidence for an immaterial self distinct from the mind-function of a brain however implemented?
No, the disembodied self I would speculate has access to everything that has ever happened to it. It is the brain which impedes those memories.What is the motivation for speculating that the self has memory?

[PixyMisa]

Originally posted by Interesting Ian
But if it were possible to create such robots, we might be able to distinguish conscious life from robots by means of anomalous cognition.Which, as we all know, does not happen.Well, I suspect it might be impossible. But it is not clear to me that there is anything wrong in principle with the self operating through an artificial simulation of the brain, rather than the biological brain itself. Replace a TV sets internal components with other components performing the same function and you will still get the same TV programme. The TV programme still isn't created ex nihilo from those internal components though.If you smash a TV, other TVs are still able to pick up the same program. Minds are invariably bound to specific brains.

If the brain is only a filter to the mind, why does the mind stop working when the brain is damaged? Why do you lose consciousness under the effects of Yahzi's Bat, rather than just losing motor control and sensory input? Why does the lack of oxygen do the same thing? Why do so many drugs change the way you think?

[synaesthesia]

Originally posted by Interesting Ian

So your assertion that "even Interesting Ian will think they're conscious" might not necessarily be true. :p

Well, when you empirically test your soul detecting ability, please be so kind as to remember me as you roll in your million. :p

[Victor Danilchenko]

Oy vey, Boring Ian just doesn't learn... Good thing I have that nitwit on ignore.

he is a textboof example of a fundamentalist as someone who can't change his mind and won't change the subject.

[The One called Neo]

I'll post as Neo since Victor has me on ignore :)

Originally posted by Victor Danilchenko
Oy vey, Boring Ian just [b]doesn't learn... Good thing I have that nitwit on ignore.



Indeed I don't learn from a numbskull such as yourself :rolleyes:



he is a textboof example of a fundamentalist as someone who can't change his mind and won't change the subject.

I'm a fundamentalist??? LMAO!!! I think you've seriously lost it.

[The One called Neo]

Oh BTW Victor. Tell me how an algorithmic process miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one! :rolleyes:

[Victor Danilchenko]

The One called Neo

Oh BTW Victor. Tell me how an algorithmic process miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one! :rolleyes:Ah, as stupid as ever.

Algorithmic processes don't necessitate conscious awareness. Minds, beuing a particular type of algorithmic process, may or may not necessitate awareness -- there is a good argument that can be made for mind being fundamentally limited in its ability to manipulate abstractions, if deprived of introspection. However, none of this has any bearing on the fact that the human minds, as a particular type of algorithmic process, are accompanied by self-awareness -- because human minds possess introspective capability. Empirical data suggests that introspective capability preceeds intelligent cognition, in that all species which are cognitively close to humans, relatively speaking (chimps, dolphins, gray parrots) are demonstrably possessed of a limited form of introspection.

And on that note, ian, let me put this incarnation of you on ignore as well. Debating with you is like talking decision theory with a well-aged bottle of paper glue.

[The One called Neo]

Originally posted by Victor Danilchenko
[b]The One called Neo

Oh BTW Victor. Tell me how an algorithmic process miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one!


Ah, as stupid as ever.



Apparently not as stupid as you who puts me on ignore but responds to a sock puppet whom you know is me.



Algorithmic processes don't necessitate conscious awareness.



Good! You're showing some sense at last! :eek:






Minds, beuing a particular type of algorithmic process,



What??? You've just directly contradicted yourself! Make your mind up! :mad:

Allow me to rephrase myself:

Tell me how some algorithmic processes miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one!



may or may not necessitate awareness



One wonders if a mind which is entirely absent of conscious awareness deserves to be labelled a mind.


-- there is a good argument that can be made for mind being fundamentally limited in its ability to manipulate abstractions, if deprived of introspection. However, none of this has any bearing on the fact that the human minds, as a particular type of algorithmic process,



Even if I were to grant algoritmic processes are necessary for the existence of minds, why should one suppose they are sufficient?



are accompanied by self-awareness -- because human minds possess introspective capability. Empirical data suggests that introspective capability preceeds intelligent cognition, in that all species which are cognitively close to humans, relatively speaking (chimps, dolphins, gray parrots) are demonstrably possessed of a limited form of introspection.



Interesting as all this might be, how this is supposed to show that certain algorithms neccessitate conscious awareness is quite beyond me!



And on that note, ian, let me put this incarnation of you on ignore as well. Debating with you is like talking decision theory with a well-aged bottle of paper glue.

Why the hell didn't you say at the beginning so I needn't have bothered responding to you!? :mad: You've said absolutely nothing whatsoever to suppose a robot would be conscious.

Come on Victor! Wheel out your robot and let's have a laugh! :p

Oh well, if you intend contributing in R&P I'll have to create another sock puppet so I can continue to keep asking all these questions that you keep dodging.

[PixyMisa]

Originally posted by The One called Neo
Good! You're showing some sense at last!He never made the claim in the first place.What??? You've just directly contradicted yourself! Make your mind up!He never made the claim in the first place.Allow me to rephrase myself:

Tell me how some algorithmic processes miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one!The word "ncessitates" is still out of place, but clearly what you are asking for is an Operational Theory of Consciousness. Guess what! He doesn't have one! I don't have one! You don't have one! Nobody has one!One wonders if a mind which is entirely absent of conscious awareness deserves to be labelled a mind.One does.Even if I were to grant algoritmic processes are necessary for the existence of minds, why should one suppose they are sufficient?Because the alternative is that consciousness is acausal. That's all that's left, I'm afraid.Interesting as all this might be, how this is supposed to show that certain algorithms neccessitate conscious awareness is quite beyond me!Well, no real surprise there.

What he is suggesting is that since brains give rise to minds, and brains are algorithmic, minds must also be algorithmic. However, the second premise (that brains are strictly algorithmic) has not been established beyond reasonable doubt, as far as I know.Oh well, if you intend contributing in R&P I'll have to create another sock puppet so I can continue to keep asking all these questions that you keep dodging. You're likely to get your account cancelled. Sock puppets are strictly taboo now. I guess existing ones are grandfathered in, like your Neo and APSP.

[Zombified]

Well, since Pixy bumped the thread anyway...

Actually, I believe we've arrived at the point where its agreed a computer or circuit can function as a brain as well as an organic one, though Ian still stipulates the origin of its programming must not be artificial, and that its "mind" is still seperate from its progamming.

I'm hoping, Ian, you can at least tell me if it is your opinion that "anamolous cognition" is the only way to experimentally distinguish a true mind from a mere mechanism. I fear the question may have been forgotten during Victor's drive-by.

[PixyMisa]

And does the complete lack of evidence for "anomalous cognition" mean that we cannot claim that consciousness exists at all? Is every failure at the million-dollar challenge a p-Zombie?

[synaesthesia]

Originally posted by The One called Neo

Tell me how some algorithmic processes miraculously and magically necessitates an accompanying conscious awareness. I'm dying to hear this one!



There is no such consciousness that 'accompanies' the brain, thus no such explanation is necessary.

[Interesting Ian]

Originally posted by synaesthesia


There is no such consciousness that 'accompanies' the brain, thus no such explanation is necessary.

Well yeah, if we're all p-zombies. The thing is, I know I'm not. A bit puzzling huh? But of course you shouldn't believe me. Not that you can really believe anything!

[Interesting Ian]

Originally posted by PixyMisa
And does the complete lack of evidence for "anomalous cognition" mean that we cannot claim that consciousness exists at all? Is every failure at the million-dollar challenge a p-Zombie?

References please.

[PixyMisa]

Originally posted by Interesting Ian
References please. Um, you want references to the complete lack of evidence for "anomolous cognition"?

Start with Flim-Flam. Proceed with The Demon-Haunted World.

Or show that I'm wrong! Provide some evidence! There's a cool million in it for you!

[Victor Danilchenko]

PixyMisa

What he is suggesting is that since brains give rise to minds, and brains are algorithmic, minds must also be algorithmic. However, the second premise (that brains are strictly algorithmic) has not been established beyond reasonable doubt, as far as I know.that's true; then again, all attempts to find non-computable physical processes which would allow for trans-Turing-machine brains, have repeatedly failed. the most anyone ever managed was vague hand-waving and claims that there could be such a process (e.g. Penrose's argument in "Emperor's New Mind"). At this point, the preponderance of evidence suggests a pretty solid conclusion -- brains are indeed a subset of Turing machines.

[Victor Danilchenko]

Zombified

I'm hoping, Ian, you can at least tell me if it is your opinion that "anamolous cognition" is the only way to experimentally distinguish a true mind from a mere mechanism.Well, before I put ian on ignore, that was pretty much the answer I squeezed out of him. He said that only anomalous cognition would allow one to distinguish a person from a perfect android; conversely, he kinda-sorta conceded that if androids turn out to be indistinguishable from humans, he would have been proven wrong about his whole idealistic belief system. I of course have monumetal doubts about him actually conceding in the face of such evidence rather than trying to do the wiggle-dance, but I guess he feels confident in stating such falsification criteria, given that a perfect android is not going to arrive at the scene any time soon.

[PixyMisa]

Originally posted by Victor Danilchenko
[that's true; then again, all attempts to find non-computable physical processes which would allow for trans-Turing-machine brains, have repeatedly failed. the most anyone ever managed was vague hand-waving and claims that there could be such a process (e.g. Penrose's argument in "Emperor's New Mind"). At this point, the preponderance of evidence suggests a pretty solid conclusion -- brains are indeed a subset of Turing machines. Agreed. I certainly found Penrose unconvincing. And even if consciousness does derive something from acausality, any physical computer can do the same, so the point is moot unless you are trying to produce a mathematical theory of consciousness based solely on classical computation.

[Zombified]

Originally posted by PixyMisa
Agreed. I certainly found Penrose unconvincing. And even if consciousness does derive something from acausality, any physical computer can do the same, so the point is moot unless you are trying to produce a mathematical theory of consciousness based solely on classical computation. That was the most disappointing thing about Penrose's argument: he's essentially attacking a straw man (the classical Turing machine).

The biggest objection I see to Turing machines, and it ain't that big, is the element of randomness. It's easy to imagine that neurons firing would be subject to some degree of noise, and that noise might matter. However, there's already a developing body of work about probabalistic Turing machines, of which standard Turing machines are subset (where every state transition has probability 1). P-Turing machines are used for comparison to quantum Turing machines where the state can be a superposition. The goal of the exercise is to prove or disprove QP > PP, that is, that there are more algorithms that run in polynomial time on a quantum computer than on a classical (but probabalistic) computer (QP >= PP >= P by construction). But its at least as likely that the noise simply doesn't matter, given that neural networks are so resilient to error.

Furthermore (against Penrose) there's no evidence that quantum coherence matters in the brain, or in the brain to a greater degree than any other biological process.

Penrose's argument also assumes an interpretation of quantum mechanics that privileges consciousness, but as usual, does not define it or its characteristics. I am very much not a fan of that kind of interpretation, though his ideas about there being some threshold of probability (which Penrose relates to the graviton) seems analogous to Omnes addressing the measurement problem in the Copenhagen interpretation with decoherence and a probability threshold based on operational measurements. Curiously enough, Cramer's transactional interpretation also uses a threshold of non-measurability to dodge to objections about causality in what is basically a hidden variables interpretation. Perhaps a theme is developing here. It's unfortunate Penrose cannot make his ideas quantitative for comparison.

In any case, I hope Ian can give me his answer himself, not that I don't trust Victor, but I wouldn't want to be accused of attributing anything to Ian by hearsay alone.

[Dancing David]

I really like the toilet papaer tube and rocks machine. How big would it have to be?

I see noticible lack of evidence here, the mechanist/materialist can say , "Well at this point we have a fair idea of how the brain works.", the immaterialst says, "You have to show me from the atom to the product each step involved."

There is no evidence that a machine which replicates a human brain would not be consious, I think we heard something along the lines of
"I would look into it's eyes and know it wasn't consious".

Hmm.

Thanks Ian your posts are better when you don't rant and go off the deep end.

[Interesting Ian]

Originally posted by Victor Danilchenko
-- brains are indeed a subset of Turing machines. [/B]

Quantum processes in the brain don't dictate otherwise?

Anyway, if we suppose this to be true how does this vindicate materialism?

[PixyMisa]

Originally posted by Dancing David
I really like the toilet papaer tube and rocks machine. How big would it have to be?You'd need about a million light-years of toilet paper, give or take a couple of orders of magnitude.

[Interesting Ian]

Originally posted by Dancing David
I really like the toilet papaer tube and rocks machine. How big would it have to be?

I see noticible lack of evidence here, the mechanist/materialist can say , "Well at this point we have a fair idea of how the brain works.", the immaterialst says, "You have to show me from the atom to the product each step involved."

There is no evidence that a machine which replicates a human brain would not be consious, I think we heard something along the lines of
"I would look into it's eyes and know it wasn't consious".

Hmm.

Thanks Ian your posts are better when you don't rant and go off the deep end.

Huh?? You like that response?? I would have thought it would have pis*ed materialists off and at least have got the statement as a nomination for BillHoyt's illogic prize!

[Dancing David]

Double I: I thought that that was all your evidence so far. Was there some I missed.

[hammegk]

Originally posted by Zombified

The biggest objection I see to Turing machines, and it ain't that big, is the element of randomness. It's easy to imagine that neurons firing would be subject to some degree of noise, and that noise might matter. ....

As you mentioned elsewhere, if Planck time & distance are the lower limit for meaningful divisibilty of space, then indeed everything must be a subset of Turing Machines.

Current math models say this is so; the real question is "do TLOP actually agree with the math?"

{Dancing David}, we don't want to see the android(brain-nervous system) constructed from atomic level, rather from the most basic "particles/thingies" we can identify, so quarks & electrons may be getting closer.

And you apparently disagree that if the construct was deemed "conscious" (and no one will ever know that answer for certain, as I don't know about you) did we "build" life & consciousness, or did "life" choose to inhabit our construct? Solve this one for "simplest life" first, unless you also believe consciousness is possible in non-life.

[aggle_rithm]

Originally posted by Interesting Ian
Now if an android were ever to be created, then the totality of its behaviour is simply a consequence of millions of lines of programmed software. The software advances from instruction to instruction in a deterministic manner. Hence everything the android says and does is fully accounted for by software, and there's no explanatory gap left over for machine consciousness to fill.

I think the reason so many people voted "no" is not that they believe there is evidence that androids could be conscious; it's just that they did not believe your arguments supported your premise.

The way you describe the hypothetical programming of an android makes it sound like accounting software. The logical programming used to create artificial intelligence is very different from the procedural method used to create most computer software. Although it may move from "instruction to instruction" on a very low level (if it is not a multi-processing program), the thought process for creating AI doesn't work that way at all.

The basic construct of AI, the "agent", is required to be "aware" of its environment and its internal state at any given time. I believe this qualifies the agent as being "conscious", although on a very primitive level.

[wraith]

If consciousness could be created through some miraculous feat of science, surely that this should be the knock out blow for deists alike and hail the triumph of materialism?

Zomby: - Going back to the brain-in-a-jar, if we replaced a single neuron with an artificial computer or electronic simulation, what changed? If we do two neurons, what changes? If we do N? At one point in the process of replacing organic neurons with simulated ones does the sum total brain lose its consciousness?

Ian: It wouldn't provided the electronic simulation provided the exact same function. Whether that would be possible I don't know.




…from a conversation between Pixy and Zomby

Zomby: If you consider those differences insurmountable, however, Turing machines are not your only option for building an artificial brain. Is there any reason to believe that a "brain" built out of (analog) transistors, resistors, and capacitors would be unconscious? Imagine building a single chip that simulates a single neuron. It's an analog device, and it may be subject to noise effects, so the objections to Turing machines do not apply. Now connect a few billion together in the same arrangement a human brain is wired. Why would we think that's not conscious?

Pixy: Right. And from there we get the argument about replacing the brain with micro-electronic components one neuron at a time. When does consciousness go away, and why?

Some interesting points to consider here.

Hey Ian, lets say that your brain sustained a considerable amount of damage that rendered you disabled and that science was at a stage whereby the damaged area of the brain could be fixed. If there was some sort of device that could become part of the brain that performed the exact same function of the damaged area of the brain, then wouldn’t you be in the same state of mind as you were before you sustained brain damage?

In principle, science could create a device that could perform the entire function of the brain.

This runs along the lines of your analogy regarding the TV….


Lets say that this device could perform every function as a “normal” brain. Through this device you could create an android. However, does this mean that the android is conscious?

If you had a device that could perform every function of your brain and you gave it a body, would that android be conscious or would it act if it were conscious? In other words, it would act exactly the same way as you would act, only the android doesn’t perceive its actions.

OR

The device can perform every function of the brain and that the android is consequently conscious.

OR

Do you think that there is an area in the brain that is the physical manifestation of your consciousness and it would be impossible to create a device that could perform the function of this area of the brain?
So it could be possible to create a device that could perform the functions of certain areas of the brain, it just couldn’t perform the whole function of the brain.

OR

Do you think that the entire brain is the physical manifestation of your consciousness?
So it would be impossible to create a device that could perform the function of ANY part of the brain.

[Interesting Ian]

Originally posted by aggle_rithm


I think the reason so many people voted "no" is not that they believe there is evidence that androids could be conscious; it's just that they did not believe your arguments supported your premise.

The way you describe the hypothetical programming of an android makes it sound like accounting software. The logical programming used to create artificial intelligence is very different from the procedural method used to create most computer software. Although it may move from "instruction to instruction" on a very low level (if it is not a multi-processing program), the thought process for creating AI doesn't work that way at all.

The basic construct of AI, the "agent", is required to be "aware" of its environment and its internal state at any given time. I believe this qualifies the agent as being "conscious", although on a very primitive level.

I'm sorry, I'm not clear what you're saying. You're saying an android's behaviour wouldn't be a consequence of step by step instructions? Why do you have conscious in quotes? Are you saying it isn't actually conscious? If you are saying it is conscious, then how does this consciousness miraculously arise? Or are you advocating philosophical behaviourism?? How is an android different from a calculater or accounting software??

[Dancing David]

Originally posted by hammegk




{Dancing David}, we don't want to see the android(brain-nervous system) constructed from atomic level, rather from the most basic "particles/thingies" we can identify, so quarks & electrons may be getting closer.

And you apparently disagree that if the construct was deemed "conscious" (and no one will ever know that answer for certain, as I don't know about you) did we "build" life & consciousness, or did "life" choose to inhabit our construct? Solve this one for "simplest life" first, unless you also believe consciousness is possible in non-life.

I know this is the irreducible argument, I can't prove that life did not choose to enter it and you can't prove that it did.

While the 'soul' part of me finds it useful to endow the world with spirit, I am an animist and a nihilsit. The rational part of me says nope.

Just like the soul part of me finds beauty in the notion of reincarnation, the rational part of me says nope.

I use the souls part of me to delve into the mythological part of the world, I use the rational part of me to delve into the rest of it. Different windows for different views.

As for the conciousness and life thing, I am fascinated by the idea that alien life could be built along non-carbon models and therefore there might be biological analogs to the android. can't prove it, mere speculation.

I think that life arose from the chaos, I myself do not believe in intelligent design or order to biological reality. Random accumulations of fortuitous combinations.

[aggle_rithm]

Originally posted by Interesting Ian

[quote]
I'm sorry, I'm not clear what you're saying. You're saying an android's behaviour wouldn't be a consequence of step by step instructions?


That's right. See below.

Why do you have conscious in quotes? Are you saying it isn't actually conscious?

An AI agent is not conscious at the same level you and I are. If you define consciousness as being aware of internal and external states, then there are many degrees of consciousness.

If you are saying it is conscious, then how does this consciousness miraculously arise? Or are you advocating philosophical behaviourism?? How is an android different from a calculater or accounting software??

Well, there are no androids, but if they were, they would probably be programmed with artificial intelligence. Most software is developed using procedural, or structural programming languages, while AI uses logical programming languages. The approach is very, very different. Programming with logical languages does not involve going from step to step, but building the program from a more holistic perspective. The implementation is also very different, because procedural programs tend to see things in terms of black or white, yes or no, while AI sees things in shades of gray. AI programs can learn from their environment because they begin with a set of random probabilities and they gradually work out the correct ones based on the input they're getting. They can also make decisions based on what they know. Again, this does not occur at the same level as it does for us (yet), but it is qualitatively similar by design.

When an AI program is run on a single-processor machine, then it does execute the instructions one after the other by necessity. (Although these instructions do not make sense in a narrow context.) With multiple processors, it can execute multiple instructions simultaneously. Our consciousness emerges from the fact that we have millions of "processors" acting together at any point in time.

(If anyone out there has more expertise in AI than I do, feel free to correct me on any of this.)

[PixyMisa]

Originally posted by Interesting Ian
How is an android different from a calculater or accounting software?? Complexity.

[hammegk]

Originally posted by PixyMisa
Complexity.

I keep thinking you have completely defined inanity, but you've exceeded even my expectations with that remark.

You really should go pound sand, or write some buggy spaghetti code, or do something you are actually capable of. :rolleyes:

[PixyMisa]

Originally posted by hammegk
I keep thinking you have completely defined inanity, but you've exceeded even my expectations with that remark.

You really should go pound sand, or write some buggy spaghetti code, or do something you are actually capable of. :rolleyes: Perhaps you'd care to point out how and why my answer is incorrect?

[hammegk]

Originally posted by PixyMisa
Perhaps you'd care to point out how and why my answer is incorrect?
Damn, your even-tempered anyway. :D

An inanity is not necessarily incorrect, just irrelevant to the point at hand. ;)

[DialecticMaterialist]

Most people should vote "No you haven't" when confronted with Ian's posts on principle. I mean, sure you could be wrong, but that's about as likely as your computer exploding. 99 times out of a hundred you'll be right. Those odds are better then Vegas.

At least that way you won't waste as much time.

[DanishDynamite]

(bump)