Has consciousness been fully explained?

[PixyMisa]

Assumption. Certainly not established by the Church-Turing Thesis.

Keep reading the thread.

No exceptions, unless there could be exceptions.

Argument from making stuff up.

Show me a peer reviewed paper where a mathematician has formally proved that a Turing machine, or lambda calculus, or recursion, any of the whole list of other computational methods can simulate a brain.

You do realise that this is what the entire field of computational physics is about, right? A brain is just another physical system.

Say it's a brain.

There is of course no evidence whatsoever that a brain is more powerul than a Universal Turing Machine but, for the sake of argument...

You want to simulate a brain with a brain?

If in some universe brains were more powerful than Turing Machines, we could build computers based on the same principle and run the simulation on those computers.

Unless you believe in magic...

If you can't explain exactly how that brain works then you aren't likely to build a machine that uses the same kind of process.

Are you or are you not invoking new and contradictory laws of physics?

If not, your argument is baseless.

If so, your argument... Is baseless, really.

Speak for yourself. Any sufficiently advanced thinking appears to be indistinguishable from magic.

And any insufficiently advanced thinking is indistinguishable from a belief in magic.

 

[Dancing David]

Give an example of something that looks like consciousness:

Everything we label as consciousness.

 

[Drachasor]

Here are some questions you've responded to without answering:

What is it that you think the definition of information is? (asked twice)
Because...? (asked twice)
Do you consider all physical interactions to be information processing? (asked once directly and once indirectly)

That's roughly 6 apiece. I'll call us even.

I've said what information is in regards to the brain. So maybe you should actually read what I wrote instead of saying I didn't answer a question that I've answered twice.

Obviously there are different sorts of information processing, but what goes on in the brain is highly organized (and again, I've talked about this in particular).

Do you know what to actually comment on my thought experiment because you don't have a sensible answer that agrees with what you've been saying?

This particular claim of Pixy's is so obviously false, and so entirely unsupported by any of the references that he's ever quoted, that one can't be surprised that he doesn't consider that any contrary arguments have been made.

It's blatantly obvious that whatever your opinions about the computational nature of the brain and the origin of consciousness, you can't remove the brain and replace it with an arbitrary computer. The person would die. And yet Pixy keeps repeating the same claim. Now, obviously he doesn't actually believe that a brain can be replaced by a computer - but since that's what he keeps saying, how can we have a reasonable argument?

That's why he uses phrases like "more powerful" rather than the more technical and more accurate language used in the real Church-Turing thesis. What Church-Turing really claims is so limited and so specific, it barely relates to this subject at all.

You could certainly replace a brain with a computer in theory. Connect up the nerves to a system that translates the impulses into something the computer can processes and respond to and the body can stay alive assuming the computer is programmed correctly. Better programming and such a cyborg could listen, talk, walk, etc.

As I've explained plenty of times - there are no physical processes going on in a computer that aren't, in differing quantities and ways, going on in a rock. That in no way implies that we can't differentiate between rocks and computers.

The main difference between rocks and computers is not in the realm of physics - it's that we can do useful things with computers. They are tools.

In particular, there is no physical activity of "information processing" that can be objectively defined as taking place in any quantifiable way in the computer and not in the rock. There's obviously an engineering difference between their behaviour, and the computer can provide information to us. However, in terms of information in its physical sense, the rock is processing information as well.

However, if you'd rather argue against the points I'm not making, I can't stop you.

Well, you're blatantly wrong. The fact that computers do useful things is an indication things are going on inside it that aren't going on inside a rock. You are contradicting yourself all over the place here, claiming there is no quantifiable difference and then quantifying the difference (such as the difference in their behavior, which is most certainly objective). The brain is much the same in this regard and it makes just as much sense to say there's nothing novel going on in it either.

 

[Drachasor]

Eh, if the anti-simulation side is sticking to the ""rocks and computers are the same" argument, then it is kind of hard to take them seriously.

 

[rocketdodger]

I hope so. This does imply that nothing of import occurs in brains that is not neuron activity, which I agree is the current consensus and may well be right. That plus the ability to actually mimic neuron connectivity and operations, and that we will eventually know enough to perform that engineering feat.

Probably do-able at atomic level rather than planck level is bonus for this approach; maybe at molecular level will do it, too.

What if you replace some of the artificial neurons with computer controlled proxy neurons -- they behave exactly like a real neuron as far as neighboring neurons can tell, but internally they just run a simulation of a neuron on a computer.

Still conscious?

 

[rocketdodger]

Eh, if the anti-simulation side is sticking to the ""rocks and computers are the same" argument, then it is kind of hard to take them seriously.

Its an old religious technique.

Claim that according to science everything is mush, thus if we are different than mush it must be because of something science can't account for.

Then reject any argument showing that science doesn't assert everything is mush I.E. "stick-to-your-strawman-as-if-your-life-depends-on-it."

Its an old religious technique.

 

[!Kaggen]

Its not issue that consciousness can be simulated which is the problem Pixy its the impression you are creating that consciousness has been simulated.

 

[westprog]

Well, you're blatantly wrong. The fact that computers do useful things is an indication things are going on inside it that aren't going on inside a rock.

Why change "physical processes" to "things"? I explained what I meant, and you unexplained it. Of course there are things going on in a computer which aren't going on inside a rock. But they break down to quite similar physical processes.

If it were a case of differentiating between the rock and the computer, then it might be quite acceptable to point out the considerable quantitative differences in the physical processes going on, and to thus provide a precise physical differentiation between them. But that's never been the point of dispute. The claim is that there's something going on in the cell and the computer, which isn't going on in the rock. This is called "information processing", and is typically justified with handwaving and not much else. I'm insisting that if there is commonality between the cell and the computer, but not with the rock, then this needs to be explained in physical terms. Since the physical differences between the computer and the rock apply just as much between the cell and the computer, then either all three should be classed separately, or all three should be classed the same.

It's not that I haven't explained this many times. Nevertheless, the claim is always repeated that I can't tell a computer from a rock. RD has now thrown in dark hints about my sinister motivations. And the point remains unaddressed, and the commonality between cells and computers is unestablished.

 

[AlBell]

What if you replace some of the artificial neurons with computer controlled proxy neurons -- they behave exactly like a real neuron as far as neighboring neurons can tell, but internally they just run a simulation of a neuron on a computer.

Still conscious?

What were you replacing neurons with in the original scenario if not already a 'computer controlled neuron'?

A simulation on a computer is what you'd need, with I/O ports to the real brain.

At the moment I suspect size will be a problem for any artificial neuron that isn't biochemical engineering.

 

[westprog]

Its an old religious technique.

You can say that again.

Its an old religious technique.

And meanwhile the physical connection between cells and computers remains elusive. But I only say that because I want to introduce prayer in schools and deny evolution.

 

[AlBell]

Eh, if the anti-simulation side is sticking to the ""rocks and computers are the same" argument, then it is kind of hard to take them seriously.

Go ahead an detail how they differ in a way that doesn't depend on our subjectively to notice the difference.

 

[westprog]

What were you replacing neurons with in the original scenario if not already a 'computer controlled neuron'?

A simulation on a computer is what you'd need, with I/O ports to the real brain.

At the moment I suspect size will be a problem for any artificial neuron that isn't biochemical engineering.

But that's not the critical issue.

A simulation of neuron behaviour is a very useful thing if you want to study how neurons behave. I'm sure that such a thing has been written many times. However, it wouldn't recognise and transmit real-time electrochemical signals in the way a real neuron would. Such a thing is not necessary in a simulation. It's not what a simulation is for.

A component which could replace an actual neuron would be an entirely different thing to a simulated neuron. A simulated neuron can receive a data packet and produce another data packet, giving us insight into how nerves work. A replacement neuron would have to actually do the work.

This is the point that Pixy totally misses in the absurd claim that a brain can be replaced not just by some kind of computer, but any computer. It involves a total misunderstanding of what the brain actually does by abstracting its behaviour into computational terms.

 

[westprog]

Go ahead an detail how they differ in a way that doesn't depend on our subjectively to notice the difference.

I expect to see the old "you can't send email on a rock" rebuttal RSN.

 

[Drachasor]

Why change "physical processes" to "things"? I explained what I meant, and you unexplained it. Of course there are things going on in a computer which aren't going on inside a rock. But they break down to quite similar physical processes.

You could say the same thing about cells and rocks or cells and computers. It's just as accurate (which is to say, it is insanely wrong).

If it were a case of differentiating between the rock and the computer, then it might be quite acceptable to point out the considerable quantitative differences in the physical processes going on, and to thus provide a precise physical differentiation between them. But that's never been the point of dispute. The claim is that there's something going on in the cell and the computer, which isn't going on in the rock. This is called "information processing", and is typically justified with handwaving and not much else. I'm insisting that if there is commonality between the cell and the computer, but not with the rock, then this needs to be explained in physical terms. Since the physical differences between the computer and the rock apply just as much between the cell and the computer, then either all three should be classed separately, or all three should be classed the same.

No. I don't have to show that a computer is the same a cell. You have to show that consciousness can only ever arise from cells. That there is something special inside the cell that creates consciousness.

In terms of processing neural signals (duplicating input and output), a computer is perfectly capable of mimicing that. Unless there is something magical about the cell in particular, which you must demonstrate, then a computer can certainly capture the important aspects of the brain.

 

[Drachasor]

Go ahead an detail how they differ in a way that doesn't depend on our subjectively to notice the difference.

You're joking, right? A rock is a disorganized mass of particles. A computer is an organized mass of particles. That's just for starters. They react very different to input (e.g. the state of the rock doesn't change whereas signals are carried throughout the computer). There are plenty of other differences as well.

Do you also think there's no objective difference between a hammer and rock? There's no objective difference between a skyscraper and a cave? How is anyone supposed to take you seriously?

 

[Ichneumonwasp]

Go ahead an detail how they differ in a way that doesn't depend on our subjectively to notice the difference.

It's an objective fact that computers work with electricity of a certain voltage and amperage running through them, and that they have logic gates opening and closing at submicrosecond speed. Rocks don't.

 

[Ichneumonwasp]

A component which could replace an actual neuron would be an entirely different thing to a simulated neuron. A simulated neuron can receive a data packet and produce another data packet, giving us insight into how nerves work. A replacement neuron would have to actually do the work.

This is the point that Pixy totally misses in the absurd claim that a brain can be replaced not just by some kind of computer, but any computer. It involves a total misunderstanding of what the brain actually does by abstracting its behaviour into computational terms.

Wait a second. I don't think anyone is arguing that the epiphenomenal simulation -- what you might see on a computer screen if we set it up so -- is doing any work. The argument is that the pattern of activity in the computer that creates that simulation is doing the work. If you simulate a neuron and then simulate neuron function, it is the function that is important; and that is always represented in movements within the computer. It's the pattern of movements in the computer that 'replicate' the function of the neuron that matter.

I think you guys are using the word 'simulation' in entirely different ways and talking past one another.

 

[Meed]

I've said what information is in regards to the brain. So maybe you should actually read what I wrote instead of saying I didn't answer a question that I've answered twice.

I've read what you wrote (quoted below) and I still don't know what you think the definition of information is or whether you consider all physical interaction to be information processing. If you think I missed your answers then please highlight them and explain.

What do you think the definition of information is?
Do you consider all physical interaction to be information processing?
On what do you conclude that everything the brain does is information processing?

Do you know what to actually comment on my thought experiment because you don't have a sensible answer that agrees with what you've been saying?

I've already discussed similar thought experiments with other posters. But if you answer my questions, then sure I'll address yours.

http://en.wikipedia.org/wiki/Information_Theory



Physical interactions within the brain is how it processes information.



It's either an emergent property or it is not. If not, then some part of the system does it specifically or it's magic. I'm assuming it isn't magic.

What does the brain do?

How does it interact with the rest of the body?

Treat it as a black box (initially).

Blood brings it what it needs to run. Nerves feed it information and output information (the blood does a little bit of that too).

We can easily define the information the brain receives in terms of nerve impulses and the bits of hormones and the like. All quite quantifiable.

It takes that stuff, processes it, and the output comes out. The output tells the body what to do in terms of information. Neurological studies on the brain further reinforce this view that it is an information processing center (complete with areas dedicated to certain kinds of processing). All this can be perfectly modeled in principle.

This isn't like a plane or a boat or water. Model the brain and you copy the most essential thing it does; process information.

Edit: Or let's put it another way. Let's say you can simulate the Earth and all inhabitants perfectly. Each person has a simulated mind, etc, etc. They are aware of what is going on around them, etc, etc. They are capable of thinking the same thoughts as their real counterpart, feeling the same emotions, experiencing the same pain, etc, etc, etc. In all respects their behavior as it relates to the simulation is indistinguishable from a real person. On what basis can you say such a person isn't conscious? On what basis can you say their thoughts and feelings don't count? Are you claiming that if we are actually "living" in that simulation then we aren't really conscious? How is that not completely arbitrary?

 

[Drachasor]

What do you think the definition of information is?
Do you consider all physical interaction to be information processing?
On what do you conclude that everything the brain does is information processing?

With a very broad definition of information, you could consider a lot of physical interactions to be information processing, though not all (given entropy).

With regards to the brain, information can be clearly defined, as I have already gone over. As to how the brain is doing information processing, I answered that as well starting with a thought experiment. If you're incapable of comprehending it, then there's not much I can do for you unless you state what is causing your comprehension problem.

I suspect your actual reason for not giving a substantive response to my post is that you don't have a good one though.

 

[rocketdodger]

What were you replacing neurons with in the original scenario if not already a 'computer controlled neuron'?

A simulation on a computer is what you'd need, with I/O ports to the real brain.

At the moment I suspect size will be a problem for any artificial neuron that isn't biochemical engineering.

Something that functions like a neuron using different materials, I.E. there is something like an action potential and synaptic channels etc but there are no organelles or any other traditional "cell" stuff. And of course the materials are synthetic and not biological.

The point is, if you agree that a brain will still be conscious when you replace some neurons with artificial ones, and then computer controlled proxies, why would just a full on replacement of the whole thing not be conscious? Where is the line drawn?