Electronic life and synthetic intelligent systems

AbstractThis paper examines the need for complex, adaptive solutions to certain types of complex situations recently typified by the strategic defense system but more aptly represented by systems proposed to manufacture complex biomolecules for diagnostic and therapeutic purposes. Since natural systems have evolved with capabilities of intelligent behavior in complex, dynamic situations, it is proposed that biological principles be identified and abstracted for application to certain problems now facing industry and defense.Two classes of artificial neural networks are presented—a non-adaptive network used as a genetically determined retina, and a frequency-coded network as an adaptive brain. The role of a specific environment coupled with the system of artificial neural networks having simulated sensors and effectors is shown to form a closed ecosystem. Evolution of synthetic organisms within the ecosystem provides a powerful optimization methodology for creating intelligent systems able to function successfully in any desired environment.The programmer/engineer interacts with the evolving system on a policy level, not at the level of programming details. This method ensures that fallible programmers are not creating brittle systems when issues of programming details and complexity would prove overwhelming. The resulting systems are expected to show adaptability to complex and changing situations as well as genetic improvement over time. Adaptive, intelligent software will replace brittle algorithmic programs in future applications to complex problems