The answer to these questions is not straightforward.
If the sole determinant of something being a piece of art was tied to its aesthetic qualities, then I think an argument could be made that for the public audience these generated works often appear to be art. Yet, as artist, researcher and educator Dejan Grba notes, art is more than just a pretty face:
Complex devices such as computers and software only represent the cumulative human creativity invested in their design, but the artists’ self-awareness, reasoning, abstraction, conceptualization, generalization, and analogy-making in dealing with these tools inform the cogency of their works. Their mental abilities, senses, emotions, passions, obsessions, and incentives determine how they interact with the world and make their art. These qualities and aspects should be in the forefront of AI artmaking. Conversely, a responsible approach to AI art requires a clear understanding that — while different forms of creative intelligence are possible and explorable — computers, robots, or algorithms are not artists because they do not embody human social embeddedness, cognitive capabilities, skills, quirks, and, most importantly, human motivations for making art.[i]
In addition to those qualities Grba submits, many would add originality, or novelty of a work. Whether every artwork introduces something novel might be open to question, but the possibility, it is argued, exists because of its human creator. Novelty sourced from an AI-based tool on the other hand is open to question. While the result itself is likely unique, its mechanism – its engine of imagination — draws from learnings based on images in its database. As these are images of existing works, the process is inherently backward-looking. Now this might be no different than the human artist who draws from the experience of others, but the distinction is that spark that Grba speaks of that enables the human to transcend into the novel and uncharted. Yet, if machine learning capabilities are able to extract the creative essence from the works in the databases they draw upon, then may be that spark will be simulated by the machine.
Another question is whether to qualify as art it should be difficult to make. Can we call something a work of art if it was easy to make? Should we expect the artist to have developed and honed their techniques over decades of study and practice? Is art in the craft of making or in the pleasure of witnessing? If we believe or at least accept the democratization process, then we would fall on the latter: the pleasure of witnessing.
One of the ethical questions that follows is whether it is acceptable to draw upon — appropriate — the works of other artists, and due to the current nature of the generation process, to do so without attribution nor an understanding of any copyright infringements. [ii]
Yet the appropriation of other people’s work has a long history in the art world.[iii] [iv] For example, Marcel Duchamp’s “Readymades” used off-the-shelf products to form his art works, one of the most famous being the “Fountain”. Andy Warhol’s Campbell’s Soup Cans appropriated the image of soup tins. Historically, many painters have re-painted variations of the works of past or contemporary artists.[v] Yet, there are limits: forgeries, which appear to be art, are clearly unethical because they intend to represent something they are not, an original, and that fact is hidden.
If Duchamp’s Fountain, Warhol’s Campbell Soup Cans and the many others, have generally been accepted as art[vi], and thus are unburdened by these ethical questions, then why not something generated by AI? Does the assembly process make a difference? In traditional art appropriation, the artist selects the objects and assembles them as necessary. The analogue for AI-generated art is the algorithm selects the items from a database and then assembles them.[vii] The differences are who makes the selection — human or machine — and maybe more importantly, the scale that a machine can achieve makes its appropriation loom so much larger.
If one of the steps towards accepting appropriation in a work is because it is considered art, then we are back to asking what is art? The definition of what constitutes art is illusive. Warhol defined it as “whatever you can get away with”.[viii]
While the aesthetic qualities of a work are often a determinant, it seems this is neither the sole nor exclusive one. Often, art is expected to convey some meaning or an idea. In the two examples cited above, the artists were confronting then-current norms. Duchamp was questioning what is art. Warhol was commenting on consumer society. Both these works raised serious question that went beyond the indifferent statements made by the objects appropriated. By establishing a new context, they created original works because they leave the viewer with a deeper insight into something beyond that of the specific elements appropriated.
It may come down to intent and meaning; there is likely a degree to which the original likeness of something appropriated can be carried over before it becomes unethical, but I suspect this constraint might have some relationship with the originality or importance of the message of the new work.
[i] Grba, Dejan. “Deep Else: A Critical Framework for AI Art.” Digital 2, no. 1 (2022): 1–32. doi:10.3390/DIGITAL2010001.
[ii] Gill, Joanna. 2022. AI-generated art is booming – but who really owns it? World Economic Forum. September 12 2022. https://www.weforum.org/agenda/2022/09/artificial-intelligence-ai-generated-art-ethics-copyright/
[iii] Artincontext. 2022. Appropriation in Art – An Overview of Artistic Appropriation in the Art World. Art In Context. May 28 2022. https://artincontext.org/appropriation-in-art/
[iv] Rowe, Hayley A. 2011. Appropriation in Contemporary Art. Inquiries Journal/Student Pulse 3 (06), http://www.inquiriesjournal.com/a?id=1661
[v] Ibid, Rowe
[vi] Ibid, Artincontext
[vii] Hsu, Wun-Ting, and Wen-Shu Lai. “Readymade and Assemblage in Database Art.” Leonardo 46, no. 1 (2013): 80–81. http://www.jstor.org/stable/23468123.
[viii] Bogost, Ian. 2019. The AI-Art Gold Rush is Here. The Atlantic March 6, 2019. https://www.theatlantic.com/technology/archive/2019/03/ai-created-art-invades-chelsea-gallery-scene/584134/
Leave a Reply