Generative AI in a space of crafting and hacking

Repositioning technologies in a crafting and hacking space, seeing them fuel, change, or break familiar practices, has been a long-standing strategy used by artists: to unpack the creative possibilities of tools in their hands. The deviation from productivity, efficiency, and other technocentric concerns has constantly pushed the boundary of our relationship with algorithms, digital software, and more recently, AI systems.

As a researcher in music technology and an electronic music producer, Shuoyang Jasper Zheng studies AI systems as materials to craft musical interactions and expressions, to understand how artistic practices impact technological advances. In his PhD research, he looked at neural audio synthesis (a family of generative AI techniques that use deep learning models trained on recordings to generate new sound) through this lens of crafting and hacking, with a focus on enabling broader participation in technological advances, from sound and music communities.

In this talk, we will take a deep dive into neural audio synthesis, and how artist-researchers at the intersection of AI and the arts tweak, craft, and break these systems to explore their use in interactive sound and music systems. We will look at a range of creative applications that emerge in this crafting and hacking space, and implications for new notations and performance techniques with generative AI.