I challenge you: your best alternate to materialism

[MRC_Hans]

Are you sure? If your calculations for leaping out of the car's way include, "I'd better leap to the right, because if the driver swerves to avoid me at the last moment he'll go farther left," and if you retain narrative memory of the event afterward, then consciousness might be a necessary result of having those capabilities.

Yes, but the train of thoughts you mention are not required to leap out of the way of an oncoming car. (Actually, I had not thought I should ever see you employ a strawman argument).

All it takes is the recognize the car as a possible threat, and flee. Even fairly unintelligent animals are able to do this.

I realize that these are not 'unconscious' as in a stupor, but they are, according to our present definition, not 'conscious' as in 'rational', which is what we are discussing here.

Or let me put it in another way: Being an electrical engineer and a computer programmer, I would have no problem building (with consumer available technology) a robot that could perform the feat.

Hans

 

[Nonpareil]

I realize that these are not 'unconscious' as in a stupor, but they are, according to our present definition, not 'conscious' as in 'rational', which is what we are discussing here.

Well, no, actually. It isn't.

We're talking "conscious" as in "self-aware". I have never heard "conscious" defined as "rational" anywhere, let alone in this thread.

Hope that clears it up.

 

[Myriad]

Yes, but the train of thoughts you mention are not required to leap out of the way of an oncoming car. (Actually, I had not thought I should ever see you employ a strawman argument).

All it takes is the recognize the car as a possible threat, and flee. Even fairly unintelligent animals are able to do this.

Well, yes, that's a simple algorithm that would require less computation. But it's not a very effective one, if the threat one is fleeing is an entity with its own volition, rather than just an object on a trajectory. If the threat is a faster predator, it will be spectacularly ineffective. If the threat is a species mate attempting to trick you into leaving your mates or young unguarded, ditto.

That seems to provide a plausible enough evolutionary drive toward decision-making methods that involve recognition of the self and others as volitional agents influencing one another's decisions. Self-awareness is probably a by-product of that useful capability, if it's not simply the capability itself.

 

[punshhh]

Why is such an experiment necessary in the first place?

To determine if none living AI are conscious, or simply automatons.

Well in this discussion, it started with a reference to a soul, or monad, which presumably would be present in an idealist universe. As opposed to the materialist view that there is no such thing.

 

[punshhh]

Assuming such an apparatus successfully mimicks a conscious being, how would you be able to know it is not consciuos?

I don't know, this problem works both ways, how would you be able to know it is conscious?

Do you have any reason to assume that it is not in fact conscious?

Only that consciousness might be a characteristic of living organisms, even the ones without brains.

In other words, do you know of ANY evidence for consciousness, other than conscious behavior?

Hans

Well for me conscious behaviour and living behaviour, to a large extent, have the same meaning. An amoeba, or a sponge follows the behaviour expected of a living organism. It is also the behaviour expected of a conscious mechanism.

 

[MRC_Hans]

Well, no, actually. It isn't.

We're talking "conscious" as in "self-aware". I have never heard "conscious" defined as "rational" anywhere, let alone in this thread.

Hope that clears it up.

OK, no problem. - Although 'self aware' is every bit as tricky as 'rational' or 'conscious' when it comes to defining it.

Hans

 

[MRC_Hans]

Well, yes, that's a simple algorithm that would require less computation. But it's not a very effective one, if the threat one is fleeing is an entity with its own volition, rather than just an object on a trajectory. If the threat is a faster predator, it will be spectacularly ineffective. If the threat is a species mate attempting to trick you into leaving your mates or young unguarded, ditto.

Well, even if it means you are still killed most of the time, it is obviously vastly more effective than just standing still.

That seems to provide a plausible enough evolutionary drive toward decision-making methods that involve recognition of the self and others as volitional agents influencing one another's decisions. Self-awareness is probably a by-product of that useful capability, if it's not simply the capability itself.

Agreed.

Hans

 

[MRC_Hans]

I don't know, this problem works both ways, how would you be able to know it is conscious?

(The apparatus mimicking a conscious being) - In the same way as I know (or assume) that thne being is conscious.

Only that consciousness might be a characteristic of living organisms, even the ones without brains.

Have you any examples of living organisms without a brain exhibiting anything that can be interpreted as consciousness?

Well for me conscious behaviour and living behaviour, to a large extent, have the same meaning. An amoeba, or a sponge follows the behaviour expected of a living organism. It is also the behaviour expected of a conscious mechanism.

Not only would that require a very unusual definition of 'conscious', but given such a definition, 'conscious' behaviour could easily be mimicked by an automation. In fact, I would claim that e.g. a robotic vacuum cleaner could then easily qualify as 'conscious'.

Hans

 

[Nonpareil]

I don't know, this problem works both ways, how would you be able to know it is conscious?

It has all the characteristics of a conscious entity. Therefore, it is a conscious entity. That is what we mean when we say that something is something.

Only that consciousness might be a characteristic of living organisms, even the ones without brains.

In which case your argument is not that a perfect mimic of consciousness would not be conscious, but that constructing a non-organic perfect mimic of consciousness is impossible.

For which there is no evidence.

 

[Belz...]

It has all the characteristics of a conscious entity. Therefore, it is a conscious entity. That is what we mean when we say that something is something.

I think the problem here is that people don't know what "to be" means. Naturally they then attach all sorts of "profound" meanings to it, all sefl-serving.

 

[Myriad]

Well, even if it means you are still killed most of the time, it is obviously vastly more effective than just standing still.

Not necessarily. In the case of the faster predator, standing still could be an equal or better strategy than running. If the organism can't comprehend or answer the question "can/does it see me?" then it's a crap shoot either way. In the case of the deceptive rival, standing still is definitely the better response.

A system that has the same response to any perceived threat can still be useful, but will have limitations. A good example is a burglar alarm. Creatures that use such strategies generally have backup strategies for species survival, such as reproducing in large numbers.

 

[punshhh]

Nit-pick: also animals.

Yes, perhaps a few mammals are self-conscious and even fewer perhaps a handful of primates actually think, "I am what I am".

Everything up to the last clause is correct.

Excuse me, but I am not aware that the experiment has been successfully carried out.

We know the brain is a computer. We know the brain is the source of consciousness. We know that altering the computation taking place in the brain alters consciousness, and we know that analyzing the computation taking place in the brain can show us details of the resulting consciousness' current state.

We know that the brain is the source of consciousness. This is not speculation.

We know that the brain is a computer and is Turing-equivalent. This is not speculation.

We know that what one Turing-equivalent computer does can necessarily be modeled by another Turing-equivalent computer. This is not speculation.

Therefore, we know that everything that the brain does can be modeled by any given Turing-equivalent computer. This is not speculation.

Everything that we know about consciousness indicates that it is computational. There is no indication, or even coherent hypothesis, of any non-computational factors being involved.

Each of those inferences is an assumption. Of course, it may be the case.

The only way to show that consciousness cannot be created on a computer, then, is to show that there is a non-computational factor.

Just screen out each of the characteristics of a human which are different to the computer, one by one, including its physical characteristics (as we are discussing materialism/physicalism).

Merely pointing out that it is an organic computer does not do this. The material that the computer is constructed of does not alter its capabilities.

As the precise origin, or process, of the phenomena of consciousness in the brain is not known at this stage. One should consider all characteristics of the mechanism which may be involved in producing the computation.

Perhaps an experiment should be conducted to demonstrate self-consciousness in a recently dead brain, by stimulation with electrodes.

If it has no self-awareness, there is necessarily a situation in which it behaves as though it does not.

The whole Pzombies are incoherent argument relies on this assumption, as I pointed out when I discussed it with PixyMisa, a while ago.

I repeat, if self-consciousness where simply computation, a mimic or Pzombie could theoretically perform any action of a self-conscious person, while entirely without self-consciousness.

If there is no such situation, is has self-awareness.

Assumption, unless it has the same physical state as the test example of a human.

Consciousness without a body - that is, without a computer to compute it - is nonsensical.

Agreed.

Let us set aside, for the sake of argument, that the lyrebird's mimicry is not necessarily perfect. If it did produce perfect copies of those sounds, yes, they would be the same sounds as the ones produced by the chainsaw.

So what?

The sound waves as they enter the microphone are indistinguishable from the real chainsaw. As it is the behaviour of the lyrebirds vocal chords which result in the same sound, it has successfully mimicked the sound. Thus mimics are possible. Also if self-consciousness is a behaviour the same process of mimicry can be carried out by sufficient mechanical hardware in the case of self-consciousness.

 

[Belz...]

The whole Pzombies are incoherent argument relies on this assumption

It's not an assumption. The P-zombie argument only works if dualism is true, and dualism is incoherent. One way or another the whole thing fails.

 

[Nonpareil]

Excuse me, but I am not aware that the experiment has been successfully carried out.

Which doesn't mean that we are just speculating.

Each of those inferences is an assumption.

No, they aren't. They are observable, rational, verified conclusions. We know that these things are true.

You really have an issue with differentiating between assumptions and conclusions. And definitions.

Just screen out each of the characteristics of a human which are different to the computer, one by one, including its physical characteristics (as we are discussing materialism/physicalism).

Which at no point indicates any sort of non-computational factor in consciousness.

The whole Pzombies are incoherent argument relies on this assumption, as I pointed out when I discussed it with PixyMisa, a while ago.

That's not an assumption. That's a definition.

You keep making this same mistake.

The sound waves as they enter the microphone are indistinguishable from the real chainsaw. As it is the behaviour of the lyrebirds vocal chords which result in the same sound, it has successfully mimicked the sound. Thus mimics are possible.

Yes. You have a perfect mimic of a specific pattern of sound waves. It is that sound. The same sound can be produced by two different means. This is precisely my point.

 

[Belz...]

Yes. You have a perfect mimic of a specific pattern of sound waves. It is that sound. The same sound can be produced by two different means. This is precisely my point.

Yes, that's probably the best analogy for machine consciousness.

 

[LarryS]

It has all the characteristics of a conscious entity. Therefore, it is a conscious entity. That is what we mean when we say that something is something.



In which case your argument is not that a perfect mimic of consciousness would not be conscious, but that constructing a non-organic perfect mimic of consciousness is impossible.

For which there is no evidence.

A couple of Saturday's ago I was listening to a science show on radio, and they were discussing a Mattel toy of some years ago called the Furby. Out of the box the Furby spoke no English, but with use it 'learned' to speak. Given enough 'on' time, the Furby would cry and say "I'm afraid" if hung upside down. The question posed on show was Is the Furby conscious because it expressed fear? Well the kids who were playing with the Furby could not keep the Furby suspended upside down . . . they were compelled to right it because they felt sorry for it. There was some consensus on the radio show that the Furby displayed a thin level of consciousness. Mattel was able to program enough behavior in the Furby to fool kids. So is that what we're talking about here - program enough behavior into a machine to fool a pool of humans in a lab into thinking a machine is conscious, or is the fooling actually a conscious behavior?

 

[Nonpareil]

So is that what we're talking about here - program enough behavior into a machine to fool a pool of humans in a lab into thinking a machine is conscious, or is the fooling actually a conscious behavior?

Whether or not an entity can make other people believe it is conscious or not is irrelevant. What others believe about it has no effect on whether or not the entity actually meets the definition of consciousness. It is highly likely that a conscious entity would be able to convince others of its consciousness, but not strictly necessary; this is an effect of being conscious, not part of consciousness itself.

Non-conscious entities, as in your post, can still make people believe they are conscious. That is, after all, the point of mimicry. But they are not perfect mimics. Furbies do not act in all situations as though they are conscious; they exhibit some associated behaviors, not all. If they did exhibit all conscious behaviors, they would be perfect mimics, and therefore conscious, but they do not.

And yes, I am aware that the definition of "consciousness" is rather nebulous at the current time. I am not aware of any definition that permits Furbies as conscious entities, and the above argument still holds regardless; it simply becomes a matter of defining consciousness.

 

[MRC_Hans]

Not necessarily. In the case of the faster predator, standing still could be an equal or better strategy than running. If the organism can't comprehend or answer the question "can/does it see me?" then it's a crap shoot either way. In the case of the deceptive rival, standing still is definitely the better response.

I think this is oversimplified. You are right in the sense that many animals tend to ignore unknown entities. For instance, animals that have never encountered humans will not often flee from them. This makes sense in that if you don't know what to do, doing nothing is as good as anything.

However, it does not require any high level of intelligence and certainly not self-awareness to be able to recognize various classes of known threats, and react accordingly, as in take cover/run/stay and fight.

Hans

 

[punshhh]

It's not an assumption. The P-zombie argument only works if dualism is true, and dualism is incoherent. One way or another the whole thing fails.

You can say it all you like. If the mimic can perform all the actions, it is conceivable that it can do it mechanically without consciousness. If it can't there must be some kind of computation requiring a fully self-aware presence.

Why would there be such a requirement? It's only computation after all.

 

[punshhh]

Which doesn't mean that we are just speculating.

Until the results of the experiment are in, that's exactly what you are doing, just speculating.

No, they aren't. They are observable, rational, verified conclusions. We know that these things are true.

They may be true, it's the conclusions you draw from them which are assumed.

You really have an issue with differentiating between assumptions and conclusions. And definitions.

I may use the terms in a slightly different way to yourself. But you seem to have difficulty comprehending the argument. We are looking at it scientifically here, definitions outside the parameters of the experiment are entirely irrelevant as I have pointed out.

Let me describe the experiment again:

What we want is two objects(physical) which are self-conscious. One is a human (the control), the other is AI(or some other construction).

We have the control, a human, but we don't know how it is self-conscious, so it is a black box. (Correct me if I am wrong).

We need to construct an object (which with our knowledge of the control is a black box) and is self-conscious.

We have no way of screening out the mimics (you have conceded that mimics are possible).


Question; how do we construct the object and how do we make it self-conscious?

Which at no point indicates any sort of non-computational factor in consciousness.

No one has carried out the screening. They may have observed cases of brain damage or which parts of the brain light up at different times etc. They haven't put a brain in a jar, or brought one back to life, or shown self-consciousness in a dead brain.

That's not an assumption. That's a definition.

that is nonsense, as the thought experiment is specifically designed with the definition that it is not conscious. So it isn't, unless it can be shown that it would be self-conscious. Show it?

You keep making this same mistake.

its not a mistake, there is an error in the reasoning which I am pointing out.

Yes. You have a perfect mimic of a specific pattern of sound waves. It is that sound. The same sound can be produced by two different means. This is precisely my point.

Ah, but the equivalent is a mimic of a human without the thinking (self-consciousness) behind the sound, a parrot, it just reals it off parrot fashion. The lyrebird does not cut wood, that bit is missing. In the Pzombie the thinking " I am what I am" is missing.