Ethics of autonomous information systems towards an artificial thinking

Abstract

Many projects relies on cognitives sciences, neurosciences, computer sciences and robotics. They concerned today the building of autonomous artificial beings able to think. This paper shows a model to compare the human thinking with an hypothetic numerical way of thinking based on four hierarchies : the information system classification, the cognitive pyramid, the linguistic pyramid and the digital information hierarchy. After a state of art on the nature of human thinking, feasibility of autonomous multi-agent systems provided with artificial consciousness which are able to think is discussed. The ethical aspects and consequences for humanity of such systems is evaluated. These systems lead the scientific community to react.