Materialists......

[Tricky]

Tricky,

You are trying to treat philosophy as if it was science. It isn't. Philosophy provides a set of contrasted and related positions and the "answers" are bound up in the relationships between the positions and between people the positions they hold. Philosophy cannot be judged by the same standards as empirical science. It it could, it wouldn't really be philosophy any more.

I can can well see why you put "answers" in quotes, because what you describe are not "answers" in my opinion. How can they be described as answers when there are no correct answers and no incorrect answers?

My own feeling is that philosophy is nothing more than the assumptions you make for your logical system. Yes, the philosophy of materialism makes the assumption that all real things are material and as such, there is material evidence for all real things. It is a very useful philosophy because it gives a consistent and demonstrable way of evaluating reality. A philosophy which makes the assumption that there are things which are non-material and thus do not have evidence accessible to material entities, cannot resolve any questions, because the answer to any question may be something without evidence. To me, that is the same as no answer.

 

[Tricky]

I never said it did. I said you cannot isolate data from holographic storage and transfer it in the way Tricky thinks you can.

That is obviously incorrect. If you cannot isolate data from holographic storage, then you cannot remember anything. You could not remember your own name because it would be impossible to isolate it from holographic storage.

But you can remember things. Very specific things. So we fall back on one of the oldest and most irrefutable of axioms: If it happens, it must be possible.

 

[UndercoverElephant]

I can can well see why you put "answers" in quotes, because what you describe are not "answers" in my opinion. How can they be described as answers when there are no correct answers and no incorrect answers?

http://www.chinapage.com/zen/koan1.html

 

[UndercoverElephant]

That is obviously incorrect. If you cannot isolate data from holographic storage, then you cannot remember anything. You could not remember your own name because it would be impossible to isolate it from holographic storage.

You don't get it. You can extract it from within. Or in the case of the A.I. you can extract it by "asking it questions" in the form of fiddling with its inputs and monitoring its outputs. What you cannot do is extract it from without as would be required by your machine (in the case of a brain) or by a software engineer trying to figure out how his neural network was storing the data. Understand yet?

 

[Tricky]

http://www.chinapage.com/zen/koan1.html

Why don't you tell me in your own words?

Let me put it simply as a pair of questions.

Are there any "correct" answers in philosophy?
If so, how do you tell if they are correct?

 

[chriswl]

I know it is hard for some people to believe it, but what has happened is that cognitive science has in fact discovered the problem I keep trying to get people to understand, and it has already turned to people like Heidegger for a way forward. There is a very good reason for this: the work of Husserl, Heidegger and Merleau-Ponty was originally designed with the overcoming of the subjective/objective dichotomy in mind. Those philosophers were trying to solve the very same set of problems that the materialists haven't yet had the stomach to admit the existence of but which the leading edge of cognitive science is all too aware of.

So let's assume that this new phenomenologically inspired approach allows AI researchers to produce machines with truly useful cognitive capabilities. I assume that's the intention. Are you saying that they will then have, in some sense, created what you would call "minds"?

If not, then it doesn't matter how many ideas or how much jargon they borrow from continental philosophers, their work is entirely compatible with materialism. All they are doing is producing physical machines that produce desired physical behaviour.

 

[Tricky]

You don't get it. You can extract it from within. Or in the case of the A.I. you can extract it by "asking it questions" in the form of fiddling with its inputs and monitoring its outputs. What you cannot do is extract it from without as would be required by your machine (in the case of a brain) or by a software engineer trying to figure out how his neural network was storing the data. Understand yet?

I understand that you are making a claim that cannot be substantiated. The brain itself is a machine. If we could get to the point where we can almost fully understand the workings of the brain, we would be able to extract the partial memories almost the same way a brain does.

Now you may argue that we will never understand the brain well enough to do that, and I would say you are most likely correct. Still, it is not impossible.

 

[hammegk]

All they are doing is producing physical machines that produce desired physical behaviour.

Agreed. And when the question "are you conscious" is put to one of those machines and the answer is "yes", will you agree it has the same standing as a living entity that you assign to yourself, and probably to other humans as well?

 

[Tricky]

Agreed. And when the question "are you conscious" is put to one of those machines and the answer is "yes", will you agree it has the same standing as a living entity that you assign to yourself, and probably to other humans as well?

Almost correct, but not quite. The machine would have to not only be able to answer that question with "yes" (we have voice recognition software that can do that now), but it would have to satisfy the same characteristics by which we define human consciousness. If it can do that, I would say it is entitled to the same standing as a conscious human.

 

[Arkan_Wolfshade]

Almost correct, but not quite. The machine would have to not only be able to answer that question with "yes" (we have voice recognition software that can do that now), but it would have to satisfy the same characteristics by which we define human consciousness. If it can do that, I would say it is entitled to the same standing as a conscious human.

Turing Test^2 then.

 

[UndercoverElephant]

I understand that you are making a claim that cannot be substantiated. The brain itself is a machine. If we could get to the point where we can almost fully understand the workings of the brain, we would be able to extract the partial memories almost the same way a brain does.

OK, I give up. I have tried to explain this several times and you just don't get it. You are wrong.

Now you may argue that we will never understand the brain well enough to do that, and I would say you are most likely correct. Still, it is not impossible.

You are wrong.

The only way to extract information from a neural net is to ask it questions. The only way to extract information from a human brain is also to ask it questions. This is not the result of technological limitations, it is the result of the way the data is stored. Please don't just reply "You cannot substantiate that". I can, I have, and I am sorry you do not understand it. This is nothing to do with philosophy, Tricky. It is to do with A.I., neuroscience and software engineering. Because it is not philosophy, but is science, I can tell you that you are actually wrong. You are wrong because you do not appreciate the basic difference between a normal computer and a neural network.

 

[Arkan_Wolfshade]

Waitaminute. Neural network <> holographic storage. Holographic storage is a way to store, and retrieve data. Neural networks are for sifting through data.

 

[Paul C. Anagnostopoulos]

So you are claiming that "subjective experience" IS IDENTICAL TO "physical brain processes". The problem with this is it is, well, completely impossible to defend.

We often go down this rat hole. The use of the word identical is tricky. You're citing a term for a set of inner experiences and comparing it to the name of a set of processes. They are certainly not the same concepts, yet one might be a result of the other and nothing else. It is like saying "you are claiming that the weather is identical to atmospheric processes." Sort of, but no one would use the word identical in this context.

~~ Paul

 

[Tricky]

Turing Test^2 then.

Yes, much like that. As I recall, the turing test only uses audio input. Satisfying criteria for consciousness might require much more.

 

[Arkan_Wolfshade]

Yes, much like that. As I recall, the turing test only uses audio input. Satisfying criteria for consciousness might require much more.

Turing test basically says; given a conversation between a person and a computer (assuming the computer can carry on said conversation), a third-party observer, if unable to detect which is the computer and which is the person, would declare the computer to have passed the test.

 

[UndercoverElephant]

Waitaminute. Neural network <> holographic storage. Holographic storage is a way to store, and retrieve data. Neural networks are for sifting through data.

They also store data - encoded in the network itself. That is what allows them to sift through other data.

 

[UndercoverElephant]

We often go down this rat hole. The use of the word identical is tricky. You're citing a term for a set of inner experiences and comparing it to the name of a set of processes. They are certainly not the same concepts, yet one might be a result of the other and nothing else. It is like saying "you are claiming that the weather is identical to atmospheric processes." Sort of, but no one would use the word identical in this context.

~~ Paul

Then don't use it in my context either. Part of the difference between weather and atmopheric processes is that weather is "atmospheric processes as they are perceived by me". You wouldn't use the word identical in that context, so why do you use it in this one?

 

[Arkan_Wolfshade]

They also store data - encoded in the network itself. That is what allows them to sift through other data.

They store/access data to support their primary use, but that storage is not the primary use. It looked like the two terms were, to an extent, being used interchangably in this thread, when they are not.

 

[hammegk]

Satisfying criteria for consciousness might require much more.

Prove you are conscious.

I "think" I am.

 

[Tricky]

OK, I give up. I have tried to explain this several times and you just don't get it. You are wrong.

Oh, I "get it" just fine. However, I don't agree. I believe you are wrong.

The only way to extract information from a neural net is to ask it questions. The only way to extract information from a human brain is also to ask it questions.

Arguably, that is the only way it can be done at present (and I believe there are those here who will argue with you even on that point), but that in no way makes it impossible that we will solve the way that the brain does it. After all, your brain extracts this information without asking questions. It happens, so it must be possible.

This is not the result of technological limitations, it is the result of the way the data is stored. Please don't just reply "You cannot substantiate that". I can, I have, and I am sorry you do not understand it.

I understand it just fine. I don't agree. It can be extracted because it IS extracted. Likely we will never be able to reproduce this phenomenon, but it is not impossible. I'm sorry you cannot understand that.

This is nothing to do with philosophy, Tricky. It is to do with A.I., neuroscience and software engineering. Because it is not philosophy, but is science, I can tell you that you are actually wrong. You are wrong because you do not appreciate the basic difference between a normal computer and a neural network.

I do appreciate the difference. But both of them exist. Both of them work. Both of them must have mechanisms for working. You are arguing that it is impossible, no matter how much we learn, for us to understand the mechanism of a neural network. I think you give us too little credit.