Co-improvising with the non-human: embedding algorithms inspired by group intelligence into semi-autonomous instruments for musical improvisation.

In this seminar, I will present my current artistic research into how non-human group intelligence and self-organisation can inform the design of multi-agent systems for musical co-improvisation. Central to my work is the Sophtar, a string instrument I developed that features an embedded system capable of running recent machine learning models and signal processing algorithms, automated electromechanical actuators, and controlled electroacoustic feedback capabilities. The instrument acts as a boundary object, a place of encounter between newly developed co-improvising algorithmic agents and human performers. I will introduce the instrument, its current algorithms and models, and those planned for experimentation during my residency at CASCB. Finally, I will outline the ideas behind a hardware co-improvising agent currently in the early stages of development.

Federico Visi is a researcher and musician based in Berlin, Germany. His work explores the interplay between human and non-human agencies in music, embodiment in networked music performance, and interactive machine learning. He has recently designed a networkable feedback string instrument with embedded machine learning called the Sophtar. He currently works as a researcher at Luleå University of Technology and he is a member of the TCP/Indeterminate Place networked hyperorgan quartet. He obtained his PhD at the Interdisciplinary Centre for Computer Music Research (ICCMR), University of Plymouth, and collaborates with research labs at several international institutions including University of Iceland, Goldsmiths University of London, and Universität der Künste Berlin.