Learning to Learn: A Reflexive Case Study of PRiSM SampleRNN

Description

Learning to Learn: A Reflexive Case Study of PRiSM SampleRNN

The emergence of neural audio synthesis technology has opened up many new creative and collaborative avenues for musical practitioners in recent years. With a growing number of software tools becoming openly accessible, many composers and sound artists start to map their music-making processes into a nebulous, data-informed collaborative framework. This often puts the practice of data curation, generative machine-learning models, as well as the artistic usage of machine-generated outputs into a state of play, whereby much of the idiosyncrasy of the resultant work is shaped by fine-tuning deep-learning algorithms. However, issues surrounding agency, distributed creativity, and access to computational resources / specialists tend to surface. This paper looks at these issues within the existing infrastructure of a Music Conservatoire, where to engage creatively and strategically with data and artificial intelligence tools becomes an increasingly important skill for artists to adopt outside their conventional musical training. Through the lens of the work of PRiSM (The RNCM Centre for Practice & Research in Science & Music) and the rollout of PRiSM SampleRNN between 2020-2022, we identify an emergent model of musical training and research that institutionally facilitates knowledge exchange and collaborative dialogues between practitioners, pedagogues, as well as research software engineers who are often not considered part of the existing conservatoire establishment.

Period	9 Sept 2024 → 11 Sept 2024
Event title	AIMC 2024
Event type	Conference
Degree of Recognition	International