Non-Linearity, Autonomous Agents, and Warfare

This proposal details an approach to quantifying the interrelationship between parts of non-linear complex systems. The approach builds on years of work at Intelligent Automation, Inc. in the use of what we call fine-grained autonomous agents to control and simulate complex non-linear systems. It also builds on extensive work we have done in analyzing deterministic chaos in non-linear systems. Our work in autonomous agents suggest Measures of Effectiveness which identify where small adjustments in a system can result in large improvements in resource allocation, schedules, target assignments, etc. Together with our work in chaotic systems, we will also suggest how to adjust systems to either reduce or increase the probability that the system's non-linearity will result in chaotic behavior. Our "fine grained autonomous agent" technology is very different from the now popular concept of a software agent. Our agents are "fine grain" agents. Their knowledge is very limited to one particular aspect of the problem, and they bid and negotiate with other agents to achieve their local goals. The global behavior of the system will emerge from the interaction of these agents. The conjecture on which the specific work herein proposed is based links the interactions of agents with the sensitivity of the global system to small changes. If this conjecture proves true, it could have important practical and theoretical implications.