Seeing red

[bennyboy]

Thank you for describing your model to me.  That does make a lot of sense.   I can't help but feel though that it is missing a command center where the output of these stages is registered.  Maybe if I had more experience with neural nets, but maybe not.

My understanding of neural nets is they exactly don't have a command "center." Basically you provide a "punishment" or "reward" of random outputs given many trials, and the weighted connections between virtual neurons are adjusted each time in a kind of cascading action. So if you have an input that is a picture of an elephant, and you want "elephant" at the end, you keep showing the picture and punishing the ANN until due to statistical changes over say a million trials it finally outputs "elephant."

[The Grand Nudger]

Which would be the NN state "elephant", a description of the system, at that snapshot in time would be what the idea of an elephant was to comp mind....in this case, even though it has nothing in particular to do with the picture of the elephant. Mind/brain isn't being rational, just performing a series of logical operations with a useful outcome.

So long as "elephant" is the right answer, I'm golden.....I don't get trampled, for example. It doesn't matter that my mind may have perfomed a billion useless steps on the variable 011100100101, rather than responded to the elephant. What matters is whether or not it works.

[emjay]

Yes I know... the brain is the ultimate in parallel processing. But the beauty of the brain is it's ability to integrate, through association, information from any source. So even if there are entire brain regions dedicated to a specific task, involving hundreds of layers and transformations, their output still collapses down to neurons firing, and those neurons can be associated with the output neurons from another system.

When you say associated, what is the nature of that association?  How is it that these different systems are brought into relation in the experience of a single mind?

People are associated, too.  We are communicating symbols through words all the time.  Does this coordination mean that there is a supermind of which we are all part?

There are two ways to think about association in the brain. The simple way is if neuron A synapses with neuron B and both A and B are firing at the same time, association is the strengthening of the synaptic strength between them as a function of the activity in both neurons, associating the two neurons so that B is more likely to fire if A is firing. This strengthening of the synaptic strengths is a function of both how active the post-synaptic neuron is (B) and the pre-synaptic neuron (A) and roughly translates to saying the more active B is the more the synaptic strength with increase relative to how active A is. This is roughly the way it is modelled in Emergent and that is based on the underlying biology. The second way of thinking about association, and the way I mostly think about it and where I see the beauty, builds on the first and is the effect of neurons X and Y both synapsing with neuron Z combined with bidirectional connectivity - which is to say neuron Z also synapses with neurons X and Y but going the other way. The same principles apply for the changing of the weights as for A and B above but in this case Z can be said to 'bind' X and Y... it becomes a detector for X and Y and is more likely to fire if both of them are active (according to the weight distribution between them). So in this usage, X becomes associated with Y through Z. And neural network dynamics will make it that the feedback synaptic strengths from Z to X and Y will tend to closely mirror the input synaptic strengths from X and Y to Z... so if I notate a synapse as X>Y then Y>X comes to be roughly the same. The effect of this is to create bias and allow for pattern completion, so if X fires and Y does not, but X still manages to activate Z even to a small degree then Z will start sending activation back to X and Y proportional to the weights. This will increase the activity of the already firing X, provide bias in Y to become active (so it requires less input current from elsewhere to activate), and increase the firing of Z as a result of the increased activity of X, which with then feedback more etc. In short it's a feedback loop that allows a whole context of related connections to be 'bootstrapped' into action very quickly based on very little initial input. To stop the feedback running away with itself, there are inhibitory interneurons which output inhibitory current to offset the excitatory current coming into neurons. So in this case Z would synapse with an inhibitory interneuron which in turn synapses with Z and fires proportionally to the activation of Z, stopping it from getting over-excited and allowing the network to settle into a stable state. So once you have a learned set of associations - spanning many related 'binding' neurons... what I call a context - this process allows it to become completely activated in leaps and bounds based on very small amounts of well-placed input from outside. Each little input will cause feedback activation that will cascade through the whole context, activating some neurons and biasing others as it goes, with the activated ones now contributing in the same way, speeding it up even more.

So if I want to recall some long distant memory, I think about it in these terms, knowing that if I can find the right input, it could bootstrap the whole thing in vivid detail. But if I keep coming up against a mental block, it means that the input I'm providing is peripheral to the context and is not well-placed enough to start a useful cascade. So for instance even if there is a whole tree of associations that I manage to activate, if they are only associated with another whole tree - the one I want to activate - through a small weight with a binding neuron, then however active that first tree is, it will not be able to trigger the binding neuron which would provide the bootstrapping feedback to the second tree. This is what I would call a red herring and is the equivalent of the context changes in Alice in Wonderland... a tear becomes a sea and the whole context changes with nothing else in common between the first scene and the second. So in that situation, remembering anything from the first scene will not activate anything from the second scene because the only point of entry is through that tear (or perhaps Alice herself... that would provide a little bit of activation, and thus bias, in the second scene) or something in the second scene.

Regarding people communication symbols, yes, words are the ultimate in association. They bind anything we can think of together, from any modality, allowing us both easy access to those representations and the ability to share them with others, to share and refine ideas (contexts). So yes, in a way you could think of words as a means of transferring representations from person to person. A book Evie put me onto, Consciousness Explained by Dan Dennett, calls such contexts 'memes' (as did Richard Dawkins), and took the evolution of mind to a whole new level, in that we could easily 'install' these useful ideas in ourselves or others and that some were more likely to take up root than others, kind of like a virus living in a host, replicating as a function of how appealing they were to us but ultimately not 'caring' about us but only their own replication (and not really 'caring' about that either - ie survival of the fittest and evolution where memes are concerned). As to 'superminds' I've never thought of it like that but you could say the meme associates (binds) two people, with their association stronger the more representations they have in common. So yeah, it would be interesting to think about the people and groups in the world (nationalities, religions, etc) in neural network terms and see if there's any similar pattern of activity

[bennyboy]

Which would be the NN state "elephant", a description of the system, at that snapshot in time would be what the idea of an elephant was to comp mind....in this case, even though it has nothing in particular to do with the picture of the elephant.  Mind/brain isn't being rational, just performing a series of logical operations with a useful outcome.

So long as "elephant" is the right answer, I'm golden.....I don't get trampled, for example.  It doesn't matter that my mind may have perfomed a billion useless steps on the variable 011100100101, rather than responded to the elephant.  What matters is whether or not it works.

That's right.  In this sense, though, I'd posit that Google may be the mind of God, essentially.  It's spooky how you type "a" and it fills in "Asian ass porn," right?

[The Grand Nudger]

There's a whole group of loonies that worry about something like google becoming "the mind of god" and summarily exterminating humanity or some-such, lol.  

Whenever I hit a, I get atheist forums, deviant.  

[bennyboy]

There's a whole group of loonies that worry about something like google becoming "the mind of god" and summarily exterminating humanity or some-such, lol.  

Whenever I hit a, I get atheist forums, deviant.  

I think if you crunch the numbers, the guy who gets "atheist forums" coming up first is statistically more deviant than the guy who gets "Asian ass porn."

Leaving that aside, I'd like to see an argument why we SHOULDN'T think of humanity as a whole as an entity perhaps with a mind of its own.  Certainly it processes world information and executes actions.  Certainly it contains all of the contents of any individual's world view millions of times over.  So if mind is just a collection of self-referencing logic gates, does this not mean that Google is the closest thing to God we have got so far?

To me, the answer is no-- but that's because I think machinery, no matter how complex, cannot amound to mind.  But I'm curious abount your stance.

[The Grand Nudger]

We do think of humanity that way, when we speak of the "spirit of the times" for example...or when we talk about people being swept up by a trend or a fad, when we discuss society.  I think we're the closest thing we have to a god so far...googles nice, but we have a hell of a headstart in the r/d dept...to the tune of billions of years.  Unsurprising, I suppose, since we modeled god after ourselves.

I'd accept that machinery more advanced than my own had a mind more advanced than my own. I'd see no benefit in denying it, and I'm already very familiar with the experience of bowing to my betters, lol. I'd definitely ask it for stock tips.

[emjay]

So Rhythm, do you think that the operation of a neuron can be modelled precisely using logic gates. Ie the difference between a neuron and an AND gate is that a neuron accepts any number of weighted inputs, and outputs an analogue rather than binary signal reflecting the degree of 'truth' rather than simply on or off. I mean I can see where you're coming from with the red and apple AND gate but neurons are far more complex than an AND gate. Is this where NAND gates come in?

[The Grand Nudger]

In principle any comp system can be built using nothing but nand...so ultimately I'd have to say that if you knew how neurons worked, to an exacting degree...you would be able to build a replica of the system in another architecture, in nand rather than neuron.  Whether or not this theoretical machine would qualify, to you, as a precise model would be up for debate, I imagine.  All of this is only applicable if mind is comp, ofc.  

I use AND to simplify, grossly, when people are stuck on the fundamentals.....when they think that it -can't- be done. You can zoom in and out with it. For example, saying that if a nueron was only able to do what an and gate could do..it could yield this and this...or up higher, saying, if two centers of the brain A and B, through vastly complicated multidirectional concurrent processing involving billions of neurons yield state C...that's red apple.

[bennyboy]

Based on what I know about neurons, I think the distribution of neurotransmitters in the brain represents at least partly a chaotic system. Minor vibrations in the skull, the particular timing of the pulse, and almost anything else could affect the trigger timing of an indidual neuron, or possibly a group of neurons. We often talk about QM unpredictability as adding an essentially random element to all events, but I think the brain is so complex that classical mechanics should make brain function as unpredictable as the weather, especially over long times.

I'd say dreaming is an excellent example. Maybe you could see the strange and unpredictable content of dreams each night as a series of butterfly effects causing parts of the brain to light up and come into coordination with each other, sometimes in ways that daily sense perceptions would never allow for.