CAIMed took part in this year's Göttingen Night of Knowledge together with the Campus Institute for Data Science (CIDAS). Together we presented the exhibit “MENACE”, a mechanical computer consisting of 304 matchboxes. 

MENACE vividly demonstrates the principle of reinforcement learning: by repeatedly playing against visitors over the course of the evening, the system learned to constantly improve its strategies in the game of tic-tac-toe. 

Reinforcement learning is a sub-area of machine learning in which an algorithm develops optimal decisions step by step through trial and error and feedback. Such methods are not only used in games, but also in the control of robots, in autonomous vehicle guidance or in the optimization of complex processes. 

The joint exhibition offered interested guests an understandable introduction to this learning principle and opened up discussions about future applications of artificial intelligence - including in medical research. We see the direct exchange with the public as an important opportunity to convey complex concepts in a tangible way and arouse curiosity for data-driven research.