Algorithmic Prediction:
To Hope or to Trust?
Prof. Dr. Elena Esposito
Professorin für Soziologie an der Universität Bielefeld und an der Universität Bologna
Diskussionsleitung: Prof. Dr. Sven Nyholm
Professur für Ethik der Künstlichen Intelligenz, Ludwig-Maximilians-Universität München
»Machine learning algorithms promise innovative ways of dealing with the uncertainty of the future, but they are opaque and often incomprehensible. How do they affect society’s approach to the open future? Attitudes toward AI fluctuate between hope for great benefits and fear of epochal threats. According to Niklas Luhmann, hope and fear are complementary reactions to an unpredictable future: both reduce complexity and enable action even without guarantees. Trust, on the other hand, is granted knowingly and revocably; it promotes learning and increases complexity. As long as AI is considered an alternative form of intelligence, the tension between hope and fear seems irresolvable. The talk shows how this dilemma can be transformed into a controllable question of trust when algorithms are viewed as tools that expand the possibilities of prediction – without being intelligent themselves.«