Multiple Drafts Model Of Consciousness

[Paul C. Anagnostopoulos]

Originally Posted by Ian

I certainly do not think it is impossible. I've often mentioned that it should be eventually possible to build an android whose behaviour in practice (including everything it says) is indistinguishable from a human being's. Just like we'll be eventually be able to create a chess program where we cannot tell that it is a computer playing (we can tell now, computers play in a particular characteristic way).

Then on what grounds would you refuse to grant consciousness to that android?

~~ Paul

[David Bailey]

Originally Posted by Paul C. Anagnostopoulos

Agreed. So let's consider an unborn fetus. Can you describe what he feels the first time he stubs his toe in the uterus? No, you cannot. It's quite likely he feels nothing, because he has not learned to hurt. How long do we have to run the simulator before we expect typical pain feelings?


Please describe how we determine whether the program is "conscious in itself."


Why not?

~~ Paul

I thought it was commonly accepted that foetuses can feel pain towards the end of pregnancy. However, this is totally irrelevant, because the simulation can start with your brain exactly as it is - with a life time of conditioning - combine it with various inputs, and generate an output. This process is really checking a theorem, and it seems pretty nonsensical to expect it to be conscious!

By 'conscious in itself', I mean that the computers can do all sorts of things if they have additional hardware attached, but you don't really want to specify additional hardware (because it is damn hard to think of what it might be - other than a Ψ-detector), so you are talking about a bare computer being conscious just because it is doing a calculation!

Do you really think that a pass of Turing's test in the manner I described would be valid? The amount of programming involved would be minimal, the thing would just be a gimmick.

David

[Interesting Ian]

Originally Posted by Paul C. Anagnostopoulos

Then on what grounds would you refuse to grant consciousness to that android?

~~ Paul

We know why the android behaves as it does. We can take it to bits to see how it works. We note that it doesn't say the things it does because of consciousness, but rather because of the execution of algorithms. So we can't infer it is conscious.

[Interesting Ian]

Originally Posted by David Bailey

I am very surprised you say that - I think it is a complete theoretical abstraction to imagine an android indistinguishable in behaviour from a human - except not conscious (which is what I assume you mean).

I see no reason why an android's behaviour could not approximate very closely a human being's behaviour. Of course in the 2 cases the origin of the behaviour are different -- a human being's behaviour comes at least partially from conscious decisions, the android's behaviour is merely the result of the execution of algorithms.

Chess and similar problems are an incredibly narrow subset of human thought, and even there the computer has to use brute force methods to essentially cheat. Artificial Intelligence has been a spectacular failure, I would say.

I presume that they thought it was relatively easy since they assume the brain creates consciousness, the brain is essentially a computer, so it shouldn't be too difficult to create AI indistinguishable from a human beings responses.

But if the brain doesn't create consciousness then obviously it will be much harder. But give AI a 100 years.

[David Bailey]

Ian,

I think you are falling into the trap of assuming that something can have real AI without being conscious. That means that consciousness is not really necessary for our behaviour. That idea features in some science fiction novels, but there is not a shred of evidence for it. My impression is not so much that AI has a long way to go - rather that it doesn't know where the hell it is going!

If someone demonstrated an android such as you describe, I would test it as best I could, and ultimately join Paul and Chris as a physicalist.

David

[Interesting Ian]

Originally Posted by David Bailey

Ian,

I think you are falling into the trap of assuming that something can have real AI without being conscious.

We have absolutely no reason to think that an android is conscious, it's just following rules.

That means that consciousness is not really necessary for our behaviour.

I don't know what you mean. At least in our own case we know that consciousness is causally efficacious in initiating some of our behaviour. Conversely we know that consciousness is not necessary in initiating any of the android's behaviour.

I'm baffled why you're having a problem with this. Surely it should be possible to create an android which is almost indistinguishable from a human being's behaviour. Not completely indistinguishable since our behaviour is non-algorithmic, but nevertheless extremely close. I would have thought it's just a question of technology.

That idea features in some science fiction novels, but there is not a shred of evidence for it. My impression is not so much that AI has a long way to go - rather that it doesn't know where the hell it is going!

If someone demonstrated an android such as you describe, I would test it as best I could, and ultimately join Paul and Chris as a physicalist.

David

But what reasons or evidence would you have that it is conscious?? You know that your behaviour is at least partially caused by your consciousness. But you know that the android's behaviour is entirely a result of the execution of algorithms. If it is conscious its consciousness doesn't do anything! It's alleged consciousness has no causal impact on the world. We know exactly why it does what it does -- and consciousness plays no part here! So we have no reason to think the android is conscious.

[David Bailey]

Ian,

I don't think consciousness can be separated from what consciousness does. It is not that I think your android would be conscious - I just think it is impossible!

I see absolutely no evidence that the subject of AI is going anywhere, even after the large sums of money that were spent on it. I don't think this is an accident - because you need real consciousness to get real intelligence!

If consciousness were not necessary to get human or animal behaviour, you can bet a lot that it would never have evolved!

David

[Paul C. Anagnostopoulos]

Originally Posted by David

I thought it was commonly accepted that foetuses can feel pain towards the end of pregnancy. However, this is totally irrelevant, because the simulation can start with your brain exactly as it is - with a life time of conditioning - combine it with various inputs, and generate an output. This process is really checking a theorem, and it seems pretty nonsensical to expect it to be conscious!

Why? Perhaps if you define what you mean by conscious we could make some progress.

But regarding where the consciousness is, I'll make a deal with you: I'll agree that a dormant computer program and its data, sitting on a shelf, is conscious if you'll agree that a human in an induced coma is conscious.


~~ Paul

[Paul C. Anagnostopoulos]

Originally Posted by Ian

We know why the android behaves as it does. We can take it to bits to see how it works. We note that it doesn't say the things it does because of consciousness, but rather because of the execution of algorithms. So we can't infer it is conscious.

Then neither can you infer that a human is conscious, because you don't know why it says the things it does.

~~ Paul

[David Bailey]

Ian,

I think what you are saying stems from a misunderstanding of Chalmers' concept of 'hard and 'easy' problems. Making an android that behaves like a human is probably just as hard as making it experience qualia - it is just that we can't prove it is as hard.

David