Governance Node Seminar - Roel Dobbe, TU Delft

During this gathering, I would like to invite you to the principle of emergence in sociotechnical systems. This principle states that, due to the ontological indistinction of social, technical and institutional elements in AI or algorithmic systems, values like safety, fairness or justice are inherently emergent system properties. Put differently, these can only be understood across the interactions of technical, social and institutional elements of a system. A knife cannot be safe, neither can an AI model. In my research, I have explored what emergence means for how we assess safety, map algorithmic harms and work towards design interventions to build “safer sociotechnical systems”. As a result, I am increasingly convinced that mapping the complex relationships in sociotechnical AI systems should be a prerequisite for normative deliberation about ethical implications, forming a valuable bridge between those disciplines designing and deploying AI systems and those developing policies and institutional arrangements to ensure their legitimacy and responsible/safe use.

Some relevant papers:

System safety and artificial intelligence

Hard choices in artificial intelligence

Bio: Roel Dobbe is an Assistant Professor working at the intersection of engineering, design, and governance of data-driven and algorithmic control and decision-making systems. Roel holds a PhD in Control, Intelligent Systems, and Energy from the Department of Electrical Engineering and Computer Sciences at the University of California Berkeley. He was an inaugural postdoc for the AI Now Institute at New York University. Roel has gained extensive experience working with industrial, public, and societal partners, through working as a consultant, a data scientist and doing engaged scholarship.