Talks:
Michael Rovatsos
Making smart societies fair
Abstract
Collective intelligence systems that exploit digitally mediated interaction among large numbers of users have the potential to combine human and machine intelligence in novel ways in order to solve complex problems that are far beyond the abilities of humans or computers alone. However, “smartness” alone does not guarantee that these systems will have a positive impact on human society, as many recent phenomena involving social media, sharing economy, and crowdsourcing technologies show. This talk will present insights on the ethical issues of smart societies we have derived in recent work, and discuss both our own contributions and ongoing efforts in developing “fair” algorithms, as well as the broader research landscape and challenges surrounding these issues.
About the lecturer
Michael Rovatsos is a Reader and Director of the Centre for Intelligent Systems and their Applications at the School of Informatics, where he has also been leading the Agents Research Group since 2004. He received his PhD in Informatics from the Technical University of Munich, and his track record includes over 85 papers in AI that have attracted over 1200 citations to date. His research is in multi-agent systems, with a particular focus on automating methods for reasoning about interaction among intelligent (human and artificial) agents. He has been involved in externally funded research projects worth over €10m as Principal and Co-Investigator, is an Associate Editor for the Knowledge and Information Systems journal, has been involved in the organisation of over 80 international scientific events, and regularly serves on the committees of most major AI conferences.