Ai Forums Home Welcome Guest    Friday, November 24, 2017
Ai Site > Ai Forums > Language Mind and Consciousness > Brain physics Last PostsLoginRegisterWhy Register
Topic: Brain physics

tkorrovi
posted 5/5/2010  16:39Send e-mail to userReply with quote
Hunt, you talked about half of the world in the physiology of sense thread, which should all properly be in separate threads. So i found no other solution than making a new thread here, which might not be proper either.

Because you are (or were) physicist, then no surprise you are interested in physics of these things.

Penrose was also much criticized because of the way how he applied the Gödel's theorem. It is possible to talk endlessly about Gödel's theorem, but in short Gödel's theorem is about Principia Mathematica. Formal systems Principia Mathematica are not complete when there are loops. So all what Gödel's theorem really says about algorithm is that there are loops in it, and indeed most of the algorithms have loops, but nevertheless, algorithms with loops can be implemented in the computer.

There is nothing mysterious in superposition. There is no way to measure something without interfering at all with the object which we measure. And this interfering inevitably changes the object, so the final measurement is the result of both outside conditions and our influence. We may sometimes say for certain in what state the object is, but sometimes we may not know all the outside conditions and how they influence the state of the object.

Microtubules form something what is called a cytoskeleton. Cytoskeleton is interesting not only because there might or might not be some quantum effects. Every cell has cytoskeleton, in general it is a system with changing topology, and it participates in almost everything in the life of the cell. More is known about some parts of the cytoskeleton, less about the others. Cytoskeleton is the only system which i know in the neuron which is able to implement my ADS-AC mechanism.

I don't know whether there are some quantum effects which happen in the cytoskeleton. Indeed there may not be clear quantum states when the temperature is high enough, but this doesn't mean that there cannot be quantum entanglement, and the systems like mine are about entanglement. David Bohm created the only causal theory of Quantum Mechanics, and in his theory everything is based on implicate order, which also is nothing bout entanglement.

Please don't now misinterpret me like i said above that there can be consciousness at the basic level of the universe. I have not said that. If even universe and mind have the same basic mechanism, then a mere presence of that basic mechanism doesn't yet mean the presence of consciousness.

 Artificial Consciousness ADS-AC project

hunt
posted 5/6/2010  02:58Reply with quote
"Because you are (or were) physicist, then no surprise you are interested in physics of these things."

Physicist by day, crackpot chatterbot developer by night. :)

"Penrose was also much criticized because of the way how he applied the Gödel's theorem. It is possible to talk endlessly about Gödel's theorem, but in short Gödel's theorem is about Principia Mathematica. Formal systems Principia Mathematica are not complete when there are loops. So all what Gödel's theorem really says about algorithm is that there are loops in it, and indeed most of the algorithms have loops, but nevertheless, algorithms with loops can be implemented in the computer."

I'll admit my understanding of his use of Goedel's theorem comes by way of translation into laymen's terms. A quick google search of the subject yielded this paper by Solomon Feferman of Standford University, which seems interesting, though involving mathematical semantics beyond my training: "Penrose's Goedelian argument" (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.130.7027&rep=rep1&type=pdf)

Feferman is also of the belief that the human mind cannot be reformulated as an algorithm, but he has this to say on Penrose:

"So on the face of it, mathematical thought as it is actually produced is not mechanical; I agree with Penrose in this respect, *understanding* is essential, and it is just this aspect of actual mathematical thought that machines cannot share with us. Beyond that, his entire drive is to nail down this conviction by showing that mathematical thought cannot even be *re-represented* in mechanical terms, as a result of the Goedel theorem. In my view, instead of increasing this conviction, this effort raises more questions than it answers and leads one off into dead-end dialectics." (pg 11)

The resource I've found most helpful is by Paul Almond: "A Refutation of Penrose's Godel-Turing Proof that Computational Artificial Intelligence is Impossible" (http://www.paul-almond.com/RefutationofPenroseGodelTuring.htm)

I think I'll determine my opinion on the subject the same way I do for all fields outside my expertise: along the lines of the majority who know what they're talking about! (And I think in this case a majority opinion clearly exists.) I may not have a thorough understanding of the proofs involved, but I believe in the scientific process. (That is, professors jockeying for authority on a subject and eager to tear through each others' work.)

"There is nothing mysterious in superposition. There is no way to measure something without interfering at all with the object which we measure. And this interfering inevitably changes the object, so the final measurement is the result of both outside conditions and our influence."

I agree with everything written above, though I want to emphasize that superposition and interference due to measurement are not related per se.

"We may sometimes say for certain in what state the object is, but sometimes we may not know all the outside conditions and how they influence the state of the object."

I want to emphasize that when it comes to "quantum uncertainty," one usually is not referring to a lack of understanding about outside conditions, but rather a fundamental limit in our ability to know the state of an object.

"More is known about some parts of the cytoskeleton, less about the others. Cytoskeleton is the only system which i know in the neuron which is able to implement my ADS-AC mechanism."

I googled this term "ADS-AC mechanism" and found what I assume is your page on SourceForge, but nothing else related to the topic. Is there a topic thread on the forum that explains what it is your working on?

"I don't know whether there are some quantum effects which happen in the cytoskeleton. Indeed there may not be clear quantum states when the temperature is high enough, but this doesn't mean that there cannot be quantum entanglement, and the systems like mine are about entanglement."

If you cannot have a quantum state you certainly cannot have entanglement! Entangled states decohere with a time scale that is (generally) inversely proportional to the temperature of the system. And that time scale is much faster than the speed of thought in something warm like the brain.

Frankly I think there is a tendency (and physicists are the worst about this) to try to interpret every new development as something exotic. Sometimes systems are just classical. That doesn't make them any less complicated or interesting though!

"David Bohm created the only causal theory of Quantum Mechanics, and in his theory everything is based on implicate order, which also is nothing bout entanglement."

As I understand it, Bohm doesn't have a theory of QM so much as a supposition that the state of a system is deterministic via a scheme of (non-local) hidden variables. A local hidden variable scheme has been demonstrably shown to be unsupported by evidence (via Bell's Inequality tests.)

"Please don't now misinterpret me like i said above that there can be consciousness at the basic level of the universe. I have not said that. If even universe and mind have the same basic mechanism, then a mere presence of that basic mechanism doesn't yet mean the presence of consciousness."

He he--nothing you have said implies to me that the universe has consciousness. Every system, at the most fundamental level, operates on the same set of rules. It is complexity that gives rise to variety, and why so many physicists across disciplines are turning there attention to the concept of emergent phenomena.


tkorrovi
posted 5/6/2010  10:07Send e-mail to userReply with quote
The problem is, not every kind of complexity, this is the most important, and the thing to think about.

About my system, i have talked repeatedly about it in this forum, so if you search my posts, then you would certainly find a lot about it. But i'm confused, i have no idea any more what is that which people don't understand, and so i don't know what i should explain. If there even is such thing, and they reject it just because it is not made by a prominent scientist, and there is absolutely nothing which i can do or how i can help in that case. You found the SourceForge project, there is also an Artificial Consciousness forum at http://tkorrovi.proboards.com .

But in general, if this might give you some idea. The cellular automata are not derived to be unrestricted, they have often made just using some arbitrary rules. And one can deny that, but it shows, it is generally not possible to interfere with the patterns development, and thus it is also not possible to train them (i talk about ca and not recurrent nn, which are nn not ca). Why, they just collapse whenever you "touch them". Thus, some different mechanism may be necessary which has less restrictions. Also essental seem to be the "non-local" connections. ADS-AC is one such proposed one. That's all, in a very short.

Last edited by tkorrovi @ 5/6/2010 10:56:00 AM

hunt
posted 5/6/2010  11:53Reply with quote
"About my system, i have talked repeatedly about it in this forum, so if you search my posts, then you would certainly find a lot about it."

I figured there must be a thread discussing it, which is why I asked for a reference. Up until very recently, I didn't frequent these boards more than maybe once every few months. (That seems to be the time scale of topics here anyway--a very inactive board in general.) I'm not that familiar with what older topics contain.

I'm currently working on a project that gives me large but awkwardly broken up periods of down time. Thus my frequent appearances here. We'll see how long this lasts...

"If there even is such thing, and they reject it just because it is not made by a prominent scientist, and there is absolutely nothing which i can do or how i can help in that case."

Don't feel bad--scientists just want to be sure that other scientists went through the same trial-by-fire that is grad school. I've come to think of it as an elaborate hazing ceremony...

"You found the SourceForge project, there is also an Artificial Consciousness forum at http://tkorrovi.proboards.com ."

Thanks for the link.

"The cellular automata are not derived to be unrestricted, they have often made just using some arbitrary rules."

I once saw a talk on cellular automata by Wolfram. (My only real introduction to the concept.) There's a guy who suffers from the same problem you do--he isn't taken seriously by the physics community. But he does make some nifty software. I guess the moral is, make a lot of money and then you can pursue what you love whether others believe in you or not. In the end, it's the results that matter anyway, right?

"And one can deny that, but it shows, it is generally not possible to interfere with the patterns development, and thus it is also not possible to train them (i talk about ca and not recurrent nn, which are nn not ca). "

I understood that Penrose's hypothesis about the quantum brain had to do with tubulin performing cellular automata-esque calculations (with quantum effects influencing the results.) I did not understand the QM to be necessary to make calculations at all. Interesting...

"Also essental seem to be the "non-local" connections. ADS-AC is one such proposed one. That's all, in a very short."

Thanks for the condensed version. :) I'll check out the link you provided for more info.

Last edited by hunt @ 5/6/2010 11:54:00 AM

tkorrovi
posted 5/6/2010  13:52Send e-mail to userReply with quote
 
hunt wrote @ 5/6/2010 11:53:00 AM:
I understood that Penrose's hypothesis about the quantum brain had to do with tubulin performing cellular automata-esque calculations (with quantum effects influencing the results.) I did not understand the QM to be necessary to make calculations at all. Interesting...

 
I don't know whether we need QM or not, but i have some reasons to say that we need something like entanglement. That is, a system with changing topology where any node can be directly connected to any other, this is also what David Bohm called a "holomovement".

BTW, Penrose may be right in a way which he perhaps didn't think about. When there is an interconnected system, so that something can be connected with something in the other end of the universe, then when we cut off some area of such interconnected system, then everything inside that area may well be non-computable, as we don't know the connections with everything outside that area.

Last edited by tkorrovi @ 5/6/2010 9:22:00 PM

hunt
posted 5/6/2010  14:06Reply with quote
I think we're working from somewhat different definitions of "entanglement". What I mean by the word is: if I have two or more entangled objects (nodes), then if I make a measurement on one node, I know the state of all the other entangled nodes. (Assuming I know what kind of entangled state they're in.) I'm not sure what you're describing requires that precise a definition to describe the relationships between nodes. But I'll have to think about it more...


tkorrovi
posted 5/7/2010  13:31Send e-mail to userReply with quote
I think what i mean is more the David Bohm's understanding of entanglement, several other phenomena, like two slit experiment, superconductivity, etc, are essentially non-local. I don't know how well established it is, but this logic brings ahead. Like, most likely i couldn't make an unrestricted mechanism using the string theory, in spite of its 11 dimensions ;)

The principle which enables direct connections between objects anywhere, essentially determines a room with limitless number of dimensions, on or beyond the 3-dimensional room, not 11 or any other number.

Last edited by tkorrovi @ 5/7/2010 1:43:00 PM

tkorrovi
posted 5/9/2010  18:49Send e-mail to userReply with quote
 
hunt wrote @ 5/6/2010 2:06:00 PM:
What I mean by the word is: if I have two or more entangled objects (nodes), then if I make a measurement on one node, I know the state of all the other entangled nodes.

 
Well, to me this merely says that the two nodes are directly connected. Now the other question is, what this connection means, what properties it changes, etc. Traditionally, they only talk about two photons with connected spins. But there are several other kind of non-local connections. Like, two non-locally connected circuits in superconductors, where the direction of one current is always opposite to the direction of the other current http://www.wired.com/wiredscience/2009/09/quantum-entanglement . I don't know whether Science News is a good enough source, anything which doesn't appear in ScienceDaily supposed to be a pure crankery. So, there is another article in ScienceDaily about the work by the same authors, which though seems to be about some experiments which they did after making their superconductive device http://www.sciencedaily.com/releases/2009/09/090923151730.htm .

But more importantly, the presence of non-local connections says us something about the nature of the things -- what there is, is a big network, instead of only a continuous 3d space, as it was thought before. And we also know that this is a changing network, because in nature everything changes. The third thing which we know is that everything must be very flexible. We may proceed from these three things and think what should be the mechanism which enables all that. One may call it a philosophy, but at least, it doesn't seem credible to me that we can solve the most profound problems, like True AI, without any general thinking.

Last edited by tkorrovi @ 5/9/2010 6:53:00 PM

hunt
posted 5/11/2010  04:23Reply with quote
"Well, to me this merely says that the two nodes are directly connected."

I think the subtlety is in the math describing how they are connected. And whether/how that connection persists after a measurement is made. A simple dependency tree, for instance, is not the same as entanglement.

"Traditionally, they only talk about two photons with connected spins. But there are several other kind of non-local connections. Like, two non-locally connected circuits in superconductors, where the direction of one current is always opposite to the direction of the other current http://www.wired.com/wiredscience/2009/09/quantum-entanglement . [...]"

Physics is an amazing beast. :)

"But more importantly, the presence of non-local connections says us something about the nature of the things -- what there is, is a big network, instead of only a continuous 3d space, as it was thought before."

To me it says that we need to revise the way we think of separate objects. In the case of the photons with entangled spin, could you not also call them both the same object, but spatially extended in a way that photons generally are not? Fun to think about.

"And we also know that this is a changing network, because in nature everything changes."

Are you talking about time evolution or what?

"The third thing which we know is that everything must be very flexible."

What do you mean by this?

"One may call it a philosophy, but at least, it doesn't seem credible to me that we can solve the most profound problems, like True AI, without any general thinking."

Certainly in any such intractable problem, like the development of AI, one must not forget the broader view during development.


tkorrovi
posted 5/11/2010  09:51Send e-mail to userReply with quote
 
hunt wrote @ 5/11/2010 4:23:00 AM:
"Well, to me this merely says that the two nodes are directly connected."

I think the subtlety is in the math describing how they are connected. And whether/how that connection persists after a measurement is made. A simple dependency tree, for instance, is not the same as entanglement.

"Traditionally, they only talk about two photons with connected spins. But there are several other kind of non-local connections. Like, two non-locally connected circuits in superconductors, where the direction of one current is always opposite to the direction of the other current http://www.wired.com/wiredscience/2009/09/quantum-entanglement . [...]"

Physics is an amazing beast. :)

"But more importantly, the presence of non-local connections says us something about the nature of the things -- what there is, is a big network, instead of only a continuous 3d space, as it was thought before."

To me it says that we need to revise the way we think of separate objects. In the case of the photons with entangled spin, could you not also call them both the same object, but spatially extended in a way that photons generally are not? Fun to think about.

"And we also know that this is a changing network, because in nature everything changes."

Are you talking about time evolution or what?

"The third thing which we know is that everything must be very flexible."

What do you mean by this?

"One may call it a philosophy, but at least, it doesn't seem credible to me that we can solve the most profound problems, like True AI, without any general thinking."

Certainly in any such intractable problem, like the development of AI, one must not forget the broader view during development.

 
> In the case of the photons with entangled spin, could you not also call them both the same object, but spatially extended in a way that photons generally are not?

We could in some circumstances consider entangled particles as the same object, but then we likely get many objects partly inside each other which is also not a conventional 3D space any more. But more than that, a particle may have one kind of non-local connection with some particles, and another kind of non-local connection with other particles. Like, electrons can be non-locally connected regarding their movement, but electrons have spins, too. Then it is almost impossible to determine, a part of what "extended particle" some particular particle is.

> What do you mean by this?

Well, you are a physicist, but i am a software engineer, so how i think may sometimes be strange to you. What i mean by flexible, is that when a system would not be flexible enough, then there would be exceptions, circumstances which the system cannot handle. The more complicated the system is, the more likely are the exceptions. And i wouldn't say that for example the Universe is a simple system... When the basic design of a system is not flexible enough, there would be too many exceptions and too much destruction because these exceptions cannot be handled, easily resulting in a system which cannot work at all.

One more thing which i would like to say about flexibility, is that mathematics is not flexible enough. In that, it cannot always describe a system with changing topology of connections. How it would appear in practice, is that we can make every kind of particular models, which also can be precise and accurate, but the problem is that no matter how advanced and complicated the models would be, we cannot make a general model. Perhaps this is not a big problem for physics yet, but it is logical to think that the time would come when it would hit that barrier.

Last edited by tkorrovi @ 5/11/2010 10:22:00 AM

hunt
posted 5/11/2010  22:02Reply with quote
"But more than that, a particle may have one kind of non-local connection with some particles, and another kind of non-local connection with other particles. Like, electrons can be non-locally connected regarding their movement, but electrons have spins, too. Then it is almost impossible to determine, a part of what "extended particle" some particular particle is."

Easy: one (unseparable) wavefunction = one object! :) Then again, at some point mathematics and pragmatism diverge, so perhaps there is no point in semantics games like these.

"What i mean by flexible, is that when a system would not be flexible enough, then there would be exceptions, circumstances which the system cannot handle."

Ah, I know all about this in the context of my AI project. Recognizing exceptions is something I was very bad at when I began, but I think as I go along I'm starting to channel my inner software engineer. It's very interesting watching one's ideas brutally misinterpreted by a machine that has the audacity to do exactly what one tells it...

"And i wouldn't say that for example the Universe is a simple system... When the basic design of a system is not flexible enough, there would be too many exceptions and too much destruction because these exceptions cannot be handled, easily resulting in a system which cannot work at all."

Black holes? I kid, I kid.

"One more thing which i would like to say about flexibility, is that mathematics is not flexible enough. In that, it cannot always describe a system with changing topology of connections. [...]"

I'm not familiar with any fundamental limit in mathematics that prevents the description of a physical system.

I think the important point is that sometimes a complete microscopic description of a system just isn't useful. The complex behaviors that guide behavior on size scales of practical interest are often not intuitable from the microscopic picture alone. It is realizations like this that place phrases like "emergent phenomena" among the new favorite buzz words of the scientific community.


tkorrovi
posted 5/12/2010  09:16Send e-mail to userReply with quote
 
hunt wrote @ 5/11/2010 10:02:00 PM:
I'm not familiar with any fundamental limit in mathematics that prevents the description of a physical system.

 
OK then. Model my mechanism using an equation. This is a simple mechanism, with only two rules, so it shouldn't be difficult ;)


hunt
posted 5/12/2010  18:49Reply with quote
 
tkorrovi wrote @ 5/12/2010 9:16:00 AM:
OK then. Model my mechanism using an equation. This is a simple mechanism, with only two rules, so it shouldn't be difficult ;)

 
Mathematics is the language with which those two rules are written. The fact that you've implemented your mechanism at all implies that it can be described mathematically. Although the results are not necessarily obvious from the written form of the rules--that is, the interesting properties are 'emergent'--does not mean that it lacks a mathematical underpinning.

So in short, I'm still confused by what you mean.


tkorrovi
posted 5/13/2010  12:32Send e-mail to userReply with quote
Yes the problem here is that they have extended the meaning of mathematics so that it includes almost everything. Yes i guess every algorithm can be written mathematically somehow, like in the Gödel's formal system. This is not mathematics which most mean by that, obviously i meant implementing with conventional equations, like with these used in physics.


hunt
posted 5/14/2010  05:55Reply with quote
 
tkorrovi wrote @ 5/13/2010 12:32:00 PM:
Yes the problem here is that they have extended the meaning of mathematics so that it includes almost everything. Yes i guess every algorithm can be written mathematically somehow, like in the Gödel's formal system. This is not mathematics which most mean by that, obviously i meant implementing with conventional equations, like with these used in physics.

 
Even most useful physical equations have no analytic form. But I think I see what you mean.


tkorrovi
posted 5/14/2010  09:27Send e-mail to userReply with quote
OK. There is one more thing about it. If we model such mechanism with Gödel's system or whatever, in essence we model a computer in a way or another, at least the computer's memory. But not the system directly, we have to provide the system separately in the memory. So, mathematics cannot describe the system directly, the only thing which can describe the system, is still the system itself. The mere fact that computer can run programs, doesn't give us much, and cannot anyway help us to analyze the things.


tkorrovi
posted 9/1/2010  00:23Send e-mail to userReply with quote
Well, and now i finally say this. ADS is some kind of model of the Universe, but i don't know how good model it is. I say it is some kind of model because it should be pretty obvious that Universe at its basic level must be very dynamic. I say it just to make clear what my point of view is, and this is pretty much all what i can say.

Last edited by tkorrovi @ 9/1/2010 12:25:00 AM

Devender Singh Rana
posted 3/18/2015  11:15Send e-mail to userReply with quote
 
tkorrovi wrote @ 5/6/2010 10:07:00 AM:

http://tkorrovi.proboards.com .



 
Hey trav can you register me at your forum with these credentials.

email: NickyBlue999@gmail.com

password Neo1596:

forum name: NickyBlue


I tried few times but I dunno why my browser crashing at last form, and I don't wanna deal with that "Prove that you are human" crap everytime I feel like posting something there, right?

So wud you pls do it on my behalf?
thanx


tkorrovi
posted 3/18/2015  16:19Send e-mail to userReply with quote
 
Devender Singh Rana wrote @ 3/18/2015 11:15:00 AM:
Hey trav can you register me at your forum with these credentials.

 
Better don't write your passwords here, this is very unsafe, better edit your post now and delete it.

But this forum, it still has anonymous posting. I know no other forum today which allows anonymous posting. I think anonymous posting is important, for freedom, and one doesn't have to reveal ones identity anyhow in the Internet. But not many people think so any more, which is very sad. Captcha is a nuisance i know, but it enables anonymous posting, which is great. Without captcha, the forum were quickly flooded by bots, so there has to be captcha for anonymous posting.

 Artificial Consciousness Forum
Last edited by tkorrovi @ 3/18/2015 4:36:00 PM

tkorrovi
posted 3/18/2015  16:43Send e-mail to userReply with quote
 
Devender Singh Rana wrote @ 3/18/2015 11:15:00 AM:
Hey trav can you register me at your forum with these credentials.

 
Ok, i registered you with that identity, using the password above *without* the colon in the end. An email was sent to you, you should activate your account with the code there.

Thank you for joining the Artificial Consciousness Forum!

But *please* edit your post above now, and remove your password from there.

 Artificial Consciousness Forum
  1  
'Send Send email to user    Reply with quote Reply with quote    Edit message Edit message

Forums Home    The Artificial Intelligence Forum    Hal and other child machines    Alan and other chatbots  
Contact Us Terms of Use