On Consciousness

[tensordyne]

It is a physical aspect of the Universe - an emergent property of suitably organised computer systems.

A very well put recapitulation of the Hard-AI school of thought. I do not buy it.

Suitably organized systems will have emergent behavior sure, but consciousness is not about behavior. Oh, and emergent system dynamics is not even necessarily about a physical aspect of the universe, it is about system dynamics.

 

[quarky]

At the small risk of going off-topic, here's a tale from my studies that sort of shook me:
(Disclaimer! I'm not advocating illegal drug use!)

When mdma was first getting popular (not when it was invented, btw) I was present for many people's first time experience with that drug.

It amazed me that something as abstract and human as empathy, could be put in a capsule. You could nearly set your watch to it; 45 minutes after ingestion, people would begin to apologize for the slimmest past offenses towards the others in the group. Then, admission of love; sweetness; gentle affection.

Seeing this repeatedly somewhat 'cheapened' all this warm-hearted expression.
It was just bio-chem. It wears off.
On the other hand, being aware of that phenomena; the cheapness of basic human emotions, like empathy, leads to a new level of empathy.

More extreme, is dealing with an Alzheimer's patient, and facing the implications of the memory of sequential events; and how the unraveling of that can make a stranger out of someone that you once knew.

Consciousness is tenuous and delicate, and it comes in degrees and it leaves in degrees.
To embrace the mystery of what it is; to see what it isn't...these are features of consciousness.

I expect some mockery for expressing this, and I'm good with that.

 

[Pulvinar]

Absolutely, I only brought up the Nazi neurological scientist to point out directly what would be needed if nothing could get in the way.

I see. "Conversations with Neil's Brain" is an excellent book, BTW; hope you've had a chance to read it.

Not so, computing is not a physical aspect of the universe, it is a computational aspect of the universe. There are many physical systems that will perform the exact same computations. Computation is not about density, it is not about mass, it is not about spin angular momentum. Computation is about things like The Church-Turing thesis that hard-AI, connectionist, etc schools take much too far.

So by "physical" you mean "physical science", not "material". That's what confused us. To me computation is about calculation, information processing, logic, in the form of patterns supervening on the physical world. Don't remember ever using the Church-Turing thesis in a machine to do a computation.

 

[PixyMisa]

Ooh ! List, please !!

A good way to think of this would be to examine Searle's Chinese Room, supposedly capable of conversation in Chinese.

Given the combinatorial number of possible sentences, a room that did so through simple lookup tables printed in actual books would be larger than the observable Universe and take a trillion years to respond to a question.

That's the unaware solution. An aware solution would be somewhat more plausible, but having (by definition) no model of its internal state, would have the same problem as before in answering questions about its internal state.

Given the parameters of the thought experiment, Searle's Chinese Room not only knows Chinese, it must also be conscious.

Another example I've mentioned before is the sphex wasp, which will robotically repeat its behaviour if interrupted in certain tasks. Lacking self-awareness, it cannot "jump out of the system" and realise that its task has been rendered futile by the neuroscientist's experimental intervention. Try that with a mouse and it will swiftly adapt.

 

[PixyMisa]

It is a closed loop. The air triggers the thermostat which runs the A/C which cools the air which triggers the thermostat.

But it has no model of itself.

If that's still not enough, what's a good minimally conscious case using electronics? Feedback loops of all types are a common and powerful engineering tool, so surely there must be some household appliance we can point to and say "this thing is very dimly conscious."

You don't really get consciousness in anything that doesn't have a computer in it. A simple 8-bit microcontroller would suffice, if it runs a watchdog routine that monitors its own state (which is a very common technique).

You could define a minimally conscious circuit in transistors; you'd need 24 for flip-flops to store a model of the external world, 24 for self-modelling, and then a bunch of NAND gates to build your attentional logic. Maybe a hundred or so. And that's for a circuit designed to be minimally conscious, not to do anything useful.

 

[Pulvinar]

Again, there are two main possibilities (unless you or anyone else can think of others) as to a scientific-type basis of consciousness: computation, physics. Maybe you can combine them. A suitably programmed computer can to whatever degree you want simulate all sorts of things. SO WHAT!

Physics is physics and computer science is computer science. The two studies can help each other in many ways but there is the real world and there is the world of abstraction; it is not a good idea to confuse the two. If I make a simulated world it will follow the rules of the simulated world.

From this simulation I can test various predictions, but I can not come up with new laws of physics from it since I would have to program them in some how. Even if I did program them in, ultimately I would have to test the resultant predictions against real world. The two domains are different.

I don't understand why you bring up simulation here. Surely you don't think that all computation is simulation. I believe it would help if you could answer my question: what knowledge do you have about consciousness that isn't itself a piece of information?

 

[PixyMisa]

Again, there are two main possibilities (unless you or anyone else can think of others) as to a scientific-type basis of consciousness: computation, physics.

No, only one: Both.

Consciousness is computational, and computation follows the rules of physics (like everything else).

Don't get hung up on the abstractions. Consciousness is something that really happens, not some purely abstract notion.

 

[tensordyne]

I don't understand why you bring up simulation here. Surely you don't think that all computation is simulation. I believe it would help if you could answer my question: what knowledge do you have about consciousness that isn't itself a piece of information?

As far as simulation goes in terms of the context, yes, all simulation is computational. Consciousness is not information, it is what it is like to be something. Consciousness is seeing red, feeling soft feathers, etc. If you do not get that I can not help you.

 

[tensordyne]

No, only one: Both.

Consciousness is computational, and computation follows the rules of physics (like everything else).

Don't get hung up on the abstractions. Consciousness is something that really happens, not some purely abstract notion.

Don't get hung up on abstractions! That is all hard-AI'ers believe in is abstraction. I agree, consciousness is something that really happens (or at least, that is what seems to make the best sense).

Computation is NOT about physics. There are lots of ways to harness the rules of physics to perform computations that are the same with only the physics being used being different. When X (computation) remains the same but is done in one of several ways Y (various ways to use physics), then X is not about Y.

Plus, physics can conceptually do things that computation can not. We could have a universe, where, for whatever reason, only water is conscious, or... That has to do with the physics that universe follows. I do not like to point out hypothetical concepts such as this but I do not know how to get across why physics can do things computation can not.

If you say the last example is contrived, my rhetorical question is, is it not likely we live in a universe where we do not know all of the physics and what it does?

 

[PixyMisa]

Don't get hung up on abstractions! That is all hard-AI'ers believe in is abstraction.

Not at all. Computation is a physical process.

Computation is NOT about physics. There are lots of ways to harness the rules of physics to perform computations that are the same with only the physics being used being different. When X (computation) remains the same but is done in one of several ways Y (various ways to use physics), then X is not about Y.

Yes, there are many ways to build a general-purpose computer. All of them are computationally equivalent, and all of them follow the laws of physics.

Plus, physics can conceptually do things that computation can not. We could have a universe, where, for whatever reason, only water is conscious

No, that's not even a coherent statement. Water is water. To be conscious, it would have to possess additional properties that would make it not water.

If you say the last example is contrived, my rhetorical question is, is it not likely we live in a universe where we do not know all of the physics and what it does?

We know that there are gaps in our knowledge. Specifically, Relativity and Quantum Mechanics do not mesh.

We also know that consciousness is a computational process, and computational processes are a question of physics.

 

[Belz...]

It could be about information processing.

And how is that not physical ?

 

[Belz...]

Surely you don't think that all computation is simulation.

As far as simulation goes in terms of the context, yes, all simulation is computational.

Nice reversal there, Tenso.

 

[tsig]

I have a food processor that I never use. Occasionally, I feel sorry for it. If I even plugged it in once in awhile, it might achieve a dim awareness.

I have a weed-eater. It channels Satan as soon as I pull the cord.

This is a silly debate; one of no consequence.
Pixy evidently isn't even going to argue with my points.The thread's consciousness is dim. Too dim.

delete

 

[tsig]

At the small risk of going off-topic, here's a tale from my studies that sort of shook me:
(Disclaimer! I'm not advocating illegal drug use!)

When mdma was first getting popular (not when it was invented, btw) I was present for many people's first time experience with that drug.

It amazed me that something as abstract and human as empathy, could be put in a capsule. You could nearly set your watch to it; 45 minutes after ingestion, people would begin to apologize for the slimmest past offenses towards the others in the group. Then, admission of love; sweetness; gentle affection.

Seeing this repeatedly somewhat 'cheapened' all this warm-hearted expression.
It was just bio-chem. It wears off.
On the other hand, being aware of that phenomena; the cheapness of basic human emotions, like empathy, leads to a new level of empathy.

More extreme, is dealing with an Alzheimer's patient, and facing the implications of the memory of sequential events; and how the unraveling of that can make a stranger out of someone that you once knew.

Consciousness is tenuous and delicate, and it comes in degrees and it leaves in degrees.
To embrace the mystery of what it is; to see what it isn't...these are features of consciousness.

I expect some mockery for expressing this, and I'm good with that.

From what you just posted there's no mystery just chemicals.

 

[Dancing David]

If you don't know you're conscious it's probably not worth discussing anything with you.

As a member of the P-Zombie Party, I ask you seriously,

and how do you know that you are conscious?

Or do you just believe that you meet a definition, like the biomedical one?

 

[Dancing David]

You know, the most interesting thing about this is that there are people who are saying you are missing a concept and yet you continue just to repeat the shibboleths of complexity, behaviorism and so on. Maybe you should try and figure out what the hell they are talking about? Even if they are wrong, try and get in their head and show why, or something.

What exactly, spelled out is that concept?

Instead of using the vague "what the hell they are talking about", explain to us exactly what the hell they are talking about.

I hope you are not going to resort to qualia and I am very hopeful it is something else.

 

[Dancing David]

Nope, attention precedes emotion.

I happen to disagree, but I am curious as to your reasoning. I shall hold mine for the nonce.

 

[Dancing David]

Behavior != Consciousness

That is a nice statement, and how would you explain that? What evidence do you have for that. And lastly, then how do you know you are conscious?

What is conscious, that is not defined by a behavior attributed to consciousness?

 

[Dancing David]

"I define consciousness to be XXXX". Fill in XXXX. Hopefully that is not too restrictive.

Try defining consciousness without reference to behavior. Hopefully that is not too restrictive.

 

[Dancing David]

For the record, I have no problem with the idea of artificial consciousness or consciousness in other types of entities. I just think, for epistemology reasons, we should figure out consciousness in ourselves first (and as much as is possible both ethically and scientifically in other animals).

I just think, for epistemology reasons, we should figure out gravity in ourselves first...



epistemology matters, why?