Abstract: When starting a new collaborative endeavour, it pays to establish upfront how strongly your partner commits to the common goal and what compensation can be expected in case the collaboration is violated. Diverse examples in biological and social contexts have demonstrated the pervasiveness of making prior commitments on posterior compensations, suggesting that this behaviour could have been shaped by natural selection. In our recent work, using methods from Evolutionary Game Theory we have shown that this strategic behaviour can promote the evolution of cooperation in a population of self-interested agents playing the one-shot Prisoner's Dilemma if, on the one hand, the commitments can be sufficiently enforced, and on the other hand, the cost of arranging them is justified with respect to the benefits of cooperation. When either of these constraints is not met it leads to the prevalence of defectors/cheaters and commitment free-riders, such as those who committed but then dishonour their commitments, and those who commit only when someone else pays to arrange the commitments.

In this talk, two approaches that may circumvent such weaknesses of prior commitments are described. The first one is through using intention recognition a priori, where we show that if agents predict intentions of their co-players before an interaction occurs and arranging a commitment only when they are not confident enough about the predictions, the chances of reaching mutual cooperation in the agent population are largely enhanced. The second approach is to use costly punishment a posteriori, after the interaction occurs, to complement prior commitments as it allows to sanction defectors/cheaters when a commitment is absent or not formed.

Bio: Dr The Anh Han is a Senior Lecturer in the School of Computing and has been at Teesside since 2014. Before that, he was a postdoctoral research fellow at the AI Lab of the Free University of Brussels (VUB) and obtained his PhD at the AI Center of the New University of Lisbon in 2012. His current research interests span a wide range of topics within Artificial Intelligence and Multidisciplinary research, including dynamics of human cooperation, AI cognitive modelling, evolutionary game theory, agent-based modelling and knowledge representation and reasoning. He also collaborates with behavioural economists to perform experiments with real human subjects in order to validate theoretical models related to human cooperative behaviour. Within these topics, his research has resulted in a number of publications in high-ranking international scientific journals and top computer science conferences, including the journal of the Royal Society Interface, Nature Scientific Reports, Artificial Life, Adaptive Behaviour, IJCAI, AAMAS, and AAAI (see webpage: https://www.scm.tees.ac.uk/t.han/).

Add to my calendar

Back to listing