A monumental LED screen covers an entire wall of the crowded NY MoMA atrium lobby. A moving image artwork morphs through endless compositions of pleasing formal abstraction, never repeating, yet always feeling familiar. The AI-generated video titled Unsupervised is an artwork by Refik Anadol commissioned by MoMA and is based on a large dataset of images from MoMA’s collection. It attempts to synthesize what modern art “looks like” by looking at all modern art and, like Borges’s totalizing library, tantalizingly suggests that every image ever made is within the AI’s grasp. Yet, as a spectacle of computation, it is unclear if the frothing sea of data holds any other meaning.
The museum text opens with a vacant enthusiasm: “What would a machine dream about after seeing the collection of The Museum of Modern Art?” The machine itself is not to be seen, replaced with a twenty-foot digital canvas which vividly displays the AI’s purportedly churning subconscious. Documentary videos of the work feature a view of the gallery empty of the normal crowds, save a lone spectator gazing upwards towards the glowing wall. The posed art viewer is dwarfed by the display, with their back to the camera, every bit a nerdy millennial stand-in for Caspar David Friedrich’s voyeur in the “Wanderer above the Sea of Fog”. The painting, an oft-used shorthand for the Kantian sublime, is also a source of internet memes which Annie Dorsen thoughtfully analyzes in her essay The Sublime and The Digital Landscape. Dorsen argues that the vastness of information on the internet has become the modern substitute for the natural landscape of the nineteenth century – “vast, unknowable, awe-inspiring.” In this way, Anadol’s Unsupervised, which purportedly processes 138,151 pieces of metadata from MoMA’s vast collection, also relies on big data spectacle that Dorsen names the “digital sublime”. Instead of the monumental vista of distant mountains, it is a vast mountain of data with infinite computational permutations.
Dorsen also describes Kant’s dynamic sublime arrived at through the apprehension of the terrific, which, for Anadol, is also lurking just under the surface. A subject who is just barely removed from the destructive power of natural forces (say earthquake, avalanche, or fire) may experience serene sublimity–so long as a thin insulation remains between viewer and natural catastrophe. This becomes a digital dynamic sublime in Anadol’s work, which is only at arm’s length from worrying existential threats predicted by the advancement of AI, thinly insulated in the safe confines of the museum white cube. The fear of job replacement, as evidenced by the 2023 writer’s strike, or the specter of total human extinction in dystopian films such as the 2014 Transcendence are only a few examples of the popular dread in circulation. Art and design professionals entranced by the infinite computability of Anadol’s work may tingle with distant dread at their own possible obsolescence.
Concerns such as these seem largely absent from Anadol’s work, which does not take any specific critical position on AI technology. Yet by paying close attention to its formal presentation and the museum text that gives it context we can infer a certain willful optimism. It also indicates troubling complicity with an AI industry of utopian accelerationist entrepreneurs, such as Sam Altman, who assume the predestination of general artificial intelligence, with radical disruptions that are only beginning to emerge..
Moreover, the artist betrays a preoccupation with the technical operation of the software tools themselves rather than any cultural exigency. There is repeated emphasis on the size of the data and the computational flex. Not surprisingly, Anadol’s own website boldly features an endorsement not from a cultural critic but from a computer scientist employed by GPU maker nVidia who is “eagerly looking forward to seeing what we can do together in the future”, suggesting that the work may also be a powerful technology demonstration for the computer graphics industry. In the long history of computer graphics demonstrations there is one most beloved and ubiquitous–the screensaver. In this light, Anadol should not bristle when critics such as Jerry Saltz claim that the artwork is a “million dollar screensaver”, for the artist extols their work in very similar terms.
To this criticism, Anadol’s fallback is an appeal to authenticity, retorting on X, “I create my work from my heart!” Wan Yii Lee argues that the claim to artistic sincerity is a well-loved strategy for those with large social media followings, as it immediately cancels conversation and villainizes critics as elitist gatekeepers. Yet the accusation of gatekeeping feels far-fetched here – an institutional exhibition at the NY MoMA puts Anadol firmly within the gates. Saltz, a perennial voice of advice to the art school crowd, has said, “Everyone is sincere. Even Jeff Koons.” One may read here that sincerity alone is not sufficient. Contemporary artistic practice has long since abandoned the virtues of artistic expression and individual genius in favor of work that is about something, that aims to make meaning and conversation within an increasingly complex world. But like most screensavers, the work Unsupervised seems to have no subject, or rather the only subject is the spectacle of its own production.
Artificial intelligence software cannot make meaning for the artist any more than the advent of paint in tubes, yet similarly AI has the potential to be a powerful co-productive tool. The most urgent challenge for artists working with technology is to proactively determine how their practice functions either in cooperation with or antagonism to the forces of technology and capital. There is a greater depth and richness in exposing co-creation, complication, and complicity with the use of technological tools by artists. The manipulation of human visual culture by an AI is done at the bidding of cultural producers such as Anadol. It is these human actors who must strive to make meaning and relevance if they wish to open the door to show us the infinite library.
Dev Harlan, 2023
Columbia University, Uni. Writing: Data and Society
ENGL GS1010 518, Prof. Kirkwood Adams