The ValuesLab works on deep integration of human values and computational intelligence. The lab uses the Socratic model, which assumes that both questions and answers typically are work in progress. Our work starts from the notion of human values which, as we conceive of it, involves ethical and moral values such as justice and respect for persons, but also epistemic values related to truth and careful thinking. We review and formulate conceptual frameworks, and give formative input on such notions as truth and falsity, lying, trust, justice, agency, linguistic communities, etc. A summary of points of connection between philosophy and AI development is here.
The ValuesLab’s PI is Katja Maria Vogt (Professor of Philosophy and Data Science Institute Affiliate, Columbia University). The lab, co-founded with Jens Haas, builds on earlier collaborative work and on three long-standing themes in Vogt’s research: the value of truth, value disagreement, and the Guise of the Good. Outputs are based on research in normative epistemology, ethics, philosophy of normative dimensions in logic and language, and philosophy of action. The lab is engaged in interdisciplinary approaches, as well as close collaboration with AI-leaders and policymakers in academic and non-academic entities.