Researchers create computer art

(07-02-2023) Using a prompt and their own programming language, a dozen UGent researchers created a work of art every day for the past month. Their results were given a daily hashtag on, a digital platform around artistic programming.

Creative with prompts

From the rather playful 'Black and white' and 'Made in 10 minutes' to abstract concepts like 'Suprematism' and 'Asemic'. All examples of prompts, broadly interpretable tasks that research assistant Michiel Rollier and PhD students Tim Van Wesemael and Thomas Van Giel, among others, have been working on regularly over the past January days.

Using such prompts, these scientists, all three affiliated with the Department of Data Analysis and Mathematical Modelling, spent a month looking for alternative ways to incorporate more creativity into their scientific research. The original idea for this extra push to came from postdoctoral researcher Michiel Stock and colleague Michiel Rollier.

On, the aim is to use written codes to generate a creative output, which is called generative art. Generative in the sense that the creation in question, by intervening in the source code of scientists worldwide, will look different every time. Instead of paint and brushes, generative artists use their programming skills to create visual or auditory art using the computer as a canvas.

Aesthetic digital results are generated thanks to commissioned algorithms. The conditions for participating (computer) scientists and (media) artists are the prompts imposed on

Generative art as an educational tool

It is Michiel Rollier, researching 'cellular automata' under the supervision of Professor Jan Baetens, who is particularly keen on this distinct form of art. Indeed, his experience with mathematical models in which imaginary spaces are divided into compartments that communicate with each other by pre-programmed codes lends itself easily to producing a generative aesthetic output.

Based on simple rules, an output is simulated that looks exactly like it ran away from nature. Among other things, this allows the idea that life consists of complex structures arising from simple rules to be playfully aestheticised. In this way, the use of art helps communicate abstract research.

The strength of lies mainly in making programming itself attractive. That it also makes communicating scientific research, as in the case of Michiel Rollier's cellular automata, a lot more appealing is a nice touch. makes concepts such as sampling, dynamic systems, geometry and image processing, topics covered in various courses at different faculties, very insightful. For instance, Bram Spanoghe delivered the prompt 'digital plants' thanks to his PhD on L-systems, and civil engineer Thomas Van Giel was able to playfully illustrate various mathematical concepts with the assignment 'Signed Distance Function'.

With Instagram to the GUM?

'gebroken' ijsvogel (Tim Van Wesemael)In addition to various contributions to, emerging generative artist Michiel Rollier is currently putting the finishing touches to a collection of artwork, with a paragraph of explanation per artist, which is due in mid-March 2023. The starting point here is the Instagram page @ugentgenerative on which generative work by various UGent scientists can be found via the hashtag #genuary.

Originally, it was mainly mathematician Sofie Van Gassen and electronics engineer Francis wyffels who started working with codes on this social networking site. Not much later, young researchers such as Tim Van Wesemael, Thomas Van Giel, Steff Taelman and Wissam Barhdadi quickly joined in.

Even staff from the Faculty of Engineering and Architecture (Dries Marzougui, Tom Neutens and Victor-Louis De Gusseme) and Koen Van den Eeckhout from the Faculty of Science also made several Instagram contributions in the meantime. All source codes of these will soon be shared via Github.

Art, then, as an exquisite way for both researchers and a wide audience to think about science. Just recently, biotechnologist Marjan De Mey presented her endorphin-producing bacteria and nowadays many follow ecophysiologist Kathy Steppe's twittering trees.

In the same vein, all UGent participants hope to demonstrate with their generative art, above all, that (computer) science does not have to be dry matter, but rather fundamentally creative. And meanwhile, Michiel and his team are already dreaming of a guest curatorship over a corner of 'Generative Art' at Ghent University Museum.

Go to ugentgenerative