{shortcode-b4f2a69b3fda2e88aaf05beffd3cb737859ab104}

In Jorge Luis Borges’ single-paragraph tale “Del rigor en la ciencia,” the Argentine short-story writer envisions a mythical kingdom where “the art of cartography attained such perfection” that its ruler decided to construct a map identical in both size and shape to the realm itself. In this fictional land, reality and its reconstruction became one and the same.

Many have argued, to varying degrees of persuasiveness, that Borges’ fable is not at all as absurd as it might seem, especially with imitating or replicating reality becoming ever easier in today’s increasingly technological society. In the 1980s, French sociologist Jean Baudillard made this very observation in “Simulacres et Simulation,” using the Borges story to introduce a dense critique of the increasing prevalence of imitation and other forms of unoriginality in mass culture. To Baudrillard, it had simply become too easy to replace something original with a copy that was sufficiently verisimilar, thus blurring the distinctions of authenticity enough to render them effectively meaningless.

Baudrillard has surely then rolled much in his grave over the past few months, as access to both natural language interfaces, such as ChatGPT, and engines for producing art from generative machine learning algorithms, such as those powering DALL-E and Midjourney, have finally been granted to the public. While computational tools have been in the contemporary creator’s toolbox for decades, the immense amount of apparent originality that these newfangled models seem to be capable of has simply been too much for some artists, as what may be best described as a moral panic has swept through creative communities and stimulated charged debate about the threat of AI to the “essence” of what art is.

Permit me to use a historical perspective to assuage these worries. I hold that, at least with respect to the production of art, the emergence of generative computational tools that many have — in my humble opinion, too hurriedly — touted as “artificial intelligence” is no different from any of the other technological innovations that have been introduced to the fine arts at one point or another in the past.

Take, for instance, something that seems as mundane nowadays as photography. When first invented and introduced to the public, the photograph drew scathing criticism from both artists of the time, such as John Ruskin, and art critics, including Baudelaire and Wölfflin.

However, over time, photography gradually became an integral part of the visual arts — first as a potent source of inspiration for painters such as Gustave Courbet, who exhibited a deep interest in interpreting photographs through painting. Today, his “Le Puits Noir” hangs prominently in a second-floor gallery of the Harvard Art Museums among the works of Cézanne, Renoir, and other contemporaries who were similarly influenced in their art by the advent of photography.

More recently, photography has even gained acceptance as a medium of artistic expression in its own right. As the multitude of photojournalistic works in the Harvard Art Museums collections — from the powerfully titled “Refugee in Hiding, Miami” to Dorothea Lange’s famous images of the Depression — demonstrate, the proliferation of a new technology need not be viewed as an existential threat to the fine arts. Rather, it should be welcomed by artists as a new medium of personal expression that can help their natural ingenuity and originality shine further.

Keep in mind that it is us humans who are the ultimate arbitrators of what constitutes “art.” The generative algorithms that have captured the public imagination rely fundamentally on building off of artistic motifs created by human agents. Their products represent the highest form of Baudillard’s simulacra, for even though they seem original, they are far from it.

Not only that, consider how important the process of making art is to its designation of originality. A human artist undergoes a process of extensive reflection, whether about their external environment or internal world, that directly affects the outcome of their work. But just like how making a gicleé of an original van Gogh requires no more creativity than it takes to use a printer, the method by which an algorithm produces a piece of art is categorically mechanistic. There is no reflection and no cognition intrinsic to its operation, and at best, it is the human who is providing the text or media input who contributes any semblance of creative vitality to the process. The algorithm couldn’t care less.

So, do not fear. As long as humans continue to be creative, the artistic cultures we cherish will continue to flourish. Yes, the economics of supply and demand suggest that the ability of “art AI” to provide vacuous eye candy in seconds will place downwards pressure on the market for simple illustrations, but if the digital camera hasn’t been able to extinguish the market for individual and family portraiture, what’s so different now?

Alexander Junxiang Chen ’24 is a Neuroscience and Chemistry concentrator in Quincy House. His column “Artifactual” appears on Thursdays.