1st CAUSE School Workshop at the Computer Science Day @UniOldenburg
24 June 2025
On the morning of the 24th June 2025 two representatives of the CAUSE research group, Moritz Buhr and Ivo David Oliveira, brought MENACE to a group of highschoolers. Unlike what the name suggests, MENACE is not a violent threat to one’s safety, instead it stands for “Matchbox Educable Noughts and Crosses Engine”. It is a playful and tactile demonstration of how reinforcement learning works with nothing more than match boxes and colorful beads.
MENACE was originally invented by Donald Michie, a British computer science researcher, in 1963. The purpose: to illustrate the, back then, new field of AI. His debut to the wider world was then published in The Computer Journal in November of that year under the title “Experiments on the mechanization of game-learning Part I. Characterization of the model and its parameters”.
Since then, much has been written about MENACE and still today it, and its variants, serve as a great, if not the best, introduction to these very concepts to curious minds on how this mysterious field of machine learning works. Inspired by this, CAUSE has been developing a didactic demonstration by making use of MENACE to illustrate core concepts of Machine Learning and inspire students to explore the field of computer science. And, what better opportunity to test our material than the open door event “The Computer Science Day” at the University of Oldenburg?
On the morning of the 24th of June many games were played and the demonstrations followed by explanations gave multiple high-school students the chance to physically interact with our RL tic-tac-toe machine. People were very interested, mainly because it gave them a very clear understanding of RL in a way difficult to find in other environments, plus, many related topics came up that had great educational value such as probability theory, adversarial learning amongst others. In summary it was a success!
Overall the demonstration was not only successful but it gave us a chance to learn ourselves how we could better adapt our material for the classroom. Thus, our next target is just that, make this more accessible and rounded for larger groups for our next iteration.