A few months ago, a painting made by a machine learning algorithm has sold for 432,500 dollars in a Christies auction. The collective that programmed the ai, suitably named “Obvious”, used code written by another artist as a starting point to train their own neural network, this mere fact, raises a number of questions about the legal, ontological and discursive status of this artwork. Artificial intelligence in one sense or another has been fuelling the imagination of artists for centuries, from Marry Shelley's Frankenstein to Spike Jonze’s Her. The possibility of synthetic beings and the consequence of their existence, fills our imagination as it confirms the most obvious existential risk for humans — refuting the idea of human exceptionality.
Art has been for centuries the very fulcrum of human exceptionality. As manifest in romantic theories of authorship, the artist as a genius and the work of art as autonomous from reality. As it emerges from an imaginary universe which is specific to a certain artist and in historical and contemporary discourse about the position of artists as ‘public intellectuals’, those that can speak truth to power and manifest critical virtue.
What can art offer in the age of machine learning, is not just its explicit description, the fact that is being made by intelligent systems, but the various questions that it raises about how do we understand what being an artist really means today, and about the notion of artificiality in relation to intelligence.