Neural Basis Function for Artificial Intelligence

The first step in the application of the ANTS paradigm was the development of a patented conceptual model, called a neural basis function, which combines the capability of autonomous and collective interaction, or bi-level intelligence, for each component, subsystem, or agent. Individual operational components connect to other operational components and to heuristic level decision making components through a specially designed interface. This structure is redundant at each level.

Every ANTS WORKER has a core heuristic and autonomous genetic code tapped into using genetic algorithms. Heuristic genetic algorithms buld a high level heuristic neural system in response to sensory input, via a self-assessment loop, using a combination of fuzzy logic and neural net approaches. This results in the decision making response in the "What do we want to do" mode. Meanwhile autonomous genetic algorithms build a low level autonomous neural system in response to sensory input, via a self-assessment loop, using non-linear dynamics. This results in the action response in the "right foot, left foot" mode. The interface between the higher and lower level modes is an essential feature of this design, allowing requests for input or output to be easily passed and implemented, and the resulting behavior, capability, and goals to evolve in response to the environment.