Employment Decisions Supporting Organizations of Autonomous Agents

Abstract

A key aspect of organization is that one or more agents perform tasks at the direction of another, supervising, agent. We model a set of individual actors using intelligent agents. These agents self-interestedly choose products to produce and to consume. Our initially-identical autonomous agents self-interestedly form organizations, with the employing and employed agents performing separate value computations. Use of agents, rather than statistical demand and supply functions, enables direct computation of every action by every agent, permitting the tracing of activity to the boundedly rational behavior of intelligent agents. Actions which incur costs, such as choosing others with whom to transact, that are often modeled using equation-based quantities in other work, are here performed by the computer code needed to achieve the desired result. A key feature of this work is that these self-reflective agents balance thinking about what they will do with actually doing it.