Social Ontology for the sake of Trustworthy AI
In my talk I will focus on social ontology and try and show how modelling social entities is not only useful but necessary for having trustworthy AI systems that interact with humans in technologically dense and risky environments. After a brief introduction on applied or computational ontologies, I will start with presenting some classical works of my lab on modelling roles, organisations, group agents and collective intentionality, which leverage well known philosophical theories and techniques, like reification, social choice theory, judgment aggregation, qua entities theory, variable embodiment and radical interpretationism. I will then turn to some open issues, like mereology and location of social entities and the relationship between the metaphysical perspective of philosophical ontology and the conceptual and pragmatic point of view embraced by our approach of applied computational ontology. I will conclude by proposing sociological studies in sociomateriality as a theoretical lens to look at such issues under a new light.
June 20th / 18:35 / Aula magna