Programme, Tuesday 13 September 20:00 @ IKLECTIK, London
Find the full text score anthology at http://milker.org/s/The-Text-Score-Dataset-10-Jennifer-Walshe.pdf
Audience text pieces generated using OpenAI GPT-3
Thoughts from an AI
Thoughts from Megan
With thanks to Jennifer Walshe, Ragnar Árni Ólafsson, David De Roure and PRiSM, RNCM.
Performing a text score is a process of decomposition: you unpick the structure, form, syntax, lexicon, imagery. You make a decision whether to approach it literally or figuratively. As an ensemble, make decisions about how far to align those individual approaches (if at all). You consider the social, historical, political and spiritual contexts. A text score is a message from the composer you’re trying to detangle through performance. The composition of a text score is an incredibly human thing – so what does it mean when a computer has composed it?
Each performer chose a piece from Walshe’s The Text Score Dataset 1.0 to feature in this gig. There were many more we wanted to perform. Getting audience members to create spontaneous new scores is a way to show the potential of these technologies. As well as creating new art they’re creating new means of communication, learning and community engagement with the arts.
My PhD project, at Royal Northern College of Music, Living Instruments, Universal Composition: New Works and Music Processes for and by Disabled Musicians is exploring the accessibility of experimental music processes for disabled, Deaf and neurodivergent performers. Text scores are accessible in their use of language over music notation or systems, and have a history of encouraging music/sound-making by amateur and community music groups. AI could further the accessibility of these processes.