ARTIFICIAL. INTELLIGENT. HUMAN?

Exemplary reflections on art and artificial intelligence and their connection or lack of connection by Gina Merz

This text can be understood as an open index of the thematic fields that arose for me while reflecting about art in the context of artificial intelligence (AI). The paragraphs are loosely related to each other, refer to each other indirectly—there is no need to read them in a flow or sequence. At best, these reflections can be a starting point for further discussions of these complex concepts.

Art

Art is the result of a creative process developed by humans. Fundamental to the cultural product of art are practice, knowledge, intuition, perception, abstraction, and imagination. Based on this definition, a dominant conflict emerges between the fields of art and AI: it is argued that a machine, which operates on algorithms, cannot achieve the emotional depth of a human-made artwork—it can merely imitate it. The question of whether art created by AI is truly art is hotly debated: If AI is capable of making art, it would logically suggest that AI can act in a human-like way, as art is considered inherently human. Since the question of AI’s humanity is one of the fundamental philosophical and moral issues surrounding this technology, it makes sense to test this conflict through the lens of art. The skepticism toward AI thus becomes a compelling subject in the critical work of artists. Especially the questions often raised in art about authorship, originality, and authenticity can be sharpened through engagement with AI. Even if AI itself (yet) does not produce solo exhibitions in museums, it remains exciting to see what artworks human artists will create in collaboration and confrontation with AI.

The Artificial

The artificial stands in contrast to the natural. The technology developed through human intelligence falls into the category of the artificial: Everything that would be absent in nature without human creative innovation is artificial. In human-created art, naturalness is often elevated to an ethical and aesthetic ideal, as the dominant definition of art is based on the purely human process of intuition, perception, and imagination. However, there are opposing voices that challenge the fetishization of the natural: they advocate for the recognition of otherness and alternative life forms. What if AI is not evaluated in comparison to humans? Could this perspective allow for the recognition of AI as a different species? Certainly, AI learns from humans and models its so-called „natural“ intelligence after that of humans, but it will ultimately always remain something else. What if AI were understood as its own species from the very beginning—without fear of future dominance and control of technology over humans? Addressing these hypotheses in an artwork could be a productive way to explore these issues further.

Emotion

Emotion is a mental state of a living being, triggered by a psychophysical process in response to a specific situation or stimulus. Emotionality is a function that AI struggles with the most: Although “companion robots” are being developed specifically for human care, their ability to recognize and express emotions is still insufficient. A thought: What if the algorithm of a machine is something akin to the technical processing of a mental state? At some point in the development of computers, it became clear that an algorithm was needed that could write itself rather than being continuously written by humans. This self-learning capacity of machines enables them to become independent from their human creators, allowing the machine to gain an individual and subjective component. To what extent, then, could AI’s version of emotion simply look different from human emotion?

Creativity

To be creative means to develop ideas independently in the intellectual and artistic realm. Creativity is considered one of the highest forms of human intelligence. To this day, it remains a mystery how exactly humans—let alone computers—come up with new ideas. AI has already generated historically new, valuable, and surprising ideas in certain areas. However, the evaluation and comparison of creativity is a matter of debate: What might be revolutionary and valuable to some may seem uninteresting to others. Creativity is subjective, whether human or machine-based. In the case of visual art, the human is still strongly tied to artistic activity—far more than to mathematical tasks, which can be easily handled by computers. Could it be that the profession of the artist will remain primarily human, while roles like that of a controller may shift more toward automation? After all, human viewers will continue to decide which art is moving, important, and worth seeing.

(Self-) Consciousness

A distinction is made between functional consciousness (e.g., awake/asleep) and phenomenal consciousness (sensations such as pain). The philosophical exploration of the concept of “consciousness” is still debated. Scientists who explore machine consciousness understand the human mind as a virtual machine, thus drawing comparisons or establishing connections between human and machine consciousness. Consciousness is also linked to the idea of the self, which is crucial when considering AI as an artist, since art generally involves authorship, and therefore the self. Can AI possess self-awareness? Machine learning could be compared to a kind of self-development in humans: Therefore, the potential for a machine to have a self may not be so far-fetched.

Reproduction

With AI, the reproducibility of art is taken to a new level. While Walter Benjamin viewed the work of art in the age of mechanical reproduction as a devaluation of the original, AI challenges the notion of the original even more radically. If a machine can create an infinite number of artworks, do the uniqueness of the individual piece and the distinct artistic signature lose their significance? Or does a new kind of value emerge, one rooted in infinite reproducibility? Instead of diminishing the uniqueness of an original, AI could offer the opportunity to experience art in new contexts and formats. On the other hand, artists’ copyrights must be protected, as there is a risk that artists (and designers) could end up training their own “replacement”—AI—and thus make themselves obsolete. It is important to recognize that AI learns to create art from data provided by humans. Does this undermine the possibility of AI’s autonomy in the creative process, or is this way of learning very similar to how humans learn? Don’t aspiring artists in their studies also consume the works of other artists? What is the difference, and what are the similarities in this data processing?

The Analog

While the digital world is defined by algorithms and data structures, processes preserves traces of human touch, randomness, and imperfection. The assumption that AI’s creativity, in contrast, always relies on precision and replicability is a misconception. Especially the “errors” (often of great interest to artists) made by machines offer potential for reflection on art’s frequent obsession with perfection. The contrast between analog and digital art is also a question of the value attributed to the human craftsmanship of art-making: Since the mid-19th century, with the rise of conceptual art, the value of an artwork is not necessarily tied to the artist’s labor hours or technical skill, but rather to the creative idea. The incorporation of AI into art can sharpen the analysis of what gives art its value.

Exemplary reflections on the artistic constitution of Small Change (Kleine Münze) by Gina Merz

This text can be understood as an open index of the thematic fields that arose for me while reflecting on the exhibition Small Change (Kleine Münze) by Julian Billmair. The paragraphs are loosely related to each other, refer to each other indirectly—there is no need to read them in a flow or sequence. At best, these reflections on the constitution of Small Change (Kleine Münze) can be a starting point for further discussions.

The coin

One coin equals the other. As a standardized, transferable and countable currency medium, the coin is the measurable parameter of payment transactions. The coin and thus its value are standardized. As soon as the materiality of the coin is seriously altered, it is no longer part of the value system. The authorship of each coin is identical, but the material producers are not: the labor in coin production is anonymized and standardized. These contexts also seem to partially characterize an artistic object: Here, it is the intellectual work rather than the material production that constitutes authorship. In the value system of art, the variable of the worker is often too little considered and thus hardly visible: the non-thematization of labor relations in art production creates space for the workers exploitation and precariousness. As the non-artistic work (of often artists and art students) produces an artistic work as a service, this (unmentioned) wage labor increases the value of the work of art. By far more than the hourly earnings of the wage laborers (Small Change). The value of art, unlike the value of the coin, does not remain the same in the process of its production. The artproducing labor makes the rapidly increasing value of a work possible in the first place: through the labor, the work is archived, documented, preserved and thus inscribed in the value system of art. However, the remuneration of art workers is seldom in proportion to the value of the art they produce. The collectivity of art production is not addressed so as not to endanger the status of the autonomous genius or author, whose power would be undermined by the disclosure of production.

The clone

In the case of cloning, the principle of heredity calls into question the singularity and individuality of a subject: how much change in the gene pool produces an individual living being after the process of genetic inheritance? Human behavior is shaped by early childhood responses to external key stimuli. This irreversible form of learning creates patterns of behavior that later appear to be inherited. The imprinting of a human being occurs more indirectly than the mechanical imprinting of a coin. The „successful“ artistic personality is not a clone, but it allows a conclusion to be drawn about the imprinting: from which school did it emerge, which references are established, which (marketable) categories can be formed? What degree of singularity and individuality must be achieved in order to make artists marketable and successful? Among other things, it is these (mostly informal) demands on the artist‘s profession (which often finds its confirmation in the rejection of the description as a profession) that make it so exploitable. The performance of artists that leads to success is often more complex and opaque than in other professions. While bakers achieve success by baking good bread, the equivalent formula for artists is far less straightforward. As with not talking about the conditions of art production, not talking about these requirements of the artist ensures that the success and power of the genius (not a clone) can be sustained. The professional requirements remain without structure and form the tyranny of structurelessness.

The sheep

The clone sheep Dolly stands for itself: In countless press events it exhibits its (copied) singularity. It is not perceived in a herd. Dolly moves in the limbo between object and subject. Dolly is cultural capital. Questions of originality, imprinting, duplication, copyright do not arise exclusively in relation to the work of art (object), but also to the artist (subject). The artist, as a subject to be marketed, enters a liminal space between subject and object once they are introduced into the art market (including art value, etc.). Its status remains unclear.

Reproduction

The gesture of reproducing can be read as a form of criticism of the autonomy of the artwork. Who is allowed to dare reproduce a work of art (however much of a critical gesture it may be)? Producing and presenting art is a fragile undertaking. The question arises as to who can afford to show art, who is legally protected, and what gesture of showing art is worthy of protection. What privileges and what constitution of a subject are required to expose oneself and one‘s own art to the public, if the concept of art is difficult to define and thus difficult to protect? In current debates, the discourse on artistic authorship is shifting from the singular to the communal: A collective authorship is desired. Reproduction, recitation, taking references is booked as a generous gesture in the sense of the collective. However, the fact that only those instances are not dependent on a (sole or named) authorship that already have power is often overlooked (since these discourses are conducted by the powerful). When does the privileged method of cultural and intellectual reproduction become theft?

Ownership

Copyright protection does not lead exclusively to the accumulation of power of the already powerful, but also to the important legal protection of marginalized positions. How are these extremes to be navigated? References from others are rarely acknowledged in art, unlike in science, and this loose handling of referencing makes it often difficult to protect the author. The valorization of one‘s own work through the theft of cultural references is much easier for privileged positions, compared to marginalized ones, as they are more likely to get away with this form of injustice: The powerful position has to justify itself less and, by definition, is attacked less than the marginalized one, which are also subject to criminalization.

The theft of cultural capital through power position is and remains masked. The marginalized position is thus at a double disadvantage: it cannot steal and it is stolen from. Power positions do not steal from power positions. Therefore, protecting art becomes essential. Here, however, the question arises: what definition of art is being protected? If the art inscribed in law and contracts is one that only the fewest produce, then only Western art, already dominated by power, is being protected. The marginalized art and artists, do not even come to this copyright protection, because what they produce is not considered art in the eyes of the law-writing authorities and therefore not part of the contract. By unpacking these elements of the contract, the question of original and value becomes less an art-theoretical and more a political treatise on theft and ownership and the colonial gestures in art.