Electronic brain model with neuron reinforcement

Abstract

A method of emulating the human brain with its thought and rationalization processes is presented here, as well as a method of storing human-like thought. The invention provides for inclusion of psychological profiles, experience and societal position in an electronic emulation of the human brain. This permits a realistic human-like response by that emulation to the people and the interactive environment around it.

1 . An electronic brain model based upon a set of electronic neurons, each neuron optionally associated with a parameter, and taking on a parameter value from 0 to 100%, the model comprising:
 a context pool memory containing contextual data and a set of neuron pointers, each neuron pointer addressing a neuron corresponding to a present situational context;   a reinforcement memory containing a set of neurons corresponding to experiential relationships;   a long term memory containing a set of neurons corresponding to experiential relationships and committed to long-term storage; and   an analyzer and correlator that creates new neurons in the reinforcement memory and promotes at least one neuron in the reinforcement memory to permanent memory based upon a predetermined activity level in the at least one neuron.   
   
   
       2 . The electronic brain model of  claim 1 , wherein the analyzer and correlator removes a neuron pointer from the context pool memory after a firing level of the neuron addressed by the pointer falls below a predetermined level. 
   
   
       3 . The electronic brain model of  claim 1 , wherein the analyzer and correlator removes contextual data from the context pool corresponding to neurons whose associated neuron pointers have been removed from the context pool memory. 
   
   
       4 . The electronic brain model of  claim 2 , wherein the context pool memory is sized to accommodate the number of neurons and an amount of contextual data required for operation for approximately one day. 
   
   
       5 . The electronic brain model of  claim 1 , wherein the analyzer and correlator reduces a firing level of a neuron substantially continuously with time and increases a firing level of a neuron when the neuron is referenced. 
   
   
       6 . The electronic brain model of  claim 5 , wherein the analyzer and correlator removes data from the context pool memory that is associated with a neuron whose firing level has been reduced below a predetermined threshold. 
   
   
       7 . The electronic brain model of  claim 1 , wherein the analyzer and correlator reduces a firing level of a neuron as a function of time and increases a firing level of a neuron when the neuron is referenced. 
   
   
       8 . The electronic brain model of  claim 1 , wherein the context pool has a size corresponding to a number of neurons firing and an amount of contextual data in a wake/sleep cycle, the wake/sleep cycle being about one day long. 
   
   
       9 . The electronic brain model of  claim 1 , further comprising an index containing pointers to all electronic neurons of the brain model and operable to reuse indices when a neuron is permanently deleted. 
   
   
       10 . The electronic brain model of  claim 9 :
 wherein the index comprises a set of unique serial numbers; and   wherein each neuron in the reinforcement memory has a unique serial number having a first sign associated therewith and each neuron in the long term memory has a unique serial number with a second sign associated therewith, the first and second signs alternately occupying a bit collocated with the unique serial number.   
   
   
       11 . The electronic brain model of  claim 1 , further comprising a clutter filter operable to accept input from multiple sources and determine whether such input is placed into the context pool based upon predetermined criteria. 
   
   
       12 . The electronic brain model of  claim 1 , further comprising a clutter filter operable to accept input from multiple sources and determine whether such input is placed into the context pool based upon configurable criteria. 
   
   
       13 . The electronic brain model of  claim 1 , further comprising a clutter filter operable to accept input from multiple sources and determine whether such input is placed into the context pool based upon criteria expressed as a parameter associated with one or more neurons of the brain model. 
   
   
       14 . The electronic brain model of  claim 1 , further comprising a clutter filter operable to accept input from multiple sources and determine whether such input is placed into the context pool based upon a temperament associated with the brain model. 
   
   
       15 . The electronic brain model of  claim 1 , further comprising a clutter filter operable to accept input from multiple sources and determine whether such input is placed into the context pool based upon personal preferences associated with the brain model. 
   
   
       16 . The electronic brain model of  claim 1 , further comprising an event memory containing electronic neurons held in waiting for a predetermined event. 
   
   
       17 . The electronic brain model of  claim 1 :
 wherein a plurality of propensities to a plurality of traits are specified based upon at least one of choleric, melancholy, sanguine, and phlegmatic temperaments, and combinations thereof, each temperament being specified by at least one parameter in the electronic brain model; and   wherein upon occurrence of a predetermined, traumatic event, the plurality of propensities reverts to a predetermined temperament, the predetermined temperament including at least one of choleric, melancholy, sanguine, and phlegmatic temperaments, and combinations thereof.   
   
   
       18 . The electronic brain model of  claim 1 , wherein the parameter value associated with at least one of the electronic neurons is derived from the parameter values of a plurality of other electronic neurons. 
   
   
       19 . The electronic brain model of  claim 1 , wherein the parameter value of at least one of the electronic neurons is directly accessible from outside of the electronic brain model.