Friday 20.09.2024

Do we still have to think ourselves when algorithms can think better and faster? In this lecture I will explain the difference between situations in which more data and calculations are actually better, and those in which the reverse is true: less is more. According to the principle of a stable world, complex algorithms work most reliably in well-defined, stable situations whereas human thinking evolved in order to overcome the uncertainties of our surroundings with the aid of simple, robust heuristics. In uncertain, unstable situations a better prediction is often made on the basis of one single data point rather than big data. The danger is not that artificial intelligence will soon render human intelligence superfluous – that is still a long way off. The danger is that people stop thinking.

Prof. Dr Gerd Gigerenzer is a psychologist and former director of the Max Planck Institute for Human Development and since 2020 director of the Harding Center for Risk Literacy at the University of Potsdam. His popular-scientific books are international bestsellers. The Swiss Duttweiler Institute includes Professor Gigerenzer among the 100 most important global thought leaders.