Rational agents: robots & artificial intelligence

The course will show what is needed to build up towards an artificial intelligent agent, in three steps (I-III). Before making those steps, we will start with a discussion of the basic components of agent architectures and the notion of "intelligence" (Prologue). The first step (I) is to develop "reflex agents" that can respond to the environment; this demands a discussion of perception & response, of "computation" and of the role of the body and the environment for intelligent agents. The second step (II) is a "planning agent" that can model ("represent") the environment and make a plan for action. This requires a discussion of representation, rational choice, goals, and search. The third step (III) is a "full agent" that can deal with uncertain environments, learn, and take responsibility for their actions. This requires a discussion of free will and consciousness. After these steps I-III have been completed, we will evaluate the ethical status of the full agent and the relation to humans, including the question of risk for humanity (Epilogue).

• Students have an understanding of the different approaches in Artificial Intelligence and their accompanying claims;
• Students have an understanding of the architecture of different types of intelligent agents and their possibilities and limitations;
• Students are familiar with the concept 'computation' and have some idea of the possibilities and fundamental limitations of computation;
• Student are able to use arguments in an essay on the (im)possibilities of Artificial Intelligence;
• Students are able to construct simple agents in a simulation environment (NetLogo);
• Students are inclined to critically analyze future scenarios for the application of Artificial Intelligence

Je moet voldoen aan de volgende eisen

• Ingeschreven voor een andere opleiding dan
• HBO-TOP Applied Physics, Schakelprogramma
• Alle onderstaande cursussen zijn behaald
• Social robots (0LEUB0)

• Articles
• Parts from: Stuart Russel & Peter Norvig, Artificial Intelligence. A modern approach, 4th Edition, Pearson Education 2020.