Explain consciousness to the layman.

[dlorde]

Contradictory?

Yes, contradictory.

Computation in a living brain = consciousness,
Computation in an inanimate machine = no consciousness.

As I said, computation is substrate independent, so you are once again simply asserting that life can be consciousness, the non-living can't be conscious. I'm asking you for your reasons, your justification for this assertion.

Perhaps you will now provide a definition?

My minimal definition would be 'a dynamic self-organising, self-sustaining structured system that responds coherently to its environment'. That could be extended to include 'reproduction [with variation]', to bring it closer to biological life as we know it.

Biological life is generally characterised by homeostasis, reproduction, metabolism, organization (structure), responsiveness, adaptation, and growth.

Are you of the opinion that consciousness will result from running a suitable algorithm in an inanimate machine?

Yes. I think a suitable algorithm running in a suitable machine could produce a conscious process. I think that is, in principle, what the brain does. In my opinion the animate/inanimate distinction is a matter of convenience, and is irrelevant to the computational nature of consciousness.

 

[westprog]

Ok but you said it's motion can be described in physical terms. If the cursor can't, how can its motion ?

You can describe what happens when a cursor moves across a screen in physical terms. Those terms do not involve the cursor as a physical object, since there is no physical object there. However, we can measure the change of state of the pixels, measure the time at which they change, and hence calculate an apparent change of position in a given time, a velocity, and so on. Does that equate to motion? Well, playing Angry Birds on an iPad doesn't involve motion in the same sense as throwing stuffed birds at a pile of wooden blocks.

How about DNA ?

How about any chemical process? Or any physical process at all?

Give a nailed-down, precise definition of computation in physical terms that isn't impossibly narrow or broad, and I'll give it a look. Just giving examples and saying "this is computation, this isn't" won't do.

 

[dlorde]

But if Data stepped into a transfer booth and somehow the original was preserved, would he off himself.

Being conscious does not necessarily mean you have the innate urge for self-preservation that evolution supplies. There are some individuals that have no fear and no drive for self-preservation due to brain damage, yet are conscious. In Data's case, he would presumably have programmed 'instincts' for self-preservation based on something like Asimov's laws of robotics.

 

[westprog]

Magic of course!

To insist that any possible theory of consciousness that isn't computational must be magic is a... what's the words... almost got it... oh, yes, it's a

False Dichotomy.

The intentionality in computer programs comes from the person who put them there. Computers executing programs have no intentionality, any more than any other physical event.

Attributing human characteristics to inanimate objects is something common to children and primitive peoples. Asserting that a computer program "wants" to calculate a list of prime numbers, or is "confused" by incorrectly formatted data is like the belief that Zeus throws down the thunderbolt or that Teddy likes to be hugged.

Claims that humans get intentionality from computer programs are exactly backwards.

 

[westprog]

What about that poor old wasp, caught in a behavioural loop by an inflexible contextual algorithm ?

What about that rock, driven to suicide by that lure of gravity?

 

[westprog]

I don't think that's a significant distinction in this context - computers and algorithms don't have to be digital and discrete in operation. The first computers were analogue systems, using cams or fluids, e.g. gun aiming systems for battleships or the MONIAC (Philips Machine) that used water flows to model the economy.

That's a very different form of computing to the kind described by the Turing Machine model.

 

[dlorde]

What about that rock, driven to suicide by that lure of gravity?

Whut?

 

[westprog]

I can't believe we are bickering about what an algorithm is in a thread about the nature of consciousness. It's a very well defined term. Can we agree to it's standard definition? I like this one:



The internal operations of brain cells are algorithms executed by chemical reactions, as are the operations of the connections between brain cells. Trillions of these working at the same time can be fairly accepted as a humongously complicated algorithm in which consciousness emerges.

If you disagree, explain how it can't.

If there is a claim that an algorithm produces consciousness as a physical effect, then "executing an algorithm" needs to be defined in physical terms. An "instruction" is not a physical quantity.

If the claim is that executing an algorithm to produce consciousness is not a physical effect, then an anthropocentric or mathematical definition might suffice. However, a claim that consciousness is non-physical in nature doesn't sit well with the claim that all the competing theories are magical.

 

[dlorde]

That's a very different form of computing to the kind described by the Turing Machine model.

The expression and implementation is different. The mathematical algorithms expressed in physical form by those analog computers can also be expressed in a form that Turing-equivalent machines can process. The digital computers that replaced the analog gun aiming computers did just that (with bells on). I don't believe it was thought practical to do the same with the MONIAC, but it could have been done.

In the same way, artificial neural circuits can precisely emulate the processing functions of biological neural circuits.

 

[westprog]

Whut?

You can claim that arbitrary physical processes are algorithmic in nature, with intention and agency, but until you can define precisely what makes one physical process the execution of an algorithm, and another not, it's just a series of Just So stories. Attributing agency to physical processes is, as a general principle, an unhelpful way to look at them. It doesn't tell us anything we don't already know about why things happen.

I know that Dennet has decided to turn against the direction of science for some hundreds of years of science and embrace anthropomorphism and the "intentional stance". I do not yet consider that position justified.

 

[dlorde]

You can claim that arbitrary physical processes are algorithmic in nature, with intention and agency, but until you can define precisely what makes one physical process the execution of an algorithm, and another not, it's just a series of Just So stories. Attributing agency to physical processes is, as a general principle, an unhelpful way to look at them. It doesn't tell us anything we don't already know about why things happen.

I don't see what that has to do with suggesting the wasp's inflexible behaviour pattern appears to have an algorithmic basis. The wasp is clearly an agent executing a specific, repetitive, and inflexible sequence of actions in an algorithmic fashion. I'm not quite sure where you feel intent comes into it - the behaviour described shows a marked lack of intent.

Are you saying the wasp is simply a collection of arbitrary physical processes that happen to give rise to a sequence of activities that look algorithmic? or what?

 

[rocketdodger]

I don't see what that has to do with suggesting the wasp's inflexible behaviour pattern appears to have an algorithmic basis. The wasp is clearly an agent executing a specific, repetitive, and inflexible sequence of actions in an algorithmic fashion. I'm not quite sure where you feel intent comes into it - the behaviour described shows a marked lack of intent.

Are you saying the wasp is simply a collection of arbitrary physical processes that happen to give rise to a sequence of activities that look algorithmic? or what?

No, he is saying that the sequence in the falling rock is just as non-arbitrary as those in the wasp.

Which I don't see an issue with -- if the behavior being described is a falling rock breaking when it lands, then the sequence is very non-arbitrary:
1) fall
2) fall faster
3) hit ground
4) break

So what? A falling rock breaking is an example of an algorithm. Anyone have a problem with that? I don't.

 

[Complexity]

That's a very different form of computing to the kind described by the Turing Machine model.

So?

 

[Complexity]

No, he is saying that the sequence in the falling rock is just as non-arbitrary as those in the wasp.

Which I don't see an issue with -- if the behavior being described is a falling rock breaking when it lands, then the sequence is very non-arbitrary:
1) fall
2) fall faster
3) hit ground
4) break

So what? A falling rock breaking is an example of an algorithm. Anyone have a problem with that? I don't.

I do. Your language is imprecise (a 'falling rock breaking is an example of an algorithm' - no, it is not) and I don't think what you meant to say helps matters at all.

This entire thread has gone off the deep end (of course). Not your fault -any thread on this set of topics is doomed.

 

[westprog]

I don't see what that has to do with suggesting the wasp's inflexible behaviour pattern appears to have an algorithmic basis. The wasp is clearly an agent executing a specific, repetitive, and inflexible sequence of actions in an algorithmic fashion. I'm not quite sure where you feel intent comes into it - the behaviour described shows a marked lack of intent.

Are you saying the wasp is simply a collection of arbitrary physical processes that happen to give rise to a sequence of activities that look algorithmic? or what?

Sorry, I might have mistaken your intent.

It's certainly the case that the behaviour of a wasp can be described by an algorithmic process, but the same might apply to a lot of processes - a ball randomly rolling around a maze, for example. An algorithm simply describes what will happen if certain states are encountered.

I don't think it's been mentioned that an algorithm is AFIAA a far more general and informal concept than that of the Turing machine.

 

[rocketdodger]

I do. Your language is imprecise (a 'falling rock breaking is an example of an algorithm' - no, it is not) and I don't think what you meant to say helps matters at all.

This entire thread has gone off the deep end (of course). Not your fault -any thread on this set of topics is doomed.

Why isn't that an example of a "falling rock breaking" algorithm?

 

[dlorde]

A falling rock breaking is an example of an algorithm. Anyone have a problem with that? I don't.

OIC. Not so sure about that - it seems qualitatively different, in that it's an implicit algorithm.

I'm suggesting the wasp is running an algorithm as an explicit sequence of instructions on a general purpose computational control system (its brain). This general purpose computational system can, and does, run a number of other algorithms to achieve different behavioural goals.

 

[Leumas]

The problem with saying that animal behavior is an algorithm is that you open yourself to the naive question of "WHO designed the algorithm?"

Much like the confusion among theists about the word "Theory". Much like they keep asking "WHO set the Laws of Nature".


Just like the Laws of nature are not LAWS but descriptions of nature's behavior so I say that the algorithms of animal behavior are not ALGORITHMS but descriptions of the behavioral manifestations.


So if you want to define algorithms in a more general way to include animal behavior then fine....but the standard understood definition implies a DESIGNER for the procedure....it implies INTENTIONS and PURPOSE for the process.

Unless we want to have the same confusing effect upon theists as when we say the "Laws Of Nature" and the "Theory Of Relativity", then we need to move away from that word due to the "designer" connotations associated with it.

 

[rocketdodger]

OIC. Not so sure about that - it seems qualitatively different, in that it's an implicit algorithm.

I'm suggesting the wasp is running an algorithm as an explicit sequence of instructions on a general purpose computational control system (its brain). This general purpose computational system can, and does, run a number of other algorithms to achieve different behavioural goals.

Yes, correct.

But now we have to define what "explicit" versus "implicit" means in terms of particle behavior.

I don't think that is hard, I think you could come up with a definition along the lines of running an algorithm "explicitly" implies the system can run a number of different algorithms, whereas "implicit" means only one.

Certainly the system of a < rock + gravity + the ground > can only run the "rock falling and breaking" algorithm, whereas the system of <wasp> can run a vast number of algorithms.

Except, westprog will point out that there are an infinite number of specific "rock falling and breaking" algorithms, so actually there aren't "more" of them. My response to this would be instead to say then that an algorithm is an equivalence class of a series of steps in a process, such that the overall process leads to a similar end behavior of the system ( or something like that ).

Then it is clear that the <rock + gravity + ground> system can only run one class of steps, meaning one algorithm, while the <wasp> system can still run many.

Finally, westprog is going to say that we are still defining "class" according to our whim, and that the universe doesn't class things. That is bollocks, however, because obviously a wasp that moves out of the way of a frog tongue leads to a different class of behavior from the perspective of both the frog and wasp -- who are not human, unless westprog wants to make even more absurd arguments. In particular, if the frog eats the wasp, the frog will survive and the wasp will not. If the wasp evades, the frog will starve and the wasp will be happy. I am pretty sure the future behavior of both of them changes significantly depending on the outcome of the <wasp> running the evasion algorithm.

 

[Complexity]

So if you want to define algorithms in a more general way to include animal behavior then fine....but the standard understood definition implies a DESIGNER for the procedure....it implies INTENTIONS and PURPOSE for the process.

Unless we want to have the same confusing effect upon theists as when we say the "Laws Of Nature" and the "Theory Of Relativity", then we need to move away from that word due to the "designer" connotations associated with it.

I don't agree.

Sure, creationists would make that assumption, but they are fools, and we should not cater to their willful ignorance. They would deliberately 'misunderstand' anything that we say, regardless of how idiot-proof we try to make it.

How are you ever going to explain to people that DNA contains much data that can be usefully interpreted as to when and how to make proteins without anyone having put that information there if you take this stance?