AI — a collaborator or a tool?

In the third day of KuvA Research days, several works done by artists with the help of AI were presented, raising – among others – interesting questions about the ‘authorship’ of the AI. I would suggest that the response that AI is a man-made tool, like artist’s brushes or musicians’ violins, tells less about the involvement of the AI and more about the interpretation of authorship, with all its benefits and responsibilities. In a market-driven art environment, crediting the AI as a co-author of the work is problematic from a financial standpoint, as well as regarding the already precarious position of an artist within neoliberal capitalist environment. Moreover, attributing authorship to the AI would create legal and ethical issues: in case if the artwork produced in collaboration with the AI proves problematic, the artist should address the issue and take responsibility, since the AI is usually trained in a society with inherent discriminatory bias and is prone to bring the issues of racism, sexism and ableism to the fore.

However, other artists, working with more traditional means, seem to be more open to the possibility of seeing their tools and the materiality they work with as agents in their work, as collaborators in the performative process of artworking. I refer to Barbara Bolt, but also Estelle Barrett, Lisa McCosh, Katve-Kaisa Kontturi, and Ellie Barrett, who have written about this issue academically, as well as artists who refer to their tools and materials as agentive participants and collaborators in speaking about their art. It seems strange that a paintbrush, canvas or even surrounding atmosphere could be seen as collaborative agents, but AI could not. I propose that perhaps this is due to the uneasy cohabiting of humans and AI; the subconscious (and often consciously voiced) anxiety that the AI could replace “us”. In the history, similar concerns have been raised about various types of technology humans have adopted. Photography and TV comes to mind readily, but already Plato expresses concern with the technology of writing – one that seems almost inseparable from our everyday life now.

I would suggest that this human/AI collaborative anxiety is actually already solved in practice: people who work with various types of AI have found a way to work around known issues and mutually enhance cooperation. We are already cyborgs, Donna Haraway said in 1985, and before her in 1930 Freud theorised of a human as a ‘prostetic god’ who depends on technology to surpass one’s ‘natural’ abilities. Inasmuch as there is an ‘authorship’ of artwork, it has always been collaborative – the materials and tools of all artists, be it paint or musical instruments or language have always been external to the person creating it and yet internalized as deeply personal. The difference between human and AI might not lie in the creativity and morals, or common sense – all of them are things that could be thought. It might lie in the obsessive need to claim personal authorship and centrality to the work which seems to be characteristic to (some) human artists, for personal, social, and market reasons. This is the need AI does not seem to have yet, and we can only hope that this is one feature it will not learn from us, instead remaining open and porous to outside influences and collaborators.

Ieva Melgalve, doctoral candidate, Art Academy of Latvia (LMA)