previous up

Seminar: Cartes Blanches (2021–23)
Human Music
Ilhan Zulji

Deep Learning has the potential to turn our visual perception upside down like no other tool since digitalization, not only for designers, the potential extends to all users of media. New questions from authorship to authenticity of what is seen or visual aesthetics force all viewers to question their own perception.

In my project I want to work with this technology. With the help of Machine Learning I create a complete music video in my own direction, without lyricist, producer, singer or film crew. I'm not only taking on new roles here, I'm also trying to question and reinvent my previous role as a creator. I am not only interested in Machine Learning as a tool, but rather as a helper and collaborator for a multimedial and audiovisual design.

From the generation of texts, the creation and manipulation of music to the visualization of the video, I use the various tools that Machine Learning offers. I take an active role as a designer. I control the outcome through the input I choose. My evaluation decides the selection of the generated output. In the elaboration of the project I decide when, where and how much influence Machine Learning gets.

I work with a lyric generator, Jukebox AI, Spade-Coco and Cinema 4D. I refer to pop culture references by reinterpreting the song "Seven Nation Army" by The White Stripes. I try to create a tension between the known and the unknown, the manmade and the generative, and the analog and the digital, both auditive and visually.