On Consciousness

[rocketdodger]

.
Unless I'm mistaken, the laws of physics are THE laws the universe runs by.
Our brain/imagination may be able to conceive of things outside those laws, but we have no way to get around them, other than by thought, which is all we do about any thwarting or work arounds.

Err, you totally missed my point, which is:

You claim that the fact that all programs can be traced back to code written by a programmer, even programs that write programs, somehow implies that the only significant causal events in the life of a program are the events of the programmer writing the code.

I understand, and even agree with this, for code like "if Ford Automotive stock is over $35.94 a share, buy 20 shares."

However for code like " if a > b return true else return false " that is less clear. In particular, if I as a programmer wrote that code, and I don't know exactly what a and b are for each instance of execution, then I don't see how the most significant causal event was me writing the code. I didn't even know what values a and b would end up taking, so how on Earth is my writing the code the significant event?

That is tantamount to saying that since the particles inside me follow the physical laws of the universe, and supposing those laws have some origin, then the only significant causal events related to me are whatever happened billions of years ago that led to those laws existing.

I disagree with that, because whatever led to those laws certainly couldn't imagine that I would be here typing this post. Even if the universe is 100% deterministic, the significant causal events related to me are what is happening now.

 

[I Ratant]

I can't see a computer, with the rules it must be assembled with, coming up with:
"It was a dark and stormy night"... or "Call me Ismael"... or "In the beginning..",
particularly the last which has what a computer would recognize as physical absurdities, much less plots that have been pre-figured in the author's mind to go in a specific direction, which may or may not be possible in a physical world.
A computer that might do that would have so few constraints its output would resemble that herd of monkeys more often than not.
Dawkins got his program to find a particular phrase in Shakespeare, but he had to direct the computer to that goal, external to the programming of the phrase generator.

 

[dlorde]

< marble is near dent > ----> given enough time -----> < marble is settled in dent >

contrast this with

< marble is not near dent > -----> given enough time -----> < marble is not near dent >

Thanks for an exceptionally clear explanation of attractors.

Takes me back to the heady days when fractals, chaos, and strange attractors were new and exciting

 

[dlorde]

... I'd have to see a direct quote from Markram, tho, to accept that he would posit that the simulator would become conscious.

Well, the Seed magazine article Pixy linked to says:

“There is nothing inherently mysterious about the mind or anything it makes,” Markram says. “Consciousness is just a massive amount of information being exchanged by trillions of brain cells. If you can precisely model that information, then I don’t know why you wouldn’t be able to generate a conscious mind.”

It seems Markram expects a human scale Blue Brain to be conscious, as the article continues:

“I think it will be just as interesting, perhaps even more interesting, if we can’t create a conscious computer,” Markram says. “Then the question will be: ‘What are we missing? Why is this not enough?’”

 

[rocketdodger]

I can't see a computer, with the rules it must be assembled with, coming up with:
"It was a dark and stormy night"... or "Call me Ismael"... or "In the beginning..",
particularly the last which has what a computer would recognize as physical absurdities, much less plots that have been pre-figured in the author's mind to go in a specific direction, which may or may not be possible in a physical world.
A computer that might do that would have so few constraints its output would resemble that herd of monkeys more often than not.
Dawkins got his program to find a particular phrase in Shakespeare, but he had to direct the computer to that goal, external to the programming of the phrase generator.

Well, let me ask you some questions.

Have you ever programmed?
Have you ever programmed any kind of A.I.?
Have you ever programmed any kind of A.I. that learns?
Have you ever programmed any kind of A.I. that learns using artificial neural networks?

If you answer "no" to any of those, then concluding that a computer can't do something the human brain can seems to be a bit premature to me.

If you answer "yes" to all of them, then I would certainly be interested in a discussion involving the results you got that lead to to this conclusion of yours.

 

[rocketdodger]

Well, the Seed magazine article Pixy linked to says:

It seems Markram expects a human scale Blue Brain to be conscious, as the article continues:

I doubt piggy considers what Markram says to be evidence of what Markram means...

 

[dlorde]

So when you say "computational equivalence", what exactly do you mean?

I mean it in a pragmatic Wolfram sense; he says

... most systems are computationally equivalent. For example, the workings of the human brain or the evolution of weather systems can, in principle, compute the same things as a computer.

Clearly we are not in a position to use weather systems to compute anything non-trivial, but if we can use a digital computer to compute the same things as a working human brain (and the Blue Brain project suggests we might), and if it is the computational workings of the human brain that give rise to consciousness, then, as Markram says, "I don’t know why you wouldn’t be able to generate a conscious mind".

 

[Piggy]

In answer to your question of evidence:

Consciousness is the ability to think about your own thoughts. Self-awareness. Computer programs do this via the technique of reflection. Not all programs do this, but many complex and/or real-time systems do, because it's as useful a technique for computers as it is for animals.

Feeling is either awareness of of an abstract representation of sensory data, or the conscious reflection of the same. So again, many computers do this.

Thinking is just processing data, so all computers do this by definition.

That's simply not a definition of consciousness that works with brains, which are the only objects we know of that produce conscious experience.

Consciousness may (or may not) be required to "think about your own thoughts", but that process is not what consciousness is.

Consciousness cranks up when you wake up, or when you dream, and it's not happening when you're in deep sleep or under general anesthesia. It is marked by the production (by your brain) of colors, sounds, odors, pain, proprioception, and other phenomena which are unique to it.

We can group these under the general category of "experiences", and the experiences of other species will overlap with ours to varying degrees, presumably moreso for gorillas than for cats or crows.

That's the phenomenon -- or cluster of phenomena -- that needs to be explained.

Designing and building computers which simply refer to data sets describing their own states is not relevant to the study of consciousness, because no one has any reason to believe (and every reason to doubt) that this will result in a machine which behaves in such a way that it performs experiences of any sort.

And in fact, it's not true that thinking is just processing data... at least, not conscious thinking, anyway.

The result of processing data is, necessarily, data. And data must be read to be of use (or even to be data).

There is no one there to read our brains, so discussion of "data" there is necessarily metaphorical to some extent, although it is immensely useful!

But as I've repeatedly said, this way of looking at consciousness which you describe is utterly unproductive in the lab and field, so it is not used. Progress is being made, however, using the functional definition.

 

[dlorde]

We think of "hallucinations" as deviations from the accurate perception of reality.
...
That's what I mean when I say that our everyday experience is a "normative hallucination".

OK; that makes sense, unlike 'normative non-normative experience'

So onto the other part, "the brain produces colors and brightness"....

Colors, smells, pain, these exist nowhere but in the brain.

Where else could these things possibly exist?

I'm just emphasising the need for care and clarity here - the brain only produces red if you stab it. It experiences the sensation or perception of red[ness] when a red object is seen (or when the relevant areas of the visual cortex are active).

 

[Piggy]

I'm sure you'll agree that a truck and a mind are two very different things and that comparing them like you did is not entirely relevant ?

A rock and a hammer are two very different things, but I wouldn't want to get hit with either one.

For our purposes, the truck and the brain are enough alike to make the point here. And that's because if you want real-time behavior in actual spacetime, then a computer can be part of making that happen, but you can't just program that.

You can't program the real-world behavior of a truck... a purely programmed truck can't get you to work. And a purely programmed pancreas can't function in your body.

A computer might be part of an artificial pancreas, or a working model of my truck, but you can't program real-world behavior and have it occur with only enough hardware to "run the logic" and none to carry it out.

And ditto for a brain. Unless you've got some reason for me to believe that this particular glob of matter operates under some very special rules.

Everything in the cluster of behaviors we call consciousness -- including colors, sounds, smells, pain, and so forth... which are all behaviors, and not properties of things out there in the world -- is, as far as anyone can tell, the result of the normal physical behavior of your brain, just like everything else your body does.

Just because we haven't figured out what's going on yet, we are not therefore relieved from our obligation not to jump to the conclusion that something other than brute physics is at work here.

We still assume that the phenomenon of consciousness is caused by the same physics that underlie all other phenomena we observe.

Yeah but using that definition makes calling your PC "computer" also useless.

Damn right. But that's how Turing saw it, you know.

If you want to use that view, which is Wolfram's view, then everything computes.

But if you're talking about symbolic information processing, which is what my PC does, then you're talking about a system which requires an encoding and decoding mind, and there's no such service for what's going on in our brains.

There is as much symbolic information in our brains as there is in our fingernails.

My brain does calculations. Computers do too. My brain does comparisons. Computers do too. My brain interfaces with my sensors and peripherals. Computers do too. Aero fits well in your purse. Milk ? Why not ?

That's pretty damn weak. Are there any two objects that you couldn't get up a list of that caliber for?

 

[dlorde]

Again I ask, does anyone think feeling thinking conscious machines already exist?

I'm happy to accept that computers can feel and think. I don't think they're conscious yet because I don't accept Pixy's definition (self-referential information processing) as sufficient; necessary, but not sufficient. Having said that, it's hard pin down anything more than that as a requirement, so I can understand why he takes such a 'minimalist' approach.

 

[Piggy]

I'm just emphasising the need for care and clarity here - the brain only produces red if you stab it. It experiences the sensation or perception of red[ness] when a red object is seen (or when the relevant areas of the visual cortex are active).

No, if you stab my brain, my brain does not produce red. It stops producing red because I lose consciousness.

If you're looking at my brain when it gets stabbed, then your brain will produce (or more accurately, perform) red.

There is no such thing as "redness" for your brain to "experience"... there are no "red objects" for you to see.

In various circumstances, your brain will perform red. That's all.

Similarly, if you stick me in the arm with a knife, I feel pain, but the pain isn't in the knife. It isn't even in my arm. Pain doesn't, and can't, exist in either of those places.

The knife in my arm triggers a nervous response, which triggers my brain to perform pain.

Similarly, when light comes from certain sections of this screen in front of me, it triggers a nervous response, which triggers my brain to perform red.

But red doesn't exist in the screen any more than pain exists in the knife.

It makes no more sense to say the one than to say the other.

 

[Piggy]

Between you saying we know nothing about it, and Pixy saying we've got a firm grasp of it, there has to be a middle ground where reality lies.

No, there doesn't.

Nobody yet knows the mechanism, and we don't even have any firm theoretical basis for proposing one.

That's just the state of things right now.

I gave you direct citations from heavyweights, several as recent as 2011, stating frankly that this is the case.

The other side hasn't provided any citations of anybody.

 

[Piggy]

Because the brain is a computer.

So say ye.

3 things I ask....

Say clearly what you mean by "computer".

Cite your evidence that the brain is one.

Explain what this contributes to our understanding of consciousness.

I'd rather ask you why you believe that's a coherent question.

Sure.

If you're talking about making a machine conscious, which is to say making it hallucinate in some way, whether normative-waking or dream-state or what we'd commonly call a hallucination -- or any other state we may be unfamiliar with -- then you're talking about making it behave a certain way.

Conscious phenomena like colors and sounds and textures and such are real-spacetime behaviors. Yes, color is a bodily function.

To get real-spacetime behaviors of any sort out of a machine, you can't just use enough hardware to "run the logic" as you have said... you have to run the logic and perform the action.

This is true of machines, including computers, that we have now and it will be true of any machine, including computers, we ever build.

So far, neurobiological research has discovered no reason to impose any symbolic or informational or computational layer in between the physical-energetic behavior of the brain and the cluster of phenomena we call conscious awareness.

Progress is being made without any of that, although the concepts of computation, computational power, and information processing have greatly contributed to our understanding of how the brain works, to our ability to talk about the brain and consciousness, and our capacity to do research.

But still, at the end of the day, if you want a liver to do what a liver does in a real body -- be it a human one or a robot one -- you have to build a liver, you can't just program it.

All neurobiological research thus far has given no indication that the same is not true of the brain and all of its real-spacetime function, including consciousness.

So if you want to build a truck to get you somewhere, you can't program it, you have to build it. And by the same token, if you want to build a thing that actually has experiences, you can't just program it, you have to build it that way.

 

[dlorde]

Similarly, when light comes from certain sections of this screen in front of me, it triggers a nervous response, which triggers my brain to perform red.

I guess 'perform' will do for 'experience'. Makes more sense than 'produce'. It may sound pedantic, but how one thinks about these things makes a difference.

 

[Piggy]

This is why it isn't even worth discussing things with you, piggy.

Someone brings up good research.

You respond with "blah blah blah, I didn't read it, I don't need to, it wasn't an MRI and it wasn't done by someone with a "biologist" in their title, so it isn't relevant."

If you think it's relevant, then let's look at it.

I'll go back and pick it up again (you were right, I didn't finish the first page) and if you don't mind just tell me in a few sentences what was accomplished and how you think it's relevant to the study of consciousness.

And we'll pick it up from there.

 

[Piggy]

Pixy writes code for a living and one could say he even does research in computer science, and he thinks in terms of the most fundamental behavior that could be required for consciousness.

Piggy can't be bothered to even read research done by others, never mind perform his own, and he thinks in terms of every behavior that might be required for full human consciousness.

This is so crazy I can hardly believe it.

PixyMisa has an idiosyncratic definition of consciousness -- as I've shown -- which is not accepted by anyone in the field, so he has no frame of reference for thinking about what is or isn't required for consciousness. Writing code for a living makes as much difference as building cars.

And no, I don't do my own research, but I do read other people's research on consciousness. Research on AI is another thing entirely.

 

[Piggy]

Massimo Pigliucci^WP has argued that emulating a conscious brain in a computer is akin to emulating photosynthesis in a computer. You don't get sugar and you don't get consciousness.

Exactly.

At the end of the day, no matter how you slice it, it really is that simple.

 

[rocketdodger]

No, there doesn't.

Nobody yet knows the mechanism, and we don't even have any firm theoretical basis for proposing one.

That's just the state of things right now.

I gave you direct citations from heavyweights, several as recent as 2011, stating frankly that this is the case.

The other side hasn't provided any citations of anybody.

Saying that the mechanism is not yet known, and that we don't have a "firm" basis for proposing one, is a far cry from knowing nothing about it.

Furthermore I question why you perpetually consider people who do little besides write popular books "heavyweights."

I personally would consider the people doing the actual research the "heavyweights," if the term is even applicable, and I doubt they share such a conservative viewpoint because the very fact that they are doing research suggests they have some hypothesis to test at least.

 

[Piggy]

Well, the Seed magazine article Pixy linked to says:

It seems Markram expects a human scale Blue Brain to be conscious, as the article continues:

Hmm... those statements were sometime after Gazzaniga, so I'll have to excuse MG for now.

And I will count this as a cite (finally) for the computationalists.

I do find this statement odd, tho:

“Once we can model a brain, we should be able to model what every brain makes. We should be able to experience the experiences of another mind.”

There's a big and obvious problem with this idea.

Our brains use some of the same real-estate, not surprisingly, when we remember or imagine an event as we use when we experience that event.

That's why you'd never be able to describe what something smells like to a person who's never been able to smell. The bits of their brain that would need to be active to understand what you're saying -- because language doesn't work by "standing for" things, it works by activating brain cells -- just don't activate.

So no, looking inside the activity of a cat's brain isn't going to make us able to know how a cat experiences the world, because all we have are our own tools, which create a human experience.

We'd have to rebuild our own brains to warp our experience catward.

Think about birds, they have some sort of way of sensing magnetic fields. Let's just suppose, for the sake of speculation here, that they're consciously aware of that somehow... that changes in magnetic fields "seem like something" to them, like chemicals in our nose seem like smells, and air bouncing in our ears seems like sounds, and such.

Even if we figured out exactly what brain activity corresponded with their experience of magnetic fields, we have no way of currently imagining how we would answer the question "what does that seem like to them?".

So I'm skeptical of that view.

As is Markram, to some extent, it seems....

And yet, Markram is candid about the possibility of failure. He knows that he has no idea what will happen once the Blue Brain is scaled up. “I think it will be just as interesting, perhaps even more interesting, if we can’t create a conscious computer,” Markram says. “Then the question will be: ‘What are we missing? Why is this not enough?’”

I think the answer might be here:

“There is nothing inherently mysterious about the mind or anything it makes,” Markram says. “Consciousness is just a massive amount of information being exchanged by trillions of brain cells."

I agree with the first statement completely.

The second statement, though, has problems, unless you're talking about information in the Wolfram sense, and in that case it's trivial because so is everything else your brain does.

If there's a theory for how and why we get from particular configurations of information (independent of particular physical behaviors) and not from others to phenomena in the consciousness cluster such as yellow, mustiness, or the sound of thunder, I'd like to know what it is.

And I'd like to know why consciousness is more likely to depend on a supposed informational layer supported by the physical behavior rather than directly from the physical behavior itself.