The advent of high-quality autonomous AI systems raises a number of questions on how these systems should decide and act when released into the wild. To avoid disasters, solutions are being proposed that aim to ensure that the use of AI in business and society is able to align both individual and societal preferences and norms, avoiding thus negative disruptions.

This ambition is captured by a booming new field of AI research that has the intention to develop cooperative or prosocial AI systems, a task which is especially difficult when the participants, being human or AI, may have conflicting interests.

Starting with an introduction on game theory and the question of cooperation, this presentation will zoom in on a recent study we made wherein people delegated their strategic choices to simple agents in a game capturing the problem of climate negotiation. Will humans or agents produce better group outcomes? If you like to know more, join this seminar during the Belgian AI week.

Details

11:35 : KEYNOTE

12:10 : Q&A

11:25 : LAST QUESTION

12:28 : CLOSING - INTRODUCING NEXT SESSION & RELATED EVENTS

-----------------------------------------------------------

For more info & joining live session: Event details.

-----------------------------------------------------------

The Belgian AI week - 15/03/2021 -> 19/03/2021

Discover also the detailed program of the week

Découvrez aussi le programme détaillé de la semaine

Ontdek ook het gedetailleerde programma van de week