Has consciousness been fully explained?

[westprog]

I'm confused why scientific objectivity can't look at how things function at levels we recognize as "mechanical".

Of course scientific objectivity can look at mechanical devices, and at living things. The problem is when we assign physical attributes to the purposes which we have assigned to those devices, and look for common physical attributes between us and our tools.

 

[westprog]

That is not what I said, though. While rock piles follow the constraints of physics, neurons follow not only the constraints of general physics but also another type of constraint the allows for what we seem to call computation (and by that I do not mean 'arithmetic').

Neurons summate inputs to a threshold and fire if that threshold is reached. The threshold (basically) is the same for the neuron at all times (the resting potential can be changed by various factors though so that it is easier for summating inputs to reach threshold). None of that happens with falling rocks; the same physcial constraints are simply not part of that system.

Falling rocks can be said to calculate only in an observer dependent fashion. The summation to threshold occurs in neurons no matter what we want to call it. Again, if we don't want to call that computation, then I have no issue with it. But I would also argue that electrons moving through gates is not computation in the same way. Personally I think it makes more sense to call both forms of computation. They just are not arithmetic.

A non-linear response occurs throughout nature. There's nothing unique physically about what happens in a neuron.

 

[westprog]

This is one of Westprogs ploys to avoid issues, you can define it all you want. Just don't discuss digital computers with him/her.

It is politer than Franko's "Then rocks rolling down hills are conscious."

I think RD has started to realise that it's not a good idea to discuss digital computers. He's already tried to rule out physics.

 

[Ichneumonwasp]

A non-linear response occurs throughout nature. There's nothing unique physically about what happens in a neuron.

Who said there was anything unique about it? It's just ions flowing through channels resulting in the opening of other channels. The same thing happens to a lesser extent in all living cells, but they don't generally have the same properties as neurons.

But falling rocks have very little in common with neurons. Sure, they're both made of matter.

 

[Meed]

I suggested toasters might be conscious, given the very broad definition of SRIP.

I never said toasters are conscious like humans, or that they exhibit human consciousness.

I might point out that the other side has implied that only healthy, normal, developed humans are in fact conscious, while mentally disabled humans, babies, monkeys, dogs, dolphins, and all the rest are not.

Oh, yes. That was your interpretation of this post:

Who's claiming something is beyond basic consciousness? Consciousness includes sofia, subjective experiences, emotions, etc. These don't go "beyond" basic consciousness. They are consciousness. Putting a qualifier like "basic" in front of "conscious" just muddies the water. Something either has conscious experience or it doesn't.

So dogs are not conscious? Nor babies? Or mentally disabled people? Where do you draw the line? Are only normal healthy humans conscious? Sorry I didn't know that was your position on the issue.

I don't know how you drew the interpretation you did. Either way, it's nothing to do with me.

 

[Meed]

I said there was no mathematically describable qualitative difference. Meaning, there isn't any fundamental system behavior exhibited by animals and not inanimate objects that isn't reducible to SRIP.

And I'm asking for you to support your claim.

Do you dispute that?

Yes, on the grounds that you haven't backed it up.

Can you find something that animals do that an inanimate object does not do that isn't reducible to a type of SRIP? Really -- can you?

I don't see why not. But you said this was mathematically supportable. Your response does not have anything to do with that as far as I can tell.

 

[blobru]

Of course scientific objectivity can look at mechanical devices, and at living things. The problem is when we assign physical attributes to the purposes which we have assigned to those devices, and look for common physical attributes between us and our tools.

Well, there's a problem if we confuse functional description with physical description, sure. But clearly both nature and we can develop information storage and retrieval systems (brains and computers). As concrete systems, necessarily, they each have a physical and a functional description. Where functions overlap, common terms only make sense.

 

[Beth]

That is not what I said, though. While rock piles follow the constraints of physics, neurons follow not only the constraints of general physics but also another type of constraint the allows for what we seem to call computation (and by that I do not mean 'arithmetic').

I'm not sure what you mean by a type of constraint that allows for what we seem to call computation. Could you give an example of such a constraint? That would help me understand how you are differentiating between these things.

Neurons summate inputs to a threshold and fire if that threshold is reached. The threshold (basically) is the same for the neuron at all times (the resting potential can be changed by various factors though so that it is easier for summating inputs to reach threshold). None of that happens with falling rocks; the same physcial constraints are simply not part of that system.

That's a fair point for falling rocks, but not all other non-living systems.

Falling rocks can be said to calculate only in an observer dependent fashion. The summation to threshold occurs in neurons no matter what we want to call it.

This is the problem. For example, substitute water reaching a threshhold and then falling. That would be more analogous to what you are describing neurons doing. The same issue remains.

Again, if we don't want to call that computation, then I have no issue with it. But I would also argue that electrons moving through gates is not computation in the same way. Personally I think it makes more sense to call both forms of computation. They just are not arithmetic.

I don't care what you call it either. My question is how can we differentiate between different types of things that do such 'computations' without needing an observer who extricates meaning from those computations?

They are, and so are you.

No. You are mistaken about that interpretation.

You can distinguish the computation done by neurons and computers from whatever rock piles do based on the simple fact that systems of neurons in people and systems of transistors in computers can react to their environment in ways that piles of rocks cannot.

Sure, we can observe them and distinguish between them. My question was how can we differentiate between systems that do such 'computations' without needing an observer who extricates meaning from those computations?

No observer is required for this effect to happen.

No one said an observer was required for rocks to fall either. It simply seems that an observer is required to extract meaning from the 'computation'.

And, if every human was to die, our autonomous robots would continue to function -- in a very different way from rocks -- as if nothing happened. None of this depends on observers.

Sure, they would continue to function - at least for a while. But why would their functioning be considered computation? What is different about their functioning besides the fact that we humans can define inputs and responses to inputs and then extract meaning from the outputs.

 

[rocketdodger]

Sure, we can observe them and distinguish between them. My question was how can we differentiate between systems that do such 'computations' without needing an observer who extricates meaning from those computations?

No one said an observer was required for rocks to fall either. It simply seems that an observer is required to extract meaning from the 'computation'.

Sure, they would continue to function - at least for a while. But why would their functioning be considered computation? What is different about their functioning besides the fact that we humans can define inputs and responses to inputs and then extract meaning from the outputs.

Perhaps I misinterpreted your meaning of "observer."

If you just mean some system that changes behavior due to the result of the computation one way or another, then yes you are correct, computation requires an observer -- but in most cases the observer is part of the system carrying out the computation itself.

For example the computation carried out by a cell's control pathways are observed by the cell itself, and its behavior changes as a result.

Does that make more sense?

Like I tried to explain before, my physical definition of computation is simply a partnered behavior in one system that another system can use (or interpret, or get meaning from, whatever) to determine which partition of the set of all states an external state is in. A thermostat uses the physical properties of it's element to determine whether the external universe is in a state within the { lower than temperature X | higher than temperature X } partitions. Thus a thermostat, using it's element, computes. The idea of a clean partition is supported mathematically because the states the thermostat resides in are stable -- regardless of whatever state the thermostat starts in, it will converge to either OVER TEMP X or BELOW TEMP X, all else being equal.

I guess you could call the thermostat itself the observer, if that is what you mean. Furthermore you could say that in a computer chip, one transistor observes the computation of any transitors prior to it in a logical operation.

But if you mean that computation requires an intelligent observer, then no that is not correct.

 

[rocketdodger]

I think RD has started to realise that it's not a good idea to discuss digital computers. He's already tried to rule out physics.

Well, I tried to rule out whatever it is that you consider physics, yes.

 

[Ichneumonwasp]

I'm not sure what you mean by a type of constraint that allows for what we seem to call computation. Could you give an example of such a constraint? That would help me understand how you are differentiating between these things.

Sure. There are several constraints that determine how a neuron functions. One is that any neuron receives information from only some sources. Each neuron also responds to only a part of the available sources at any one time (depending on which neurotransmitter is used in the vicinity and what receptors are available on the neuron). It takes multiple inputs for any cortical neuron to fire and they must summate in a way that exceeds the cell's threshold at the axon hillock. The duration of the stimulation from EPSPs (subtracting out IPSPs) determines the firing rate at the axon hillock (the firing rate being one of the primary means of coding information). And when the neuron reaches threshold it doesn't just sort-of-fire, but either does it fully or not at all (all or none phenomenon). The big constraints on any single neuron, then, are the funneling of information coming from other neurons and the duration of the stimulation to determine a firing rate and duration of firing. Within the much larger system, there are other issues at play having to do with recurrent loops, etc.

That's a fair point for falling rocks, but not all other non-living systems.
This is the problem. For example, substitute water reaching a threshhold and then falling. That would be more analogous to what you are describing neurons doing. The same issue remains.

I'm afraid that is not at all analogous to a firing neuron unless there are additional constraints on the system -- as nature provides for neurons. Water spilling over a bucket is much closer to the type of graded potential one sees in an EPSP or IPSP not the all-or-none action potential. But one could devise a system in which any water being present is important; one might be inclined to say that this is only possible if there is an observer to define the water flow as important, and for such an artificial system you'd be correct (when viewing it as a top-down design). But that doesn't matter for neurons, because threshold is built in to their function. No one needs to be there to define it, as they would need to be with water overflowing a bucket to make do something or mean something. The action potential simply follows.

I don't care what you call it either. My question is how can we differentiate between different types of things that do such 'computations' without needing an observer who extricates meaning from those computations?

It is based in the constraints that are already put in place within the system of the neuron. An action potential already has a meaning built into it by natural selection -- move the flow of information along. Water over flowing a bucket does not unless someone decides to impart meaning to it. The difference between the nervous system, as I have mentioned before, is that its simple 'meaning' (and that is way too loaded a word to use here) is built in from the bottom up. When we design computers we do the same thing but from the top down.

 

[Beth]

Perhaps I misinterpreted your meaning of "observer."

If you just mean some system that changes behavior due to the result of the computation one way or another, then yes you are correct, computation requires an observer -- but in most cases the observer is part of the system carrying out the computation itself.

To be honest, I was thinking of neurons, which are part of a conscious observer. Would you say that the output of a single neuron has no more of an 'observer' than the normal environmental surroundings of a pile of rock? - i.e. atomic particles bouncing around constantly? After all, we don't ever observe such outputs at a conscious level outside of a lab. Or would you say that being a part of conscious observer automatically bestows such an observer to all parts of such nervous systems?

If the first, then you haven't distinguished computation with neurons from computation with piles of rocks. If the second, then the distinction is based on the presence of a conscious observer, which is what I am saying seems to be required.

Like I tried to explain before, my physical definition of computation is simply a partnered behavior in one system that another system can use (or interpret, or get meaning from, whatever) to determine which partition of the set of all states an external state is in. A thermostat uses the physical properties of it's element to determine whether the external universe is in a state within the { lower than temperature X | higher than temperature X } partitions. Thus a thermostat, using it's element, computes. The idea of a clean partition is supported mathematically because the states the thermostat resides in are stable -- regardless of whatever state the thermostat starts in, it will converge to either OVER TEMP X or BELOW TEMP X, all else being equal.

I guess you could call the thermostat itself the observer, if that is what you mean. Furthermore you could say that in a computer chip, one transistor observes the computation of any transitors prior to it in a logical operation.

But if you mean that computation requires an intelligent observer, then no that is not correct.

How is it that the actions and/or outputs of a falling pile of rocks are computationally different from the actions and/or outputs of a your hypothetical robot world with no humans?

It seems to me that either you accept that computation requires a conscious observer to extract meaning - i.e. neurons compute but a pile of rocks doesn't or conclude that functioning robots* do the same sort of 'computation' that falling rocks do in a world with no conscious beings.

* For the sake of this hypothetical argument, all conscious robots were also eliminated by whatever tragedy struck all conscious life forms dead in this hypothetical world.

 

[Beth]

It is based in the constraints that are already put in place within the system of the neuron. An action potential already has a meaning built into it by natural selection -- move the flow of information along. Water over flowing a bucket does not unless someone decides to impart meaning to it. The difference between the nervous system, as I have mentioned before, is that its simple 'meaning' (and that is way too loaded a word to use here) is built in from the bottom up. When we design computers we do the same thing but from the top down.

This is interesting. Thanks. I'll think about it for a while.

eta: I think the idea that the 'meaning' is built into neurons is important.

 

[Ichneumonwasp]

This is interesting. Thanks. I'll think about it for a while.

Yeah, sorry, like with the discussion over EEGs there are certain things that I think are self-evident because I deal with them all the time, which is one of my big problems with communication. One of the things that I think is self-evident but which probably doesn't come out in what I wrote above is that location (or rather how things are linked together) has meaning in the CNS. So a particular input to a neuron is not just any input, but it is an input that already comes with meaning attached to it; though a lot of the meaning is trained into the brain through experience (this is a bit of a complex topic but the basic issues can be covered fairly quickly if anyone cares). This is simply a feature of our nervous system that we inherited through natural selection. I can't think of any other good examples from nature that have all of these features attached to them.

With computers we are forced to do everything from the top down because to try to make all of this evolve from the bottom up would simply take too long and we wouldn't be able to control the outcome very easily. Since computers are our tools we always want to control the outcome, so we design them from the top down and supply meaning. The bottom up design from nature supplies some bits of meaning. We get other bits from a sort of top-down approach -- learning, but with bottom up constraints.

 

[AlBell]

It is based in the constraints that are already put in place within the system of the neuron. An action potential already has a meaning built into it by natural selection -- move the flow of information along. Water over flowing a bucket does not unless someone decides to impart meaning to it. The difference between the nervous system, as I have mentioned before, is that its simple 'meaning' (and that is way too loaded a word to use here) is built in from the bottom up. When we design computers we do the same thing but from the top down.

This is interesting. Thanks. I'll think about it for a while.

eta: I think the idea that the 'meaning' is built into neurons is important.

To me the top down design is a major part of the problem achieving 'consciousness' in a machine. Don't the current proponents seem to ignore that exact specifications of all the interacting parts will be required?

"At planck scale" seems meaningless until that design spec is understood (as does any programming effort).

 

[PixyMisa]

It is not possible to set an experiment to define what is conciousness.

We all know we share different conciousness because we are the experiment itself.

Conciousness is not part of the time and space dimensions, therefore it cannot be fixed to be identified.

It is Ouroborus, the cycle which never ends.

http://www.ouroborus.dk/

No scientific method will ever provide the delimitations of conciousness...

Conciousness is the experiment and science is the result.

You do know that there's no prize for "Least coherent post in thread"?

 

[PixyMisa]

I’m clipping out the areas where we seem to be in agreement.

The existence of constraints and rules are not the issue regarding computation. It is the need for an interpretation of the resulting output. As you note above, rock piles and neurons are not functionally different. Both have constraints and follow rules. How can you distinguish the ‘computation’ done by neurons and rock piles without requiring an observer to extract meaning from one but not the other?
No one is arguing this point. The question I’m asking is: How can you distinguish the ‘computation’ done by neurons, rock piles or even computers without requiring an observer to extract meaning from one but not the other?

Switching.

 

[PixyMisa]

This is one of Westprogs ploys to avoid issues, you can define it all you want.

Ask him why he uses a computer rather than a rock to connect to the internet, and enjoy the silence.

 

[rocketdodger]

To be honest, I was thinking of neurons, which are part of a conscious observer. Would you say that the output of a single neuron has no more of an 'observer' than the normal environmental surroundings of a pile of rock? - i.e. atomic particles bouncing around constantly? After all, we don't ever observe such outputs at a conscious level outside of a lab. Or would you say that being a part of conscious observer automatically bestows such an observer to all parts of such nervous systems?

If the first, then you haven't distinguished computation with neurons from computation with piles of rocks. If the second, then the distinction is based on the presence of a conscious observer, which is what I am saying seems to be required.



How is it that the actions and/or outputs of a falling pile of rocks are computationally different from the actions and/or outputs of a your hypothetical robot world with no humans?

It seems to me that either you accept that computation requires a conscious observer to extract meaning - i.e. neurons compute but a pile of rocks doesn't or conclude that functioning robots* do the same sort of 'computation' that falling rocks do in a world with no conscious beings.

* For the sake of this hypothetical argument, all conscious robots were also eliminated by whatever tragedy struck all conscious life forms dead in this hypothetical world.

No, there is no conscious observer and no, it is not the same as a pile of rocks.

The observer(s) of a neuron firing are other neurons, or muscle tissue, or anything else that radically changes its behavior based on said neuron firing. I wouldn't consider worms conscious like humans yet their neurons compute just the same as ours.

A pile of rocks might exhibit some of the behavior of a neuron -- once. And such behavior might influence the behavior of other systems in a similar fashion, E.G. if one pile of rocks slid down the hill and triggered another pile that fell in turn. To be honest, I could construct a series of dominoes that would carry out any calculation a series of transistors could, so if the rocks were the right size and configuration anything is possible.

But the similarity stops there, because you won't find many such piles of rocks in series and you certainly will never find piles of rocks that return to a different state over time on their own. Contrast this with neurons and transistors, which are hooked up to other neurons or transistors, and can return to their unexcited state without any outside help, etc.

So I am not saying a bunch of falling rocks never carry out a computation. I am saying that neurons always carry out computation, since other neurons always do something with the excitation. Life isn't different from non-life because only life features computations -- that isn't anything I have ever said. I said from day one that life is different from non-life because only life features computations that cascade upon the results of other calculations in a serial way that can and does lead to a longer existence of life.

Think about this -- life has existed for how long on Earth? Lets use a conservative estimate and say 3 billion years. Assuming abiogenesis didn't happen too many times, that means every single cell on Earth has been around for 3 billion years. How could that happen? How is it possible for a system of particles to keep itself around in a relatively similar form for so long, despite all the things that could have destroyed it -- especially since the particles that compose the system are constantly replaced? Really, how is it possible?

The only way it could be possible is if such a system was able to change its behavior based on the state of the environment around it so that it could avoid the natural destructive forces of the universe. Period. How can a system change its behavior like that? There is only one way -- computation. Being able to determine which partition the state of the universe is currently in, in order to act accordingly.

Everything *can* compute. Computation *can* happen anywhere, if the constraints are in place. But life and the things life builds feature far more coordinated computation than anything else in the universe. Because only life has figured out how to use coordinated computation to make itself last longer, and very lately, in order to get parts of the universe to do things we want.

Understand?

 

[PixyMisa]

The idea that you need a conscious observer for computation to be meaningful boils down to:

Computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because computation can't produce consciousness because...

And so on for 1919 posts.