Progress in consciousness?

stuh505 wrote:

Quote:
Each neuron may be connected to thousands of other neurons. The network "reaches a decision" by adding and subtracting neurons to and from the network as their input reaches or fails to reach the threshold (in other words, it accesses memories, logical processes, and emotions) until a condition is achieved where the network is providing sufficient stimulation to the nerves that tell the body to do something.

ok, i dont work with neural networks (yet...)...my mom does, but i know that the network learns by modifying the multipliers of each input signal relative to the difference in the desired result. it reaches a decision the same way it learns; by simply plugging in the inputs and seeing what the output neaurons yield, which can only be a 0/1. but a combination of 0's and 1's can give a more complex answer of course. im pretty sure that there is no adding or subtracting of neurons from the network, that would seem to defeat the purpose.

i think i can speak for Terry, in saying i am sure she meant "by taking the 'algebraic sum' of the neural outputs" which means the same as you are saying.

stuh505 wrote:

bogowo, i think i promised to get back to you on the evolution thing...anyway i was wrong. i could have sworn there was another process in effect but if there is i coudnt find mention of it...............

having a healthy scepticism toward any statements, can never be described as being 'wrong', stuh! Laughing

0 Replies

Quote:

i think i can speak for Terry, in saying i am sure she meant "by taking the 'algebraic sum' of the neural outputs" which means the same as you are saying.

well the neural outputs are 0 and 1, taking the algebraic sum of the outputs eg 1010101011 would yield 6 which is definitely not the right way,

the only place where a sum is involved is the activation function WITHIN the neuron, each input would be multiplied by a scalar weight and then these values would be summed and compared to a threshold value to determine the output

is that what you meant?

but this is actually the outdated method, i guess now we use a differentiable activation function so the output would not longer be a step function

0 Replies

OK, this isn't the thread topic.... well maybe a little

http://www.onelife.com/evolve/brain.html

0 Replies

Terry, more importantly...neural networks could not POSSIBLY be used to make decisions the way you describe. Let me explain.

In a neural net you first have to give it a set of datasets and corresponding desired outputs. Using these datasets, the neural net is trained to produce the desired outputs whenever it gets the given inputs...and this training allows it to generalize, so that when it gets a dataset which does not exactly match any that it was trained on, it will yield a result that is a compromise between the closest matching input datasets.

Suppose you wanted to use a neural network to make decisions in a chess game. First, you would have to show it a whole bunch of chess boards with randomly set up pieces...and then tell it exactly what move is the best move for each case. Then when it played, it would try to make those moves whenever it could...and if the setup wasnt exactly the same, it would figure out what setup it was most like and make that move. As you can see, this would not work as a decision making algorithm....because it could never plan ahead, or have any knowledge of what was actually going on in the game...it would be making every move on a turn by turn basis.

In a game of chess, you would want to give it 1 goal: win the game. But this could never be programmed into a neural net, because you cant train a neural net based on GOALS, neural nets cannot come up with solutions to problems, they can only recognize patterns that you give them. So you'd have to train it based on giving it SOLUTIONS based on each set of input parameters.

But when you give it solutions, it's not actually "figuring anything out" it's just finding the pattern. A neural net could NEVER make an actual decision on it's own, at least not the kind of neural nets we have today.

0 Replies

stuh

Have you tried Capra "The Turning Point" for an alternative to neural nets. ?

<< ...the systems view of mind seems perfectly consistent with both the scientific and mystical views of consciousness, and thus to provide the ideal framework for unifying the two.

The systems view agrees with the conventional scientific view that consciousness is a manifestation of complex material patterns. To be more precise, it is a manifestation of living systems of a certain complexity. On the other hand, the biological structures of these systems are expressions of underlying processes that represent the system's self-organisation, and hence its mind. In this sense material structures are no longer considered as primary reality. Extending this way of thinking to the universe as a whole, it is not too far-fetched to assume that all its structures - from the subatomic particles to galaxies and from bacteria to human beings - are manifestations of the universe's self-organising dynamics, which we have identified with the cosmic mind. But tbis is almost the mystical view, the only difference being that mystics emphasise the direct experience of cosmic consciousness that goes beyond the scientific approach. Still, the two approaches seems to be quite compatible. The systems view of nature at last seems to provide a meaningful scientific framework for approaching the age-old questions of the nature of life, mind, consciousness and matter...>> (Chapter 8.)

0 Replies

first of all i do not deny that our brain uses neural networks...neural networks in computers are just a dummified copy of the neural networks in our brains, all i am saying is that our neural networks as we understand them cannot do anythign other than analyze the data from our five senses and possibly make predictions, something else must make our decisions.

as a fun experiment to understanding neural networks, ive started another small project in my spare time...writing a computer chess AI that uses a neural network to learn from it's opponent

sorry fresco i cant say im a convert to the systems view

0 Replies

sharing.......

Daniel Dennett's works...
discuss any?
-c

0 Replies

nolanguagenrlungs

Dennett wrote (debate with Chalmers) in advocating the by-passing of "conciousness"

" It still seems that the sun goes round the earth...., but we've learned not to credit those intuitions."

Italics=Dennett Bold=Me

It appears from this that Dennett could be making a fundamental error of assuming the heliocentic model has some intrinsic validity relative to the geocentric model, and by extrapolation that an "AI approach" to "thought" has a similar "validity". I will read more but this apparent error - of ignoring the "purposes of observation" makes me skeptical of his approach especially when he too simplstically relates the concept of "geocenticity" to the concept of "vitalism" ...thence "consciousness".
This would mean his argument is vacuous on two counts.

0 Replies

I agree, Fresco, that purpose and standpoint are critical epistemological factors. From one standpoint the heliocentric model is appropriate, but mankind has been served very well by the geocentric model throughout most of his career--I still behave most of the day as if the world were flat. To respond, as most will, that "the real" or "objective" standpoint, the one taken from a scientific analytical or empirical perspective, is the only one that "counts" for Truth is not wrong. But that "true" perspective does not cover all possibilites. As such it seems narrow and naive.

0 Replies

I agree, Fresco, that purpose and standpoint are critical epistemological factors. From one standpoint the heliocentric model is appropriate, but mankind has been served very well by the geocentric model throughout most of his career--I still behave most of the day as if the world were flat. To respond, as most will, that "the real" or "objective" standpoint, the one taken from a scientific analytical or empirical perspective, is the only one that "counts" for Truth is not wrong. But that "true" perspective does not cover all possibilites. As such it seems narrow and naive.

0 Replies

i suggest that you gentlemen who are well versed in the practice of 'meditation', take up a comfortable position looking west, on a clear night at "sundown".
As you watch the setting of the sun, approach if you can, a trance state, in which the earth is rotating toward you, and the sun, a fixed object in the sky, is slowly being eclipsed.
So much for geocentricity! Rolling Eyes

0 Replies

0 Replies

BGW

I think Denett should try selling "heliocentic reality" to those Scandanavian woodsmen suffering from SAD (Seasonal Affective Disorder) but he should keep one eye on that axe on the cabin wall ! :wink:

0 Replies

Fresco, a quick question if I may.
Could 'harmonic vibration' have anything to do with consciousness or levels of same?
Just curious.

0 Replies

fresco wrote:

nolanguagenrlungs

Dennett wrote (debate with Chalmers) in advocating the by-passing of "conciousness"

" It still seems that the sun goes round the earth...., but we've learned not to credit those intuitions."

Italics=Dennett Bold=Me

It appears from this that Dennett could be making a fundamental error of assuming the heliocentic model has some intrinsic validity relative to the geocentric model, and by extrapolation that an "AI approach" to "thought" has a similar "validity". I will read more but this apparent error - of ignoring the "purposes of observation" makes me skeptical of his approach especially when he too simplstically relates the concept of "geocenticity" to the concept of "vitalism" ...thence "consciousness".
This would mean his argument is vacuous on two counts.

He didn't assume that the heliocentic model has intrinsic validity in concerning the functionalism of the geocentric model. He doesn't assume that the "experience of" consciousness has any validity on the materialistic functioning either. He does however acknowledge qualia of what's it like to experience something. I came to the understanding that he is not truly an eliminativist. However I think I see what you're talking about...Break me free if I am wrong, but are you saying that he can be seen as inconsistent and can be refuted because his problem implies two options,
having the actual apprehensibiity of qualia without the intrinsicality
of qualia, ....and/or vise versa?
If that is what you mean, than I guess I'll be spending my reading time, re-reading.

0 Replies

nolanguagenrlungs

I think Dennett's attempt to side-step "consciousness" is doomed because he still clings to a concept of anpthropocentric "science" irrespective of his take on "qualia". Such "science" (as we know it) has a particular "functionalism" built in - that of "prediction and control".This is often overlooked. It is only perhaps by taking a wider view of consciousness which "transcends" such ephemeral aspirations of humanity that alternative "systems" can be appreciated, (e.g. Capra) in which "qualia" may have as little significance to the overall system as "sparks" do to an understanding of "electricity".

Gelisgesti

Some esoteric systems indeed use "resonance" as an explanatory model for their dealings with "consciousness". However, I am not aware that this is a central concept (yet) in "systems dynamics" which is a less mystical approach. But it is signifant that "equilibrium" is a central issue in the latter and this has semantic leanings towards "standing waves" in the former.

0 Replies

fresco wrote:

nolanguagenrlungs

I think Dennett's attempt to side-step "consciousness" is doomed because he still clings to a concept of anpthropocentric "science" irrespective of his take on "qualia". Such "science" (as we know it) has a particular "functionalism" built in - that of "prediction and control".This is often overlooked. It is only perhaps by taking a wider view of consciousness which "transcends" such ephemeral aspirations of humanity that alternative "systems" can be appreciated, (e.g. Capra) in which "qualia" may have as little significance to the overall system as "sparks" do to an understanding of "electricity".

Fresco,
Is Capra leaning towards eastern philosophy? I have never studied him. Also, if I were to start reading (Fritjof)? Capra's works, what would you suggest I start with? I am mainly interested in the possibility of other species' consciousness. And that consciousness isn't as specialized as we deem it to be. I've delved into the moral philosophy a smidge, with Peter Singer, but I am interested in a more analitical philosophy. If you can start me off with a beginner's title to his works, then I'd much appreciate it.
Thanks,
c

0 Replies

nolanguagenrlungs

Yes, he does lean towards the eastern tradition - first movements perhaps in "The Tao of Physics".
However I suggest you start with the more recent "Web of Life" (or Google references thereto).

I would add that I am not a particular fan of Capra's "deep ecology" stance, but it is one way of moving away from problems of "observer-observed". You might note also that the "eastern slant" in general is criticised by Dawkins who often makes sense elsewhere.

Progress in consciousness?

Related Topics

Quick Links

My Account

able2know