Log in

View Full Version : A conjecture of some interest



CommieBastard
11th February 2005, 02:57
This is a conjecture i recently discussed with a friend on msn, I am going to cut and paste the relevant sections.
I admit the limits of my knowledge in all of the fields discussed, and welcome being corrected in any particulars.
Didn't have the time to edit it properly, will maybe do so in future. Either way, though disjointed i think you may still be interested in it

right, first off is string theory i guess. basically it's premise is that the universe is composed of strings of energy, and that these string's interactions are what produces physical effects. The strings apparently overlap, and it's possible that there are multiple overlapping universes, that interact on some level, but which don't seem to effect our reality in any meaningful way

next is neural networks. normal computers run linear calculations, you put a set of inputs in, you get a definite set of outputs out. This is very useful for certain kinds of calculations, like mathematics or complex programming. neural networks run on a different principle. basically you dont need to input 0's and 1's, the neural network can produce results in analogue, that is to say on a scale. i cant demonstrate it properly without a diagram, but ill try. neural networks work on the basis of analogue, so their calculations are on the basis of levels of refinement. thats why we can work out the world and objects in detail, or look at them generally. we can also think conceptually in different levels of refinement. our level of refinement is constrained by the complexity of our network, which is in turn constrained by it's size. so, a neural network of different sizes is useful for different purposes, a small one can work out things in greater detail, but cant make large generalisations, whereas a large one can work out greater generalisations but not greater detail. in order to get around this it would be theoretically possible to make an enourmas neural network that is split up into smaller parts, that reference back to the mainframe only at certain times.

so, to explain in analogy to buddhism our life is when our neural network is being used to run computational processes
we interact with other sections of the neural network via our inputs and outputs with a virtual reality that the overall reality creates as a medium for our interaction
this is when we communicate, either verbally, or by moving objects, such as textually or with other signs

were all smaller networks in a biggr one

when we die
our neural network stops simulating our personal virtual reality, and stops referecning with the overall one
it begins to download it's calculations into the overall mainframe of our universe
once it has completed this, it becomes a part of a greater section of the neural network
that is set to the task of processing this date
this is basically the consciousness of our universe
in time, the neural network might get reassigned to another part of it to continue a new process
here is another tie in with buddhism
the neural network is not consistent in it;s quality
so it assigns the best parts of it to the most important tasks
the worst parts get the lower tasks

right, we are in the network of life, that is simulating our 'physical' 'material' reality (that is actually virtual).
nirvana is when we become conscious of the overall consciousness, and move into it
can we reach the universal consciusness without dying? it might be possible if a particular section of the network performs outstandingly it rewritees the universe to move it into better purposes
so thats when we become a part of the overall process of the universe

what is the goals of the performance
maybe to find the meaning of life? who knows. I dont think we can in our level of consciousness. maybe only the multiversal consciousness knows the real purpose. maybe the multiverse is contained in a multiverse and it goes on recusively for at least a while
maybe only the highest level knows the real goals
maybe even they dont

afterall, we are talking about interaction between neural networks that can run the same equation and get different results

you may be aware that in child psychology they have shown that very young children interact with their experiences along pretty much scientific principles of experimentation and reproduction of results
once we learn language, we have a new source for information than our own computational processes
our computational processes running on their scientific basis recognise the benefits of this source, and so seek it out. However, as society stands children are taught very early on about authority and obedience
authority varies, but it has some common principles
that to gain new beneficial information, one has to accept blindly the validity of certain behaviours (where behaviour can include expressing or holding beliefs)
and also the stick and the carrot
if you do not comply in believing, then they employ this method. Your scientific principles tell you to do something that gives positive results, and avoid things that give negative results
this is where it relates to the previous conjecture on the nature of the universe
it is the case that because the neural networks get different results
to gain some form of co-ordination it requires authority
on whether they are capable of coming to the same conclusions when enacting computational processes, i.e. they have to agree complicitly on a wide plethora of matters this is what authority simulates because we act as if we do agree even though we dont
if we were able to find a set of inputs that lead to the same outputs, we would have found grounds for agreement. Basically, im talking about the meaning of life
because if we have a set of principles that cannot be doubted by human minds, then we would not have to force one of the plethora of conjectures on our children, we could communicate it, and they would develop in their neural processes almost optimally
though the authority and obedience paradigms are accepted, and then hinder the scientific paradigm of learning, the scientific paradigm is reintroduced later in life once it can be set into the context of the reinforced (by this time) overall paradigms of authority and obedience. Such that though a person is capable of rational doubt, they have an enourmas set of material to work through with
their doubt once they gain it, and they cannot get to the root point without an awful lot of introspection
because only introspection leads to the real truth
information from the outside cannot be trusted unless verified by information from the inside.
this is what meditation is for in buddhism and why buddhist monks contemplate the universe
i remember someone said that in the past it was noted that great buddhist sages agreed on an incredible number of matters, whilst lower sages tended to disagree more the lower down you went


maybe early on the universal consciousness realised it needed authority, but it was flawed, so it set some of its neural networks to try and work out other means of running the comnunication i.e. a search for universal or indubitable knowledge

encephalon
11th February 2005, 08:32
interesting take, I can sort of see where it's going; a lot of it, actually, reminds me of Jung's work, with neural networks super-imposed on top.

The problem is there's no real substantiated connection between the two concepts, only an explanation of what "might be" without any evidence; it first assumes that buddhism is correct, and then--a concept I have a genuine distaste for--this lovely baseless conjecture:

"...only introspection leads to the real truth"

Even though I'm one hell of an introvert, I find that blatantly false; introspection leads to personal, subjective epiphany, perhaps realization of an external objective truth, but is otherwise entirely subjective.

CommieBastard
11th February 2005, 12:32
I did say it is only a conjecture, i do not believe it to be the case, though i do find it a compelling concept.

As for when i mentioned the real truth, that is from another part of my philosophy, that I do believe to be the case.

Whilst 'external' sources can provide information, if this information is simply accepted then it is functionally useless, as it inhibits the functional development of a system of evaluation and selection.

The use of external sources is that they provide a prompt for us to think in and of ourselves, which gives us the chance to test within our personal perceptual reality the applicability of the suggestions of the external sources.

Introspection is required to know the truth, external sources just help us in the introspection.

CommieBastard
12th February 2005, 14:44
Another point.
The human mind is a multiversal recurring neural network consciousness.
If we do exist within a universe composed of 'strings', then that means that our very brain exists within a multiverse.
What is more though, we can show that our mind exists within a multiverse as well.
We definately do exist within a virtual reality, our mind, being an incredibly powerful computer, creates one for us to exist within and interact with so that we can function within the shared universe of interactions.
Through thought experimentation we can demonstrate that there is a conscious and a non-conscious element to our computational processes.
For a neural network to have consciousness is to say that beyond the development and interaction of the data it processes, it processes data about the means by which it processes data, and the way that this data is constrccted into a supposedly coherent system.

Since we know our consciousness recurs into greater refinement, how do we know it does not recur into greater generalisation?

Imagine if you will a neural network where the nodes themselves are neural networks.
Here is a question i have been pondering, could there be a system of sufficient complexity such that it can simulate a system of greater complexity?
If such a system is possible then it would be possible that it could contain an infinitely recurring set of simulations of systems of exponentially greater complexity