Two sides of the same coin
In the theatre the curtain falls, the lights go up, and the audience applauds with gusto. The applause itself begins as a deafening noise, with hundreds of hands clapping individually and all at once. But after just a few seconds, things change: a rhythm emerges from the chaos. Suddenly, everyone in the audience is clapping together in time. Who gave the order for this to happen? No one. The audience has synchronised itself.
Physicists call this process “self-organisation” – and it is a natural phenomenon. Irrespective of whether it is people applauding, or thousands of birds flying across the sky without colliding with one another, or billions of drops of water forming a hurricane: when lots of small parts coordinate themselves, a new order emerges.
This basic principle is something which interests not only researchers in the field of Physics but also scientists dealing with artificial intelligence (AI). In a new “Topical Programme”, the University of Münster is pooling a variety of research strategies under the heading of “Artificial Intelligence and Complex Systems”. There exist close conceptional and methodological connections between the two areas. As Dr. Oliver Kamps, Administrative and Scientific Director of the Center for Data Science and Complexity (CDSC) points out, “Research into complex systems and the development of artificial intelligence are two sides of the same coin. Principles found in nature inspire the development of AI. In return, AI helps us to understand complex natural processes.” The Center serves to build up scientific networks at the University of Münster in the fields of data science, complex systems, artificial intelligence, machine learning and mathematical modelling.
The collective behaviour of individual parts, such as swarming birds, can lead to organised “intelligent” behaviour. The human brain is another example of a complex system, consisting as it does of around 90 billion interacting neurons. “AI attempts to replicate certain intelligent feats which the brain carries out, such as image or speech recognition,” says Kamps. “Basically, the aim is to learn complex relationships between data points.” A good AI model will, he says, reduce the immense complexity of a reality to a level that remains mathematically manageable and machine-readable, while still retaining the essential features. The focus which researchers have in the Topical Programme is on continuing to build the bridge between Physics and AI.
Research in the field of AI is multifaceted and is gaining in importance in almost all fields of research and areas of life. Combining AI and Physics earned John Hopfield a Nobel Prize in 2024. He showed how, using simple principles of physics, computational models could be constructed which recall the ability of the brain to learn and remember. Without his ideas, many of the things which artificial intelligence is capable of today – for example, facial recognition, language recognition or recommendation systems – would be inconceivable.
Author: Hanna Dieckmann
Background: Topical Programmes
With its Topical Programmes, the University of Münster opens up themes which are suitable as future focuses for research and gauges the opportunities for large-scale projects to be undertaken in research alliances. The aims are to strengthen networking with external scientists and other (non-)university research institutes and to promote international collaboration. Advanced initiatives can be given support up to a level at which funding applications can be made. The “AI and Complex Systems” programme is being funded by the University to the tune of around 40,000 euros. Nine researchers from CDSC are involved. A digital “knowledge base” is being built up to push the integration of topics, as well as to train students in these areas. This knowledge base is an interdisciplinary platform which makes available central concepts and methods used in the field of the topic in the form of tutorials and code examples.
This article is taken from the university newspaper "wissen|leben", issue no. 2, 1 April 2026.