Deliberate evolution in multi-agent systems

Abstract

Artificial societies of agents make it possible to experiment with deliberate evolution processes without ethical dilemmas. Moreover, interesting application areas have emerged: applications related to the Internet which can highly benefit if societies of agents can evolve automatically, and if within such an evolution process the agents can deliberately influence the direction of the evolution. An architecture of agent capable of deliberately creating or modifying agents was designed and implemented. The factors needed for this architecture are the model for deliberation and the model for effectuation.