sculptures as performers
concrete, plastic, Deep Learning, electronics, soundscape, videos
The concrete sculptures are singing a basically infinite number of songs generated by artificial intelligence (AI).
These sculptures are not fixed things. They exist in a state of potential. It is about relations of species without separation of the biological and the social in the context of “natureculture” thinking (Donna Haraway). What changes, and above all how does something new emerge, not necessarily the better.
The songs are generated with the help of an AI algorithm. This means that text and melody are generated automatically and the singing voice synthetically.
I train my AI software with pieces of music from the Internet that have been marked with the keywords “love, hate, life, death, ecstasy, drunkenness, euphoria”.
My music-making machines don’t really work any differently than any composer. First they learn about existing styles and then create similar technical reproductions. The AI recognizes patterns in the structure of a piece – for example the chords used, the notes of a melody, repetitions or variations. Based on these patterns, the AI recreates the “logic” of a musical genre and finally creates an infinite pool of new compositions, which are based on originals or consciously differ from them.
The AI is able to compose songs in any genre or style. The result depends only on the data material that has been supplied.