Ravyne
The program uses asc for input but turns it into codes relative to each characture,word,fragment,page, or file.
the codes are feed to the neural systems as a pattern ???.
Feed back is done in the coder as each characture,word,fragment,page, or file as needed,thus full control.
The codes can be weighted on or off or status "1001101100"
A lot of potential for monitoring and testing and changing output code.
in a network logic as well as neural would compliment each other.
The final out put would be the most logical as well as probable.
My idea was to walk thru the program,point out the existing characteristics,and controls.
Show what I do with it,and let someone do their own conclusion if there is something new.
The codes are not factored only changed,related,and directed.
I was only going into detail if someone was really interested in studying it,then I'd explain what i can,part by part.
You don't see that unless you play with the program.
even if each neuron had a small logic system ,you'd have a lot of logic.
These are my thoughts not facts.