The IDEA of Us: Engineering Socially-Aware AI Systems
                    
                    
                    
                        
                            Artificial Intelligence (AI) systems often require interaction and cooperation with humans to achieve their goals.  However, human behaviour is uncertain and complex, and so it can be difficult to reason about it formally. Specifying, designing, implementing, and deploying AI systems able to cooperate with humans is challenging but crucial for reliability and trustworthiness of those AI systems.  In this talk, I will present IDEA: an adaptive software architecture that enables cooperation between humans and autonomous systems, by leveraging in the social identity approach. This approach establishes that group membership drives human behaviour. Identity and group membership are crucial during emergencies, as they influence cooperation among survivors. IDEA systems infer the social identity of surrounding humans, thereby establishing their group membership. By reasoning about groups, we limit the number of cooperation strategies the system needs to explore. IDEA systems select a strategy from the equilibrium analysis of game-theoretic models, that represent interactions between group members and the IDEA system. I will show how this approach can extend AI systems’ capability to enhance the resilience of local communities and emergency services in the aftermath of a mass emergency in an urban area by facilitating their effective cooperation.
                        
                    
                    
                        
                            
                                
                                     Friday, May 24, 2024
                                    @13h00
                                
                            
                            
                                3195, Pav. André-Aisenstadt
                                    Friday, May 24, 2024
                                    @13h00
                                
                            
                            
                                3195, Pav. André-Aisenstadt