|
Not exactly, it's not just anything. If you want to look more widely then there are two things we can talk about artificial consciousness, and these are definition and things related to that, and the systems what implement it. But then there are only two ai systems what are fully self-developing and implemented by computer program, and these are cellular automata and absolutely dynamic systems, but cellular automata are not exactly unrestricted.
Why I want to talk about it in my forum is that the things would be in one place then, nobody wouls start to search them in this forum. I didn't get any spam after registering in proboards, but you can post without registering and still with your name, this is what I always prefer to do because logging in kind of too inconvenient.
I cannot comment everything what you said because I didn't think about it so exactly, I derived, implemented by computer program and tested the mechanism, and tested that it learns. Now how it differs from many "theories of consciousness" is that they often don't argue anything, so you cannot prove them wrong. The things they say are obvious, known to everybody, carefully chosen not to be anything completely new or original, and just said in other words. Nobody proved me wrong yet, but I can be proved wrong or right because I argued something and implemented something very exact. If you for example can prove something wrong then it would be very useful because I will implement it in my program then and you would be a co-author of it. Or if you can show that there is another mechanism what can be implemented by computer program and what is unrestricted (as unrestricted as possible).
> The few knots and links "surround" the subtracted items effectively making them cells surrounded by membranes?
> I suggest would it work to teach it to build "cells" of common ground (so allow common knots but separate them by "cell walls" of non-common knots)?
> As soon as you introduce restrictions like angles and nodes you would perhaps make your system "stringy" ?
These are in fact part of one problem. It seems that in any self-developing system would form some "nodes" what are connected to each other. Concerning consciousness they must form themselves, but we must implement them if we want exactly model something. The nodes must be separated enough or be self-repairing so that some change in environment will not destroy them. If you for example can implement such node with my program, then we may not be far away from modelling, say, two slit experiment with it. Such nodes may change connections, create new nodes etc, in that way they are rather foamy :-)
Well, and whether such self-developing and unrestricted system exists on quantum level, we may find out on quantum computer one day.
Tarvo |
