Emotions and a Prior Knowledge Representation in Artificial General Intelligence

Article English OPEN
Gavrilov, Andrey (2008)
  • Publisher: Institute of Information Theories and Applications FOI ITHEA
  • Subject: Emotions | Neural Networks | Knowledge Representation | Hybrid Intelligent Systems | Artificial Intelligence | Cognitive Simulation

In this paper a prior knowledge representation for Artificial General Intelligence is proposed based on fuzzy rules using linguistic variables. These linguistic variables may be produced by neural network. Rules may be used for generation of basic emotions – positive and negative, which influence on planning and execution of behavior. The representation of Three Laws of Robotics as such prior knowledge is suggested as highest level of motivation in AGI.
Share - Bookmark