Creative Computers, Improvisation and Intimacy

Autonomous musical machine partners, ‘live algorithms’, are able to collaborate with human improvisers on an equal footing. Adaptability can be a significant factor in human/machine interaction in this context. Intimacy is an additional factor; intimacy might be achieved if human and machine performers can adapt to each other and learn from one another. Previously associated in computer music with ideas of embodiment and HCI, ‘intimacy’ as more widely understood, refers to the interpersonal process enjoyed between individuals, in which personal self-disclosure finds validation through a partner’s response. Real intimacies are learned over time, not designed, and are based upon an evident reciprocity and emergent mutuality. In the context of musical expression, a social – rather than a biological/technological –discourse can be applied to live algorithms with a capacity for learning. This possibility is explored with reference to the author’s various improvisation/composition systems including au(or)a, piano_prosthesis, and oboe_prosthesis.