Has consciousness been fully explained?

[Beth]

Real-time scans revealing brain activity show characteristic patterns of activity across the brain corresponding to the mental tasks being undertaken. Many of these patterns are broadly similar between individuals, and sufficiently consistent in a single individual that they can be used to tell which of a limited set of objects the individual is visualising. Pretty crude at the moment, but it's a start.

Keep in mind the "limited set" is actually just two images (at least in the study I've seen). And successfully determining which object they're visualizing, just means "right more than 50% of the time", according to a statistical significance test.

It sounds pretty amazing. I'm sure they will get better at it over time. But it's got a long way to go and no telling how far we get with it.

Doesn't a pattern necessarily contain information? How else does one recognize it as a pattern?

Yes. Interestingly enough, the more regular the pattern, the less information. Densely packed information is basically indistinguishable from random noise. I don't think there is any way to separate information from pattern or from random noise. But if somebody has any good ideas about it, I'd sure like to hear them.

But this is the logical extension of their claims.

Unless, of course, they can explain why consciousness is some sort of exception.

Which so far, no one has.

Actually, I think there have been some good posts answering the question of why consciousness is an exception. I'm too lazy to go back and find them and you replied to them so I know you saw them too. Suffice it to say that while you may not find them convincing (I agree), I do find them as least reasonable and consistent if you accept that premise.

What do you think a pulse is? Do you think it's along the lines of Platonic dualism or is it the functioning of the body? If you believe the first, then certainly a simulation with a pulse is possible. If you believe the second, then in a perfect simulation of the body, a pulse is inevitable.

I think this is a very good summation. It's why I find it odd that the strict materialists argue for consciousness being an exception. To claim that consciousness is solely an informational construct is, to me, much the same as claiming the existence of an eternal soul for every conscious creature. Patterns exist, as do all mathematical constructs, outside of time and space. And clearly mathematics also exists with us in our time and space.

Saying you can program a computer to engage in real-world behavior is ridiculous, whether you're talking about pumping blood, regulating temperature, or turning the process of conscious awareness on and off.

Actually, I don't think it's that odd of a notion. Certainly, it's well captured in robot lore. Think of C3P0 in the original Star Wars movie. Top down programming would make it quite easy to implement. If we eventually program conscious machines, I'd be surprised if we didn't program something like that. We certainly program machines to regulate temperature and pump all sorts of liquids around, both physically and in simulations. If someone develops an app for consciousness, I'd expect it to be on everyone's i-phone in no time.

Your induction has failed you. I am quite agnostic.

Unfortunately, just as some people post here who apparently honestly believe that there are no true atheists, that instead they must be angry with god, others here apparently cannot believe that any intelligent adult could be agnostic. Either you are an atheist or you must be religious. Apparently, if you don't agree with the truth of the way they see life, the universe and everything (there ain't no god, no how , no way, no where), then you must be secretly religious, harboring some yearning to believe. Otherwise, you would be an atheist. I don't know how people can be so convinced about the motivations of other posters, but it is quite clear to me that people who become convinced about the secret motivations and agendas of people who disagree with them on the internet did not reach their conclusion via logic and evidence.

In response to a possible similar challenge next time I'll try to stick to asking for maybe the mathematical description of a couple falling dominoes

There are many interesting and unsolved mathematical problems. It is unknown if theoretically we could describe everything in the world mathematically. It is a faith-based claim to assert that it is true. It may be true, but extrapolation doesn't always go where you think it will based on past history. Might turn asymptotic at some point and only approach a constant limit.

 

[Drachasor]

I think this is a very good summation. It's why I find it odd that the strict materialists argue for consciousness being an exception. To claim that consciousness is solely an informational construct is, to me, much the same as claiming the existence of an eternal soul for every conscious creature. Patterns exist, as do all mathematical constructs, outside of time and space. And clearly mathematics also exists with us in our time and space.

A simulated brain is still made out of matter. There's no particular reason to think consciousness has to arise from a particular brand of matter. It isn't like some atoms produce bits of consciousness on their own. Indeed, we have evidence of simulated mechanics working regarding cyphers, computer hardware (ZSNES is one of many examples), and other things.

Frankly, I don't see how one can argue the brain is not an organ that processes specific kinds of information. We have input going into it via nerves and input leaving it via nerves. How is that information processing? I'll grant it is a bit more complicated than that given hormones and how it relates to blood. So in a full-on replacement, you'd need a more complex device to duplicate the role of the pituitary gland, but again, that doesn't seem to be a factor in consciousness. People with damaged pituitaries still got "it."

Secondly, you're the one abandoning materialism by claiming patterns and math exist outside of time and space. Though, I guess you already know that. Math doesn't exist outside of a particular context anymore than the works of William Shakespeare exist outside of the context of this universe (or any other universe that might exist where they have also been created).

 

[quixotecoyote]

What do you think a pulse is? Do you think it's along the lines of Platonic dualism or is it the functioning of the body? If you believe the first, then certainly a simulation with a pulse is possible. If you believe the second, then in a perfect simulation of the body, a pulse is inevitable.

See how ridiculous that is?

False analogy.

A better analogy would be pulse rate. Your simulated heart would create a simulated pulse. However, the simulated pulse would have a rate. This pulse rate is the abstraction of what the heart is doing. This informational pattern will inevitably arise every time you perfectly simulate a heart.

All analogies break down, but yours was a false analogy from the getgo.

Consciousness is a bodily function.

A conscious machine will have to do something equivalent physically, whatever that turns out to be.

On the other hand, if you only use enough resources to run a program, that's all you're going to get.

Saying you can program a computer to engage in real-world behavior is ridiculous, whether you're talking about pumping blood, regulating temperature, or turning the process of conscious awareness on and off.

And this is the erroneous distinction that you keep coming back to.

You keep thinking there has to be a physical component for consciousness but you can't argue why, you have to keep resorting to analogies and so can never make headway in your argument. So, just for a refreshing change of pace, here is a completely analogy-free argument.

Consciousness is not a physical object or a physical action. While definitions for consciousness vary, it's commonly referred to as a certain kind of state. Specifically, one thing that crops up in a lot of definitions is 'awareness.'

We know that the brain can create such a state of awareness through the interactions of various biological components, primarily neurons.

So if we replicate those various biological components in a software simulation, we would expect to create the same state. The only difference is that the state would be formed within software rather than wetware, which is not a requirement to fit the definition of conscious.

 

[Meed]

False analogy.

A better analogy would be pulse rate.

Why is that a better analogy? And why does it matter? A rate is just a measure of changes per time.

Your simulated heart would create a simulated pulse. However, the simulated pulse would have a rate. This pulse rate is the abstraction of what the heart is doing. This informational pattern will inevitably arise every time you perfectly simulate a heart.

Well, depending on the implementation of the simulation, the rate won't be the same in real time.

You keep thinking there has to be a physical component for consciousness but you can't argue why, you have to keep resorting to analogies and so can never make headway in your argument. So, just for a refreshing change of pace, here is a completely analogy-free argument.

Consciousness is not a physical object or a physical action. While definitions for consciousness vary, it's commonly referred to as a certain kind of state. Specifically, one thing that crops up in a lot of definitions is 'awareness.'

We know that the brain can create such a state of awareness through the interactions of various biological components, primarily neurons.

So if we replicate those various biological components in a software simulation, we would expect to create the same state. The only difference is that the state would be formed within software rather than wetware, which is not a requirement to fit the definition of conscious.

Some major problems with this argument. First, you say it's not a physical thing or a physical action. While this is debatable, you go on to say that it is created through physical interactions in the brain. So it doesn't really matter if it "is" a physical action if it's created by physical action. Second, software simulation cannot "replicate" biological components, nor does it create the same state. If conscious is the result of physical interactions, as you put it, then you must acknowledge that the physical interactions within a brain are very different from the physical interactions within a simulator.

 

[quixotecoyote]

Why is that a better analogy? And why does it matter? A rate is just a measure of changes per time.

The measure of changes per time is not the thing changing over time. eta: Don't expect me to fight hard on this one, it's just an analogy and argument by analogy is probably the weakest way of making an argument. It's not bad for teaching or explanation, but it's **** on toast as far as argumentation goes.

Well, depending on the implementation of the simulation, the rate won't be the same in real time.

Perfect simulation, remember?

Some major problems with this argument. First, you say it's not a physical thing or a physical action. While this is debatable, you go on to say that it is created through physical interactions in the brain. So it doesn't really matter if it "is" a physical action if it's created by physical action

You jump from my statements to your conclusion without making any arguments in between. No dice.

Second, software simulation cannot "replicate" biological components, nor does it create the same state.

Bare assertion. No argument. No dice.

If conscious is the result of physical interactions, as you put it, then you must acknowledge that the physical interactions within a brain are very different from the physical interactions within a simulator.

Bare assertion. No argument. No dice. So no I mustn't.



To elaborate, when I say consciousness is not a physical thing or a physical action, but but a particular state/pattern/arrangement of neurons and the like whether or not those neurons & co. are physical or simulated, it makes absolutely no sense for you to point out that the simulated neurons aren't physical as if it's a point against my argument.

 

[Drachasor]

If conscious is the result of physical interactions, as you put it, then you must acknowledge that the physical interactions within a brain are very different from the physical interactions within a simulator.

Not so. Simulations are done precisely because they duplicate the mathematical relationships in real objects. An accurate simulation of the brain will duplicate all important mathematical relationships within the brain. Since, as best we understand reality, all things can be described with math (leptons and bosons compose everything, and they and their interactions are described wonderfully with math), there's no reason to think the simulation would miss out on anything essential to modeling the behavior of the brain.

Just like a simulation of water and its interactions will accurately depict how it responds to stimuli, a simulation of the brain would accurately depict how it responds to stimuli. This would include all important features such as learning, emotions, consciousness, etc. Granted, the simulation would have to supply stimuli to the brain and properly model the input and output to it from nerves, but this can be done easily enough via simulating a complete environment or simply attention sensors to the outside world.

Granted, the gross physical characteristic of the brain won't exist except in the simulation. So this brain wouldn't be made of organic matter. That stuff would only exist within the simulation. That's irrelevant, unless you are proposing there is something special within the actual matter that can't be captured mathematically. Again, this simulation would duplicate the response to information coming in on nerves that a real brain does, it would model the growth of new dendrites and new neural cells, and if you connected it to the real world via sensors and other interfaces that translate between the simulated world and real one, it could interact just like any human.

Now, you might try to argue that the interface somehow contains consciousness, but this is like saying consciousness exists in our eyes or in the nerves on the skin. It's completely disproved by everything we know about the brain and how consciousness works. Also, you'd be insulting my mother,* so I'd take offense.

So the real question is, do you think consciousness is a result of neurons and other parts of the brain interacting with each other, or do you think it comes from somewhere else? If you think the the former, then the simulation will perfectly capture that interaction and will be conscious (well, as much as a human in a similar state would be). If you think the latter, then you are either arguing for dualism or claiming there is something about matter that can't be modeled with mathematics. That said, we have no evidence at all that anything in reality can't be modeled with math and in fact as best we understand it, everything can.

*She's blind. : )

 

[Meed]

Perfect simulation, remember?

A Turing equivalent machine can't necessarily replicate real-world rate of change. The simulation won't change at the same speed as the real system.

You jump from my statements to your conclusion without making any arguments in between. No dice.

I don't get your point, though. You say consciousness is not a physical interaction. Then you say it's produced by the physical interaction of neurons in the brain, so duplicating that physical interaction should produce consciousness. So how does it help your point to say consciousness is not a physical interaction?

Bare assertion. No argument. No dice.

Double standard.. It's a bare assertion for me to say software cannot replicate biological components, but it's not a bare assertion for you to say that software can replicate biological components?

http://dictionary.reference.com/browse/replicate
1. ( also intr ) to make or be a copy of; reproduce

If I can distinguish between real biological components and software simulatoins of them, then how is it "replicating" them?

Bare assertion. No argument. No dice. So no I mustn't.

How are the physical interactions in a software simulation the same as the physical interactions in the brain? If you're talking about unimplemented software, then there isn't really any physical interaction. If you're talking about, say, a computer implementation of the software the physical interaction involved with electrons moving through gates is the same as the physical interaction involved with neurons firing in the brain?

To elaborate, when I say consciousness is not a physical thing or a physical action, but but a particular state/pattern/arrangement of neurons and the like whether or not those neurons & co. are physical or simulated, it makes absolutely no sense for you to point out that the simulated neurons aren't physical as if it's a point against my argument.

You're saying simulated neurons aren't physical? If so, how are they simulated?

 

[Ichneumonwasp]

Nothing, as long as the programming is controlling some process that actually makes the behavior happen.

You can program a computer to play a CD, or to run a fan, or paint cars. But you can't get any of these real-world behaviors (or any other) via programming alone.

And yet the computationalists make a magical exception for the behavior of consciousness. That, they say, can be done with no direct physical cause -- without ever explaining how this can be achieved.

Then I think we are agreed.

I think rather than computationalists actually thinking that it occurs through programming alone that this is probably a case of poor language use. I know I fall into that trap all the time. But that is just conjecture on my part.

 

[Ichneumonwasp]

Yes, but no one is claiming that a heart rate generates consciousness.

On the other hand, if you claim that a machine running a sim of a human body itself becomes conscious by virtue of running the sim, then you're claiming that the machine running the sim does in fact adopt the behavior of the thing being simulated despite the fact that it is not doing any of the physical stuff that the actual human body physically does.

If that's the case, then it's impossible to understand why the computer would not start exhibiting all of the behaviors of the human body. Or, for that matter, all of the behaviors of anything that it simulates.

So if a computer running a sim of a racecar does not go zooming down the road, or have an oil pressure, then there's absolutely no reason to claim that it somehow begins to exhibit one particular (and only one) behavior of the human body when that system is simulated.

But here's the thing -- in the case of mental activities, it certainly can do the physical stuff the human body does. That is why earlier I made reference to why we created computers in the first place -- to do some of our mental activity. Computers do replicate some of our mental activity because they move electricity around in a way similar to, but in no way identical to, how a brain functions. So, I think it is easier to "see" how recreating the action patterns of anything mental occurs in a computer rather than recreations of other mundane functions. A sim of a heart could recreate all of the relationships between cells and the way the entire structure pulsates but there is no way to see this happening in a real object unless we link up the info from the sim to real matter and show that we have recreated in the sim the proper relationships between cells, etc. The way of showing it on a computer screen is simply an epiphenomenon. It exists in the real world as a representation when it is on a computer screen, but there really are recreations of the relationships between all the working parts of a heart in the functioning of the sim. That is why we can speak of an orange being an orange in the simulation -- because all the relationships are preserved. It's a bit of a fiction because what's really happening is all those whizzing electrons and not an actual orange.

Same with 'mental' activities, but the movements of electrons and translation of those relationships into action we can see in the real world are much easier to visualize -- because our I/O ports (mouth, ears, etc.) are not that complex relatively speaking.

 

[PixyMisa]

Nothing, as long as the programming is controlling some process that actually makes the behavior happen.

You can program a computer to play a CD, or to run a fan, or paint cars. But you can't get any of these real-world behaviors (or any other) via programming alone.

And yet the computationalists make a magical exception for the behavior of consciousness. That, they say, can be done with no direct physical cause -- without ever explaining how this can be achieved.

Category error.

 

[dlorde]

Keep in mind the "limited set" is actually just two images (at least in the study I've seen). And successfully determining which object they're visualizing, just means "right more than 50% of the time", according to a statistical significance test.

Like I said, pretty crude

But it's a start.

 

[dlorde]

So, for the computationalists, let's say that instead of running a sim of one person, you have your computer run a sim of several people.

Now, your computer is conscious of several different minds at once.

Oh noes! Your computer would go insane!
...

No, the computer won't be conscious, the simulated individuals would be conscious.

 

[Ichneumonwasp]

No, the computer won't be conscious, the simulated individuals would be conscious.

The computer would be conscious. But each consciousness would be localized so there would be no problem.

 

[dlorde]

A Turing equivalent machine can't necessarily replicate real-world rate of change. The simulation won't change at the same speed as the real system.

We can't actually do the simulation at all for the forseeable future, it's a thought experiment in which we assume that the computing technology becomes available. So lets assume that it will be comfortably fast enough to match the real world - after all, the 'switching time' of neurons (in as much as they have one) is pretty slow compared to modern electronics.

But even if the simulation were to run orders of magnitude slower than the real world, it wouldn't matter. In the simulation, unit time doesn't have to match external reality, it can be faster or slower. It would just mean we couldn't interface with it in real-time, and if we wanted to study it at real-time equivalent rates, we'd have to record its activity and play it back speeded up or slowed down.

 

[Earthborn]

You can program a computer to play a CD, or to run a fan, or paint cars. But you can't get any of these real-world behaviors (or any other) via programming alone.

Nothing can happen via programming alone. Not even programming. A computer can only function if it has the hardware to receive input from outside (for example a keyboard or mouse) and send output to the outside (for example with a screen).

If one were to hook up a computer to a robot body, (or theoretically even a human body) all that is needed to produce conscious behaviour is the right programming. Of course that is also the point at which many robot projects fail, as no one has a clue on how to program such behaviour; it would require making a simulation of a brain inside the computer, which is at the moment only theoretically possible.

And yet the computationalists make a magical exception for the behavior of consciousness. That, they say, can be done with no direct physical cause -- without ever explaining how this can be achieved.

I don't think anyone in this thread has argued such a thing.

 

[dlorde]

The computer would be conscious. But each consciousness would be localized so there would be no problem.

How (in what sense) would the computer be conscious? there would be consciousnesses within the simulation running on the computer, are you saying this alone makes the computer conscious?

It seems to me that if you could interact with one of the simulated persons, you would be interacting with the simulation, the computer is just the substrate that supports that simulation.

 

[Ichneumonwasp]

How (in what sense) would the computer be conscious? there would be consciousnesses within the simulation running on the computer, are you saying this alone makes the computer conscious?

Yes. Consciousness is an action, and the only actual action occurring is in the movements of electrons through logic gates.

The simulation is supposed to make it easier to see that consciousness should occur when the computer is running; but the program itself isn't conscious, the computer is. Running the program makes the computer conscious. Each individual consciousness is easiest to conceptualize, as you say, in each simulated person; but there isn't an actual person there to house a consciousness. We could set up an interface with what is happening in the computer so that we could see the simulation unfolding on a computer screen, but what we would see on the screen is just a representation.

 

[dlorde]

...
The simulation is supposed to make it easier to see that consciousness should occur when the computer is running; but the program itself isn't conscious, the computer is. Running the program makes the computer conscious. Each individual consciousness is easiest to conceptualize, as you say, in each simulated person; but there isn't an actual person there to house a consciousness.

OK, I see what you mean. I think it's just a difference in viewpoint. I prefer to see it as the simulation (the result of the execution of the software by the hardware) that's conscious, rather than the hardware or the software. Maybe it's because it all reminds me uncomfortably of Searle's Chinese Room...

 

[Malerin]

Yes. Consciousness is an action, and the only actual action occurring is in the movements of electrons through logic gates.

The simulation is supposed to make it easier to see that consciousness should occur when the computer is running; but the program itself isn't conscious, the computer is. Running the program makes the computer conscious. Each individual consciousness is easiest to conceptualize, as you say, in each simulated person; but there isn't an actual person there to house a consciousness. We could set up an interface with what is happening in the computer so that we could see the simulation unfolding on a computer screen, but what we would see on the screen is just a representation.

So consciousness would arise when a particular combination of switches are utilized?

That would create a scenario where enough monkeys flipping switches like these would create consciousness:

Doesn't that seem kind of odd to you?

 

[Ichneumonwasp]

OK, I see what you mean. I think it's just a difference in viewpoint. I prefer to see it as the simulation (the result of the execution of the software by the hardware) that's conscious, rather than the hardware or the software. Maybe it's because it all reminds me uncomfortably of Searle's Chinese Room...

Which is fine as far as I am concerned, but there has been considerable acrimony in this thread over the use of particular words and ideas.

Talking about the simulation is just an easier way of discussing it, I think. It is a higher level abstraction, more on our normal level of thinking.