An Articial Volition Architecture for Autonomous Robotics

Abstract

We introduce a new computational architecture capable of exhibiting an archetypal type of volition in a simplified modularized version of M. Minskys hive-mind. In this model, three relatively independent computational cores which themselves can also be whole multi-agent systems are engaged in an endless interaction each one representing the internal “imaginative” world, the external world interface and the arbitrator or Internal Observer. Volition then is expected to occur as the result of an endless antagonism for control between the internal and the external world models.