Every Lecture Gets a Film Poster

This semester, every lecture of Inf-Einf-B, the introduction to computer science at the University of Bamberg, gets a film poster. It is drawn by a pen plotter, a machine that moves felt-tip pens across the paper.

David J. Malan teaches CS50 at Harvard, an introduction to computer science whose lectures are freely available online. In them he tears a phone book in half to show how to search a sorted list, and he brings volunteers on stage to act out an algorithm with their own bodies. This is called theatrical teaching.

Inf-Einf-B builds on CS50, and I think Malan’s theatrical teaching is a good approach. Props make students curious about the subject, including those who have had little to do with computer science so far. They also give students a picture on which to hang a suitable mental model: if you have seen the phone book tear down the middle and one half put aside, you already have an idea of binary search before the first line of code appears on the projection screen.

In the first weeks of this semester, a remote-controlled robot dog walked through our lecture hall. A pyramid of ten cardboard cubes stood on the table, and letter cards hung from a clothesline. When the card with the end-of-text marker was taken down, the program kept reading past the end of the word.

So every lecture gets a film poster. Its title is also a term from the lecture: the film for the lecture on nested loops is called VERSCHACHTELT (German for “nested”; Schachtel means box), and the poster shows the pyramid of boxes. Above the title sits a tagline, as on cinema posters, and below it a cast list with the week’s props and an age rating with a joke from the course. Film 000, for lecture 0, is “rated 0 and up”.

The posters do not summarise slides. They are meant to make sense to those who were in the lecture, and to make everyone else wonder what happened there.

The plotter can only make areas by hatching, so from a distance you see areas and up close individual lines. Depending on the motif, one poster takes 40 to 110 metres of line. The video above shows what that looks like in time-lapse.

The program that turns the description of a motif into files for the plotter was written by Claude Code, an AI tool that writes and runs program code on its own. Motifs, titles and taglines come out of conversations with the tool. I decide what goes on a poster.

The further posters will not appear as separate posts on this blog but on a page of their own that grows over the semester. For each poster it says what happened in the lecture.

I cannot say whether the posters help students learn. Which props have worked in other courses? I would be glad to hear examples.