Move Along, There’s Nothing to See Here: A Thought Experiment, Illustrated (2023) consists of nine red monochromes made of handmade paper by the Dallas artist Tino Ward plus nine panels of text generated by prompts by Michael Corris addressed to the Artificial Intelligence program ChatGPT, which is designed to interact with users in a conversational way.
The initial question posed to the AI program requests an explanation of Arthur Danto’s thought experiment involving an exhibition of nine identical red monochromes. In Danto’s original account, found in his book The Transfiguration of the Commonplace: A Philosophy of Art (1981), some of the nine red monochromes are titled, as though actual works of art, while others are merely described, indicating their identity as ordinary material. [See below for titles.]
The philosophical puzzle at the heart of the Danto’s thought experiment asks us to consider “how is it that some of these identical red monochromes—indiscernible from ordinary, red-colored materials—are reckoned to be works of art while others are merely ordinary things in the world?” For Move Along, There’s Nothing to See Here: A Thought Experiment, Illustrated, ChatGPT is queried using the standard interview technique known as “funneling”. That is, the interrogator poses increasingly specific questions in order to probe the depth and coherence of the subject’s knowledge. Here, we are attempting to determine the accuracy and sophistication of ChatGPT’s representation of Danto’s philosophy of art. By assessing its performance as a knowledgeable interlocutor in a philosophical dialogue, we are hoping to reveal ChatGPT’s potential to go beyond paraphrasing existing philosophical arguments associated with art and indiscernibility.
Thus far, one stunning instance of what is picturesquely called “hallucination” has been noted. This is the instance where ChatGPT spontaneously generated the titles claimed to have been assigned by Danto to the nine indiscernible red monochromes. [See below for titles.] All ChatGPT’s titles differed significantly, and hilariously, from Danto’s. ChatGPT’s titles range from the evocative—“A flag for a parade of communist labor parties”—to the absurdist—”A piece of fabric intended to cover a hole in a larger piece of fabric.” On the other hand, Danto’s titles—such as “Israelites Crossing the Red Sea”, “The Red Lead Ground Prepared for Giorgione’s Sacra Conversazione (1505)”, and “Kierkegaard’s Mood”—reference his philosophical and art historical acumen. What has yet to be simulated by any AI program is the philosophical discourse stemming from Danto’s epiphany upon beholding Andy Warhol’s “Brillo Box” at Stable Gallery in 1964.
The second dialogue—Move Along, There’s Nothing to See Here II (2024)—is a revised version of the first conversation. While dramatic liberties are taken, the essential structure and content of this reconstructed transcript is similar to that of the first. The characters of the “Artist” and the “Philosopher” make their appearance as interlocutors emotionally and intellectually wed to their respective positions. The language of this dialogue is polemical and subjective as befits a conversation where the stakes are high.
The third project in the series consists of two parts—Philosopher: An Illustrated Thought Experiment #1 (2024) and Philosopher: An Illustrated Thought Experiment #2 (2024)—that explore human-AI interaction in the context of art illustrating philosophy. The first part consists of nine red monochrome canvases whose titles are taken from a fortuitous “hallucination” generated by ChatGPT. The grid-like installation of the canvases is interrupted by the addition of the second part: nine excerpts from the second ChatGPT dialogue rendered as digital prints in the form of typographic samples. While these two projects have been conceived as two distinct works of art, they are installed so that the component parts of one disrupt the order of the other, producing a kind of visual dissonance. One way to consider this third project is as the literalization of Danto’s exposition of his thought experiment: nine indiscernible red monochromes embedded in (fragments of) theoretical discourse framing them as objects of philosophical interest.
* * *
Engaging with AI as a conversational partner is the stuff of cinematic musings like “Her.” How can it be taken seriously as a significant means of making art, let alone a worthwhile philosophical pursuit? Yet engaging in a language-based exchange with a computer program is not new. In 1950 Alan Turing conceived of it as a means to determine if a computer program could be written such that a human being would be convinced that they were communicating with another person. The Turing Test—or imitation game as it is known—seems to be on everyone’s mind these days as generative AI programs become increasingly complex and adept at simulating human language skills, especially mimicking empathetic verbal interaction. From an artist’s point of view, the act of engaging with ChatGPT is an opportunity to test the limits of technology as an adjunct to the making of art.
Some AI advocates argue that the continued development of large language models points inevitably to the emergence of sentient computers. The so-called “singularity”—that is, the point at which autonomous machine consciousness will overtake and supplant us as the dominant species—is highly controversial and in the opinion of most experts implausible given the current state of computational technology. Nevertheless, the notion of thinking machines continues to capture the imagination of respected computer engineers, some philosophers, and the popular media, conjuring apocalyptic Darwinian scenarios of robots outstripping human intelligence and ultimately threatening our existence.
A more immediate consequence of the rapid development and unrestricted application of generative AI is the threat it poses to the livelihood and esteem of artists and performers. While some contemporary artists enthusiastically—and sometimes critically—embrace AI as a creative medium, it is apparent to others that their integrity, autonomy, and intellectual property rights are at risk. Intellectual property rights are particularly vulnerable owing to the unrestricted and covert mining of internet data that is required to train large language learning models. Clearly some visual artists are willing to strike a Faustian bargain with AI because they believe that its negative aspects will be more than compensated for by its potential as a technology offering visual culture a spectacular aesthetic future. Perhaps these artists envision AI to be the dominant artistic medium of the twenty-first century; one whose impact will rival photography’s rise as the dominant visual form of the twentieth century.
Others in the creative professions—particularly screen actors, musicians, graphic designers, and writers—may not fare well compared to the visual artist embedded in the fine art market/museum network. The current organization of production in the cultural industries, coupled with a lack of regulations governing the impact of AI on intellectual property rights suggests that workers in creative professions are already facing a profound existential threat to their expertise and livelihood. The labor force of the cultural industries—artists and writers working in cinema, online streaming, and advertising—is constantly under pressure from automation in the form of generative AI posing a threat to their livelihoods, cultural status, and expertise. Organized labor in some sectors of the culture industry has already recognized this and is actively engaging in the task of preventing their skills from being grossly devalued, stolen, or rendered obsolete.
* * *
Titles for Nine Red Monochromes
Arthur Danto’s Titles (The Transfiguration of the Commonplace. A Philosophy of Art [1981], pp. 1-3)
1. Israelites Crossing the Red Sea
2. Kierkegaard’s Mood
3. Red Square: A Moscow Landscape
4. Red Square: A Minimalist Painting
5. Nirvana, or ‘Red Dust’
6. Red Table Cloth
7. Preparatory Ground for Giorgione’s “Conversazione Sacra”
8. Sample: Red Lead
9. Untitled
ChatGPT’s Titles
1. A Philosophical Meditation on the Nature of Art
2. A Cheap Poster
3. A Piece of Fabric Intended to Cover a Hole in a Larger Piece of Fabric
4. A Panel in a Red Room of a Temple Devoted to the Worship of the Sun”
5. A Flag for a Parade of Communist Labor Parties
6. A Study for a Painting by a Great Artist
7. A Poetic Masterpiece
8. A Small Work in a Large Series of Monochromes Painted by a Madman
9. An Early Work of a Famous Painter
Michael Corris 