My take on why indeed the study of consciousness may not be as simple

[!Kaggen]

As I mentioned, the collection of mental processes that we call "consciousness" are emergent properties at least at the level of brain structure. You don't get consciousness in a single atom or subatomic particles. Even when physicists ask questions about how a particle "knows" about the spin of a linked particle, the scare quotes are used to recognize that they're using the pathetic fallacy.

How many brain cells do we need before consciousness emerges?

 

[drkitten]

Are you prepared to concede that he is, in fact, a physicist? Indeed, a very eminent physicist?

Not in the slightest, no.

His work as an algebraist and geometer has been of tremendous use to physics, just as Kahneman's work (and Simon's) work in psychology has been of tremendous use to economists. This does not make Penrose a physicist any more than it makes Kahneman or Simon economists.

And I hold to this statement despite the fact that Kahneman and Simon have received Nobel prizes in economics. Because -- as they would be the first to admit -- their work is psychological, not economic.

 

[Dancing David]

who says some discovery's aren't

Now I did not mean that. I meant that translating the idiomatic meaning of pictorial art that is filetered through personal and cultural filters and then trying to interpret symbolms across time and culture is not wise.

You are safe in saying 'flying people can be interpreted as shamans','flying people can be interpreted as angels','flying people can be interpreted as being launched from a catapult' but you can only speculate as to the meaning of pictorial art unless you have a good contemporary understanding of the symbology. And even then it can be questionable.

 

[Dancing David]

It's called that by people who disagree with it. People who are neutral call it the Lucas-Penrose thesis. There are of course a number of versions - inherent in the fact that it's Lucas-Penrose, not just Lucas.

Do you mean the impossible QM effects? I will check.

Nope something else

The probably most prominent and most articulate argument why the whole field of AI would be doomed to failure is expressed in the so-called Lucas-Penrose argument, which can be summarized as follows: Since Gödel proved that in each sound formal system - which is strong enough to formulate arithmetic - there exists a formula which cannot be proved by the system (assumed the system is consistent), and since we (human beings) can see that such a formula must be true, human and machine reasoning must inevitably be different in nature, even in the restricted area of mathematical logic. This attributes to human mathematical reasoning a very particular role, which seems to go beyond rational thought. Note that it is not about general human behaviour, and not even about the process of how to find mathematical proofs (which is still only little understood), but just about the checking of (finite) mathematical arguments.

http://www.cs.bham.ac.uk/~mmk/papers/05-KI.html

What if the brain is analog and fuzzy?

 

[drkitten]

Do you mean the impossible QM effects? I will check.

Nope something else

http://www.cs.bham.ac.uk/~mmk/papers/05-KI.html

What if the brain is analog and fuzzy?

Hell, what if the brain is "unsound"?

Who would think that humans were capable of making mistakes in reasoning? Certainly not Penrose, which is one reason that he is "simply" wrong.

 

[westprog]

Yup. I'd accept that we can get advances from fields like psychiatry, neuroscience, even psychology before I would expect any insight into consciousness coming from physics.

Of course. There's so little known about consciousness that there's nothing much for physicists to work with.

The fact remains - for something to be completely understood, there needs to be a physical explanation. Go deep enough and it's all physics.

Somewhat of a derail. . .I understand the TV show NUMB3RS is pretty entertaining, but the first episode I watched turned me off completely because it made the same mistake people here are making. The hero was using some mathematical model of the criminal's behavior to predict what he would do next. At one point, the prediction was dead wrong. Aha! he says, I forgot to consider the observer effect. He then invokes the Heisenberg Uncertainty Principle and claims that any time you observe something you alter it. He then leaps to the conclusion that since the bad guy knows the cops are after him, he has to change the mathematical model to take that into account. . .

Bleah!

What applies to subatomic particles does not necessarily apply to human brains (at that level of organization). The characteristics that emerge from such higher levels of organization that give rise to consciousness are NOT the characteristics of subatomic particles--especially the "weirdness" which is pretty much defined as the characteristics of subatomic particles that are NOT observed in the macro world.

From a logic point of view, it is the composition fallacy to think that if an object is composed of smaller objects with a given property that the larger object must also have that property.

As I mentioned, the collection of mental processes that we call "consciousness" are emergent properties at least at the level of brain structure. You don't get consciousness in a single atom or subatomic particles. Even when physicists ask questions about how a particle "knows" about the spin of a linked particle, the scare quotes are used to recognize that they're using the pathetic fallacy.

Nevertheless, any large-scale physical phenomenon reduces to a composition of small/atomic scale physical phenomena. If consciousness is an actual physical phenomenon, then it will be possible to analyse it at all scales. If consciousness is something that emerges on a large scale, it emerges as a result of small scale processes combined in some way not yet known.

 

[westprog]

What if the brain is analog and fuzzy?

Then strong AI is wrong, since it assumes that the brain is digital and precise.

 

[drkitten]

Then strong AI is wrong, since it assumes that the brain is digital and precise.

No, it assumes that the necessary fuzzy aspects of the brain can be modelled by something digital and precise.

Since we've had precise formulations of fuzzy logic^WP for decades, that's not a particularly controversial assumption.

 

[westprog]

Not in the slightest, no.

His work as an algebraist and geometer has been of tremendous use to physics,

Has it? That's beside the point. Penrose has not been working as a mathemetician solely. There are plenty of pure and applied mathematicians whose work has been used by physicists. They don't win physics prizes.

just as Kahneman's work (and Simon's) work in psychology has been of tremendous use to economists. This does not make Penrose a physicist any more than it makes Kahneman or Simon economists.

And I hold to this statement despite the fact that Kahneman and Simon have received Nobel prizes in economics. Because -- as they would be the first to admit -- their work is psychological, not economic.

Let's stick to science, shall we?

In 1965 at Cambridge, Penrose proved that singularities (such as black holes) could be formed from the gravitational collapse of immense, dying stars

That sounds sort of physics-ish to me. Or how about

In 1969 he conjectured the cosmic censorship hypothesis. This proposes (rather informally) that the universe protects us from the inherent unpredictability of singularities (such as the one in the centre of a black hole) by hiding them from our view behind an event horizon. This form is now known as the "weak censorship hypothesis"; in 1979, Penrose formulated a stronger version called the "strong censorship hypothesis". Together with the BKL conjecture and issues of nonlinear stability, settling the censorship conjectures is one of the most important outstanding problems in general relativity. Also from 1979 dates Penrose's influential Weyl curvature hypothesis on the initial conditions of the observable part of the Universe and the origin of the second law of thermodynamics.[7] Penrose wrote a paper on the Terrell rotation.

Sort of kind of like physics, isn't it?

And what about

In 2004 Penrose released The Road to Reality: A Complete Guide to the Laws of the Universe, a 1,099-page book aimed at giving a comprehensive guide to the laws of physics. He has proposed a novel interpretation of quantum mechanics.[9] Penrose is the Francis and Helen Pentz Distinguished (visiting) Professor of Physics and Mathematics at Pennsylvania State University.

So, he's made discoveries in physics. He's got physical theories published in Physics journals. He's written a big book all about physics. (And it's a very big book, I can assure you, and it's not at all easy reading). And he's a professor of Physics. He is not a mathematician whose ideas have been seized on by physicists. He's a mathematician who's used his mathematical knowledge to do physics. Precisely what else could he possibly do to qualify as a physicist?

Here's a suggestion:

Stop digging.

 

[westprog]

No, it assumes that the necessary fuzzy aspects of the brain can be modelled by something digital and precise.

Any device that exists in the real world will necessarily be fuzzy and analogue, whether it's a brain, an abacus, or a digital computer. The Strong AI hypothesis assumes that the essential element is the emulation of digital processing.

Since we've had precise formulations of fuzzy logic^WP for decades, that's not a particularly controversial assumption.

 

[Dancing David]

Then strong AI is wrong, since it assumes that the brain is digital and precise.

Really, the fuzzy logic people would disagree.

Sounds like a true Scotsman argument to me, is there something that supports your position?
ETA
It may have been defined that way in the past, but that may not be true for all people into AI.

 

[Dancing David]

Any device that exists in the real world will necessarily be fuzzy and analogue, whether it's a brain, an abacus, or a digital computer. The Strong AI hypothesis assumes that the essential element is the emulation of digital processing.

Um digital by defintition is not analog, you can emulate analog on digital.

What are you trying to say?

 

[JoeTheJuggler]

How many brain cells do we need before consciousness emerges?

I don't think any number of disorganized brain cells will result in consciousness emerging. (Nor will any number of aggregated atoms or molecules. Nor will a collection of amygdalas or a collection of nerve fibers.)

However, the question as to what specific structures give rise to what specific components of "consciousness" (again stuff like memory, proprioception, language, empathy, etc. are all components of consciousness) and how they do it are questions that I predict will be answered by neuroscience (or related fields) and not by physics.

 

[JoeTheJuggler]

Nevertheless, any large-scale physical phenomenon reduces to a composition of small/atomic scale physical phenomena. If consciousness is an actual physical phenomenon, then it will be possible to analyse it at all scales.

No. That's not true. That is the composition fallacy.

Consider a structure like a house. It has emergent properties, such as the ability to keep a person protected from the weather.

A brick, which is a component of that house, does NOT have that property. You can analyze a brick all you want, and it will not exhibit the properties that emerge at higher levels.

Similarly, you can analyze a hydrogen atom all you want and you will not find in it the properties of water.

It is as silly to consider the human brain to be a collection of atoms as it would be for your auto mechanic to try to diagnose a problem with your car's engine by considering it to be a collection of atoms.

QM will no more explain consciousness than it will explain why your car won't start (or, for that matter, why it does start when it's working). That's why a competent automotive mechanic doesn't have to learn QM.

 

[JoeTheJuggler]

I still see no answer to the question I asked at the end of post #118. I'll ask it again,
has trying to understand consciousness by invoking QM led to any practical applications (as examining it with neuroscience has)?

How's that "quantum anaesthesia" for surgery coming along? (Again, the sensation of pain is a part of consciousness.)

 

[westprog]

No. That's not true. That is the composition fallacy.

Consider a structure like a house. It has emergent properties, such as the ability to keep a person protected from the weather.

A brick, which is a component of that house, does NOT have that property. You can analyze a brick all you want, and it will not exhibit the properties that emerge at higher levels.

Similarly, you can analyze a hydrogen atom all you want and you will not find in it the properties of water.

It is as silly to consider the human brain to be a collection of atoms as it would be for your auto mechanic to try to diagnose a problem with your car's engine by considering it to be a collection of atoms.

QM will no more explain consciousness than it will explain why your car won't start (or, for that matter, why it does start when it's working). That's why a competent automotive mechanic doesn't have to learn QM.

Of course a mechanic doesn't need to know physics in order to understand how a car works up to a point. But if he wants to know how chemical energy is converted into motion, it is necessary to understand the workings of the car on an atomic level. And we do understand it on an atomic level. We can actually see how the chemical energy contained in the fuel and oxygen produce heat, which produces pressure, which involves molecules moving faster and striking the piston head, which produces force on the crankshaft which turns the wheel. If we don't know how all this happens on an atomic scale, we don't know how a car works. Saying that a car moves because it's an emergent property of fuel air and steel is just saying we don't actually know how the car works.

Until we can trace the effects of consciousness from the micro level, tracking the forces involved, then we don't have a physical process.

 

[westprog]

Um digital by defintition is not analog, you can emulate analog on digital.

What are you trying to say?

I'm saying that all real world computing devices are in fact analogue in nature, and emulate an ideal digital computer. There is no such thing in reality as a digital computer, since nature at the level at which such devices operate is not digital in nature.

It might be possible, theoretically, to create a genuinely digital device, but it hasn't been done yet.

 

[westprog]

I still see no answer to the question I asked at the end of post #118. I'll ask it again,
has trying to understand consciousness by invoking QM led to any practical applications (as examining it with neuroscience has)?

How's that "quantum anaesthesia" for surgery coming along? (Again, the sensation of pain is a part of consciousness.)

It's not to be expected that theory would lead to more practical applications than experimental research. AFAIAA, no theory of consciousness has produced any practical applications so far.

There is no dichotomy between experimental research and forming theories. It's impossible to devise theories without data, but uninterpreted data is of no value.

 

[RandFan]

How many brain cells do we need before consciousness emerges?

How many atoms are needed in the wing of a plane before flight emerges?

 

[RandFan]

I'm saying that all real world computing devices are in fact analogue in nature, and emulate an ideal digital computer. There is no such thing in reality as a digital computer, since nature at the level at which such devices operate is not digital in nature.

It might be possible, theoretically, to create a genuinely digital device, but it hasn't been done yet.

A semantical argument. Digital has a real and fundamental meaning when it comes to data transmission. That all underlying medium is analog does not obviate that meaning. Digital transmission has high fidelity (not perfect). Analog, by comparison, does not.