AEIOU: aeio unit : affert expectation intention output unit
a *e *I o
Future a and o train *e and *i, respectively
a and e are continuous
*i and o are ballistic
A and e can modulate *i and o
Output trains intent!
And thus every module can coexist, synergize, and cross train at once. So it is with arbitrary modules. So it is with us!
Ultimately we intend because we do, not the other way around. It is only the nuance of priority and learning between the modules that indicates any difference at all. With repetition, it becomes true regardless.
So we stack all these same modules with varying specialties on top of each other in a network. And sensors, preceptors, motivators, coordinators, and actuators emerge. Within the unit there is no motivator. It starts with a few initial instinctual rules, later modules by the complex interaction of the modules learning and adapting.
I believe therefore that behavior can be modeled completely by a reasonable number of modules of symbolicly focused units that learn from one another by means of:
1. Input teaches forward expectation
2. Output teaches retroactive intent
