introduction post

[martin of redwall]

im trying to make an A.I. "computer" with adaptive learning. ok so i have a small three input with 36 outputs based on the inputs given. it does work. in oder to explain i will break the system into five parts then explain their relationship. part one input: the three inputs are pressure plates they feed into a one wire bus this bus can be contined an infinite distance. at the end is a component that decodes which input was used. (the flaw here is that the incoming signal has to be a pulse not a hard signal.) part two: the "ALU" this is a parallel of 36 order sensitive two input and three input decoders with built in RAM which allows the decoder to remember the last input since it will need either two or three inputs to active its output. once the output is sent the decoder resets to zero. part three: the direct output: all decoder singals go on to the same one wire bus to a decoder then the proper output is sent. the outputs for now are lamps. part four: counters and EEPROM: counters will count the amount of time a particluar sequence is used i cunnently have it set to three times this is also a self resetting component once the counter reaches three it sends a pulsed signal to EEPROM where the infromation is stored for a later time. part 5: quick lookup table: is is a parrallel of AND gates inwhich the primary input and EEPROM inputs are sent.
relationship: the primary input is sent ao a one wire bus to a bus decoder that signal is sent to the decoders processing wires each input is bussed parrallel to the decoders to allow one main signal source. the primary signal is also sent to the quick lookup tables (covered later). once at the decoders to imformation is processed by either a three input or two input decoder if insuffienct data is avaiable the current data is stored in its internal RAM once enought information is vailable the decoder sends out a signal to a bus line where it is incoded for transmtion along the one wire bus. the bus decoder will activate to proper output. the decoder signal is also sent to a counter which will count the number of times it that event happened once it reaches three (but can be changed to count up to 576 or 1,151 if you dont want auto reset) the signal is sent to EEPROM which is a slightly modified T-flipflop this information is a hard signal to the quick look up table to its event AND gate this turns on half the needed data. the primary input is sent and if its event has happen 3 times a signal is send straight to the output taking 2.3 seconds of delay out of the response.
the A.I. is the fact it can track inputs and a decide a direction of travel. the adaptive part is where an event is stored after three detections to quiken its response.
ill let that sink in and anwser any questions before i tell you some problems ive had. but it does work.

woah A.I with adaptive learning :O.
I'm gonna need a simplification, I can only take in so much at a time. Is it a machine that takes inputs, based on the inputs decides a destination and if the same input is sent 3 times it quickens it's response? I was not able to quite understand how it works, but I think I can figure it out if you explain the problem

ok here is an attatchment of a diagram of my system i hope it worked. im just going to send it now. i can make a way more detailed diagram if you need it i mean way more. but i made this before i started it as a template