On Monday 22 April, Enrico Liscio was the first PhD student from the Hybrid Intelligence Center to defend his thesis, entitled “Context-specific value inference via Hybrid Intelligence”. The premise of Enrico’s thesis is that in Hybrid Intelligence systems, agents will need to be aware of the norms and values of the humans they collaborate with. Enrico devised a method that allows the artificial agents to identify, classify and estimate those values from observations of the behaviour of the human agents in the system, as summarised in the diagram below. 

To validate his results, Enrico performed experiments in the domain of policy making (understanding the concerns of citizens), and in the domain of behaviour change (learning to live with diabetes).

The committee awarded Enrico’s PhD with a “cum laude”, a predicate that is reserved for the top 5% of PhD theses in the field.