Situated co-creation: Configurations of artistic practices with generative AI for arts education

Abstract

Beginning with an exploration of the potential of generative AI (gAI) for arts education, I examine how co-creative practices with gAI are configured in contemporary art through techno-material entanglements and artistic attitudes. Rather than anthropomorphising AI or reducing it to a passive tool through anthropocentrism, I understand co-creative practices with gAI as network practices composed of human and non-human elements, as outlined in Bruno Latour's (2005) actor-network theory. Using situational analysis (Clarke et al 2022), I investigate four exemplary artistic positions by examining how agency is negotiated across discursive and praxeological dimensions, as well as through the techno-material affordances of generative networks. In doing so, I develop a typology of artistic attitudes: attributed control, affirmative release of control, loss of control, and shared control. Rejecting the tool metaphor, I tentatively propose alternative terms for AI as connectionist counterparts and collaborators, depending on the praxeological distribution of control. The study is limited by its focus on four case studies, which provide conceptual orientation while risking the production of overly stable distinctions. Taken together, these entangled attitudes towards control offer a condensed account of situated co-creative practices with gAI and an analytical framework for encountering gAI in arts education.