LSI is an experimental platform designed to explore the artistic potentials of generative AI within immersive spatial installations. Visitors are invited to step into a fluid audiovisual environment that can be transformed through 3D video projection mapping.

In a collaborative process, participants assist open-source generative AI models in augmenting the environment. Models generate artwork based on user input, while visitors modulate the resulting visual and auditory output via a touchscreen interface or through other sensor-based interactions.

By using LSI’s abstracted interface, people not only influence shaders of the video projection mapping but also play a sound synthesizer. Consequently, visitors become audiovisual artists, collaborating to create spatial works of art.

Today, many people communicate with generative AI systems in chat environments in which they send their commands to a “cloud”, which sends back an algorithmically generated image after a certain waiting time. This process is carried out using human language, which reinforces the impression that you are dealing with a human being to whom you are giving commands.

The abstraction of said process through a graphical user interface based on symbols and the realization of the concept of “co-creation” enables people to develop a new perspective on the topic of generative AI. Users become creators of a collaborative audiovisual experience instead of being mere clients of artworks.

Video projection as a medium of spatial design and influencing human perception of space with audiovisual means have been the central fields of research of our media art collective MO:YA since it was founded in 2011. We aim to create mesmerizing geometry that follows mathematical rules by merging modern technological research with classical principles of graphic and spatial design.

MO:YA (Roland Mariacher & Werner Huber) | Instagram