Computer software designed for the arts (design, illustration, cinema, music, fine art, performance, etc.) is now a mature, multi-billion-dollar global business. While computer systems support the production and organisation of creative outcomes effectively and extensively, the vast majority of these systems are founded on the assumption that the users – not the software – supply all the imagination in the evolution and performance of a work. Software is considered an inert tool that should passively support an artist’s inventiveness, but never intervene as a creative contributor. In short, the computer’s involvement has been conceptualised as reactive, not creative.
What if the computer could act more like an equal, or at least as an inspirational assistant, understanding an artist’s own style and abilities, interacting with them as they generate new art? The computer could then challenge the artist to forge new ideas, and support or critique them as they emerge – akin to working with an artistically sensitive human partner who understands and buoys one’s own originality. This system would encourage learning and practice by adapting to each individual’s current style and technical abilities, developing in sophistication and empathy just as the human artist’s own abilities mature.
The focus of this ARC funded research project is in designing and building computational systems for improvisation. Improvisation is one of the most demanding yet rewarding creative acts. While well studied in human psychology, only recently has the idea of people improvising with machine partners gained serious attention. The aim of this project is to advance the state-of-the-art in human-machine improvisational creativity.
This project is on going. More information can be found on the project web site.