I challenge you: your best alternate to materialism

[barehl]

Since no one has produced artificial consciousness

Quite true.

nor even a theory on how to produce artificial consciousness -

I have most of a theory but I can't say that it would definitely result in artificial consciousness.

Granted, Materialism predicts that consciousness is producible from material, but we don't know if this assertion is true - that consciousness is even create-able . . . and if consciousness is producible, that it can be produced via inert material.

I don't know what this means. As far as I can tell, consciousness is only possible based on materials and material processes. In other words; I think I can falsify Idealism. I'm not sure what you mean by inert though. Any cognitive system would need both energy and organization.

 

[barehl]

This not what I am saying, I have stated a few times that I have no objections the possibility that consciousness can be produced artificially, for example in a robot. But that if it were the case, there would be found in that artificial consciousness apparatus, a component producing consciousness, distinct from the component producing intelligence(a computer).

Comprehension, intelligence, and free-will are not the same thing. You need free-will for consciousness. You can have comprehension without intelligence. You can't have intelligence without comprehension. Both comprehension and intelligence are necessary for free-will.

I don't know of any way to get consciousness without free-will. I cannot say at this point that free-will requires consciousness.

 

[dlorde]

Comprehension, intelligence, and free-will are not the same thing. You need free-will for consciousness. You can have comprehension without intelligence. You can't have intelligence without comprehension. Both comprehension and intelligence are necessary for free-will.

I don't know of any way to get consciousness without free-will. I cannot say at this point that free-will requires consciousness.

Can you explain what you mean by free-will in this context?

 

[Nonpareil]

Can you explain what you mean by free-will in this context?

I'd also like to ask for clarification on the claim that the brain does not "function by computation", as the brain is, as has already been established, at least Turing-equivalent, which means that it does compute and is therefore a computer in the strictest sense. It is certainly an enormously complex computer, but it definitely appears to be a computer all the same.

 

[barehl]

Can you explain what you mean by free-will in this context?

Harris argued that humans didn't have free will. This is a completely rational assumption based on the idea that actions must have either random or fixed origins. In other words, it is arguable that neither random nor fixed rules can lead to free-will.

Dennett argues that free-will is the result of a momentarily dominant coalition of neural signals. Dennett argues that free-will does exist but that it isn't the classical definition because that would require indeterminsm to be true.

Both of these ideas are incorrect. Free-will is simply the adaptive decision making part of the brain. In order to make a decision you have to have free-will. Having free-will seems to lead to consciousness.

 

[barehl]

I'd also like to ask for clarification on the claim that the brain does not "function by computation", as the brain is, as has already been established, at least Turing-equivalent, which means that it does compute and is therefore a computer in the strictest sense. It is certainly an enormously complex computer, but it definitely appears to be a computer all the same.

No, this is not the case at all. The previous assumption was that cognitive theory was a subset of computational theory. But when I ran through the boundaries of Turing computable, consciousness wasn't possible. So, then I realized that I had it backwards: Computational theory is actually a subset of Cognitive theory. So the brain can make use of computation for some of its functions but no Turing Machine will ever become conscious.

 

[Belz...]

Primary knowledge is something in our experience which cannot be doubted.

And yet much of our experiences and memories are deceptive and faulty.

 

[Belz...]

Comprehension, intelligence, and free-will are not the same thing. You need free-will for consciousness. You can have comprehension without intelligence. You can't have intelligence without comprehension. Both comprehension and intelligence are necessary for free-will.

I don't know of any way to get consciousness without free-will. I cannot say at this point that free-will requires consciousness.

Twelve posts in a row, really ?

 

[Nonpareil]

No, this is not the case at all. The previous assumption was that cognitive theory was a subset of computational theory. But when I ran through the boundaries of Turing computable, consciousness wasn't possible. So, then I realized that I had it backwards: Computational theory is actually a subset of Cognitive theory. So the brain can make use of computation for some of its functions but no Turing Machine will ever become conscious.

I'm curious as to what it is that you're working on, and in what context, that leads you to this belief.

It also seems that we're talking past one another a bit. The brain is at least Turing-equivalent, since it can act as a Turing machine, so it is actually a computer and does function via computation - though it may not be based on Turing-equivalent architecture. It may be entirely true that a machine that is just Turing-equivalent can never become conscious, as I am unsure whether or not a Turing machine could fulfill Koch's requirements for integration. It may be possible that one could, but I wouldn't be surprised to learn that it couldn't, and that consciousness has to be based on a more intricate architecture.

But the brain still very definitely seems to be a computer, and the material which a computer is built out of still doesn't seem to matter. What project are you working on, exactly, that leads you to think otherwise?

 

[barehl]

I'm curious as to what it is that you're working on, and in what context, that leads you to this belief.

I'm working on concsiousness, AI, and free-will.

The brain is at least Turing-equivalent, since it can act as a Turing machine, so it is actually a computer and does function via computation - though it may not be based on Turing-equivalent architecture.

No. This assumption is incorrect. I'm familiar with computational theory, language theory, and Turing architecture. What you are talking about is completely correct within computational theory. Anything that can be computed by a modern computer can also be computed by a simple machine that uses nothing but punch cards and mechanical logic. Turing computable has been shown to be accurate for derivative computers.

The mistaken assumption is that cognition is bounded by computational theory. Computational theory is adequate for organisms that are environmentally reactive but not adequate for those that have environmental awareness.

It may be entirely true that a machine that is just Turing-equivalent can never become conscious, as I am unsure whether or not a Turing machine could fulfill Koch's requirements for integration. It may be possible that one could, but I wouldn't be surprised to learn that it couldn't, and that consciousness has to be based on a more intricate architecture.

It isn't a question of may or may not. I've already disproven this. That then leads to the question of whether enhanced architecture could become conscious. This is rather complicated so I don't have a proof yet. However, I haven't thought of anything that would prevent an enhanced architecture from having free will.

But the brain still very definitely seems to be a computer, and the material which a computer is built out of still doesn't seem to matter. What project are you working on, exactly, that leads you to think otherwise?

The brain does not fit within the bounds of Turing computable. So, if you are using this to define a computer then the brain is not. I'm working on a theory of cognition.

 

[Myriad]

Every extension or enhancement to a turing machine that I'm aware of, that can be defined in terms of how it operates, has been proven to be equivalent to (that is, can be simulated by) a turing machine. This includes such enhancements as multiple tapes, multiple heads, registers, stacks, nondeterministic rules, and even multiple whole machines interconnected to message one another by writing on each other's auxiliary input tapes.

(The only extended capabilities that make a machine provably non turing equivalent are "oracle" devices defined not by how they function but what they can accomplish, such as determining whether a given turing machine will halt.)

This strongly suggests that the ability to interact with an environment (whether in an "aware" manner or not; you'd have to define that to go any further) only makes a machine non turing equivalent is if the environment itself is non turing equivalent. Which hypothesis seems to risk begging the question.

 

[Nonpareil]

I'm working on concsiousness, AI, and free-will.

Yes, I gathered that. I meant in what context. Are you a professor? A neurologist? A computer scientist? "Consciousness, AI, and free will" is so broad a topic that it isn't at all helpful as an answer to the question.

No. This assumption is incorrect.

Well, no. To start, it's not an assumption. It's a demonstrable fact. The brain can act as a Turing machine. It is, therefore, at least Turing-equivalent. It may be more than Turing-equivalent, but it is at least Turing-equivalent, and is therefore at least partly a computer.

I'm familiar with computational theory, language theory, and Turing architecture. What you are talking about is completely correct within computational theory. Anything that can be computed by a modern computer can also be computed by a simple machine that uses nothing but punch cards and mechanical logic. Turing computable has been shown to be accurate for derivative computers.

The mistaken assumption is that cognition is bounded by computational theory.
Computational theory is adequate for organisms that are environmentally reactive but not adequate for those that have environmental awareness.

Well, again, this isn't an assumption. It's a conclusion. I'd be open to hearing whatever evidence you have to the contrary, however.

It isn't a question of may or may not. I've already disproven this.

Wonderful. Where, how, and with how much support from other cognitive scientists?

This is an actual inquiry, by the way. I'm not being facetious. I want to know.

 

[barehl]

Every extension or enhancement to a turing machine that I'm aware of, that can be defined in terms of how it operates, has been proven to be equivalent to (that is, can be simulated by) a turing machine. This includes such enhancements as multiple tapes, multiple heads, registers, stacks, nondeterministic rules, and even multiple whole machines interconnected to message one another by writing on each other's auxiliary input tapes.

You are completely correct within computational theory. This is beyond that.

(The only extended capabilities that make a machine provably non turing equivalent are "oracle" devices defined not by how they function but what they can accomplish, such as determining whether a given turing machine will halt.)

You are talking about hypercomputability. I'm aware of the Halting Problem but it isn't relevant in a cognitive context. You keep making the assumption that cognition is a subset of computational theory but it isn't. Cognitive functions can't be predicted. That would require absolute determinism such as you have for computer programs. You don't have that for consciousness.

 

[barehl]

Yes, I gathered that. I meant in what context. Are you a professor? A neurologist? A computer scientist? "Consciousness, AI, and free will" is so broad a topic that it isn't at all helpful as an answer to the question.

I guess I'm a computer scientist although this is beyond computational theory.

Well, no. To start, it's not an assumption. It's a demonstrable fact. The brain can act as a Turing machine. It is, therefore, at least Turing-equivalent. It may be more than Turing-equivalent, but it is at least Turing-equivalent, and is therefore at least partly a computer.

You have this backwards. Using conscious reason, you can duplicate the steps used in Turing computation. This means that the brain can simulate a Turing machine. That does not imply that the brain is based on a Turing architecture or that a Turing machine can think. The brain does seem to depend on neurons with logic functions. This is somewhat like a computer but not like any Turing architecture. I know people talk about neural networks but I haven't heard of one that actually accomplished anything in terms of general functions.

Well, again, this isn't an assumption. It's a conclusion. I'd be open to hearing whatever evidence you have to the contrary, however.

Well, the fact that all AI projects to date have failed is a pretty good clue. Of course, like the gambling addict, you can believe that maybe the next one will be the big winner. I'm saying that no current computational theory will ever succeed in producing robust AI.

Wonderful. Where, how, and with how much support from other cognitive scientists?

Zero. I haven't published anything yet. There are only two possibilities: I'm an overconfident crackpot with no understanding of computational theory and nothing but delusional speculation to fix non-existent problems or...

 

[Nonpareil]

I guess I'm a computer scientist although this is beyond computational theory.

You "guess" you are a computer scientist?

You have this backwards. Using conscious reason, you can duplicate the steps used in Turing computation. This means that the brain can simulate a Turing machine. That does not imply that the brain is based on a Turing architecture or that a Turing machine can think.

Well, actually, yes. This does imply - in fact, it confirms - that the brain is at least partially based on computation which could be carried out by a Turing machine. It does, in fact, mean that the brain is a computer. That is what being Turing-equivalent means.

If you claim that this goes "beyond computational theory", then your claim - if it is to have any possibility of being true at all - must not be that the brain is not a computer, because that is demonstrably untrue, but that the brain is not just a computer, or at least not a computer that can be modeled in its entirety via Turing machine.

Which is itself a problem, because then you run up against either the Church-Turing thesis (that anything computable can be computed via Turing machine) and/or a complete lack of evidence. If you say that the brain is a new kind of computer that has computational abilities beyond that of a Turing machine, well, the Church-Turing thesis has borne up under a lot of scrutiny over the years, and you need a hell of a lot to overturn it. If you say that the brain is flat-out not functioning through computation, well, there's just no evidence in support of that.

The brain does seem to depend on neurons with logic functions. This is somewhat like a computer but not like any Turing architecture.

Any logic function or series thereof can be modeled via Turing machine, period. The lack of complete neural replication is due to the fact that there are a hell of a lot of connections between a hell of a lot of neurons that are hard to map accurately on account of them being inside our heads, not because Turing machines have some arbitrary number of logic operations at which they stop working.

Well, the fact that all AI projects to date have failed is a pretty good clue.

No, they haven't. What are you talking about? The field of artificial intelligence makes incredible leaps on a regular basis - and even if it didn't, that wouldn't necessarily indicate what you say it does.

I'm saying that no current computational theory will ever succeed in producing robust AI.

And given the aforementioned issues regarding lack of evidence supporting this, no one is particularly inclined to believe you.

Zero. I haven't published anything yet. There are only two possibilities: I'm an overconfident crackpot with no understanding of computational theory and nothing but delusional speculation to fix non-existent problems or...

I refrain from comment.

 

[Dancing David]

The brain has a semi probablistci analog function, at teh neural level each neuron is potentiated or attentuated in its firing rate by the other neurons that contact it.
So if the cell receives an input and then fires it is potentiated to fire by that inputting cell, it the input it received and the cell does not fire then it is attenuated as an input cell .

So over time each cell has a 'matrix' of potentiating cells and attenuating cells. The base chance of a neuron firing is based upon its own base rate and the input of the attenuating and potentaiting neural contacts. And then there are other inputs as well.

However this means it is semi computational and based upon associative learning.

(Granted this is all old and probably out dated by now)

 

[Fudbucker]

Actually, this doesn't follow. This is what I've been working on for more than a year. Your assertion is correct in terms of naive philosophy but falls apart under any rigorous structure. Any functional system needs an energy source. Your rocks don't have that.

Here's a link to the comic: http://xkcd.com/505/

I think there's a whole lot more wrong with rock-consciousness than a missing energy source.

But it's not my pet theory, it's what some people here actually believe- that if a person moves enough rocks around "in the right way" (and it's never specified what that way actually is), a whole universe of conscious beings can be simulated.

Of course, the causal explanation for how moving rocks around in "the right way" can produce a simulated universe is not clear. If the guy moves rocks around to spell "Eat at Joe's", there's no simulated universe. But if he moves them around in just the right way (in some way that he knows will produce a simulated universe, I guess), then you get a simulated universe. It's also unclear what, exactly, is producing the simulation: the person's brain? The moving rocks themselves? The brain and the rocks? The special pattern the person is following? Which of these are necessary to simulate consciousness? Which are sufficient?

But it is taken for granted by some people here (who should know better), that if you move rocks around you can simulate conscious people. So the next time you see some kids in a dirt-clod fight, tell them to show a little respect for the fleeting conscious entity that may have just popped into existence.

 

[dlorde]

Dennett argues that free-will is the result of a momentarily dominant coalition of neural signals. ...

...these ideas are incorrect. Free-will is simply the adaptive decision making part of the brain. ...

Dennet's claim is that 'adaptive decision making' is the result of a momentarily dominant coalition of neural signals (although that's rather a weak summary of his argument). You have asserted that it isn't without explaining why you think so.

Equally, if we define free will as adaptive decision making, then 'fixed rules' (as per Harris) can also lead to this kind of 'free will'. For example, consider Google's autonomous cars; once a goal (destination) is set, they can plan a route and navigate it to the goal, adaptively allowing for other traffic and diversions. To do this, they use a (very extensive and complex) set of fixed rules. More trivially, a sat nav that re-routes you if you make a wrong turn makes an adaptive decision in doing so. If you feel that a sat nav with free will is absurd, please explain how what it does is not adaptive decision making.

If you feel that what an autonomous car does is not adaptive decision making, please clarify more precisely what you mean by the phrase, or why an autonomous car does not make adaptive decisions.

 

[Myriad]

It's also unclear what, exactly, is producing the simulation:

The computation.

 

[MRC_Hans]

But it is taken for granted by some people here (who should know better), that if you move rocks around you can simulate conscious people.

I really think not. I expect we all know that this is just an extreme analogy to illustrate that consciousness is not hardware-dependent. While a simple computer can be made using rocks (or any other element that may represent a logic state), even a PC class computer using such cumbersome methods would be impossibly huge and slow.

Hans