Leo Goldsmith
Generative AI in Teaching, for Remixing the Moving Image
In this class we will explore the relationship between form and content — how is meaning constructed and communicated through visual language? Through observing, collecting, analyzing, writing, and form-making, students will apply design processes involving visual research, concept generation, and craft skills. Driven by research interest, students will use digital and analog means to build visual archives.
Faculty Reflection
This course looks at the history of remix across a range of media (mainly moving image, but also music, collage, digital media, etc) in light of technological shifts and changes in law and artistic practice. Appropriation/remix challenge our sense of artistic originality and the role of technological reproducibility in art. This course asks students to take a critical view of claims about these ideas, and even to ponder what kinds of appropriation are ethical or not (valences that can shift wildly depending on the example, from sampling in rap to textual plagiarism, etc.).
Generative AI raises still more questions, as it seems to borrow some of the practices of modernist appropriation but usually lacks the sense of conceptual artistic intention or a criticality about the economics and politics of art. Yet it remains difficult to parse how artistic forms of remix differ from those performed by AI, and how we might understand and frame those ethical or aesthetic distinctions.
Throughout the course, students used genAI tools to investigate how they functioned, and to then contrast their results with both historical examples of appropriation/remix and works of their own making. Students were frequently resistant to using AI for a number of reasons. Even though many confessed to using these tools elsewhere, they also saw them as unethical, both because of their detrimental ecological impact and because of their questionable ethics. I also saw these as ambivalent reactions to being asked, in a classroom setting, to explore similar practices that they have been repeatedly taught are “wrong,” such as plagiarism and piracy.
There was also considerable frustration with the tools themselves. I decided to mostly have the students explore tools that are available online for free or available through their New School accounts (such as Firefly). What they discovered was that these tools required lots of human effort, and were often very frustrating to control. Frequently, they were unable to make sense of their results, or to get these platforms to produce even basic iterations of what they were working on. Also, remix art tends to foreground its processes of appropriation, sourcing, materiality; genAI often obscures its sources and textures for a cleaner final product.
This led me to reflect on the value of integrating genAI into a classroom setting. Even as I invited criticism and skepticism about AI and its role in art-making and society at large, I wondered if there were other ways of exploring and understanding these technologies—ways that perhaps would not require the students to use them directly. This is a larger pedagogical/methodological question for media studies, one that poses a question about engagement and complicity. Can we critically engage with genAI and not be party to its ethically questionable operations?
Curriculum and Student Projects
Questions to keep in mind:
What can you discern about the functionality of your genAI tool? What are its capabilities and limitations? Is its process of association/remix similar or different to your own? Does it betray certain latent (or blatant) prejudices or biases?
Project 1: Make an original work of your own using one of the following styles: collage, montage, compilation, appropriation, or supercut.
Remixing the Moving Image, Spring 2025
Project 2: Experiment with a similar work made using an A.I. app, which tries to approximate the same goals as your own artwork, and you may need to try several different prompts to make this work.
Project 3: Provide a response paper that addresses the motivations/inspirations behind your original work, and discuss which style you are using, and on the similarities or differences between your work and the AI-generated one.
Learning Outcomes:
- Build and utilize that critical vocabulary in aesthetic and political analysis of film/moving-image media, including film, video, digital culture, and generative A.I.
- Develop a critical understanding of the connections and distinctions between a number of related media practices, their politics and aesthetics, as well as their technological histories (film, video, digital culture, generative A.I.)
- Develop an understanding of the historical development of appropriation practices in a number of media and across genre
- Grasp a range of media styles, genres, production contexts, and significant directors/artists, and contrast these with gen A.I.’s forms of appropriation
- Articulate interpretations and analyses of media texts and contemporary issues around aesthetics and technology through written and creative exercises and in group discussion
Video Media by Issabella Moreno, “A.I. Disruption.”
“A.I. is the ultimate plagiarist. It’s the recycling of what came before it. It steals from what is established and contorts it into an artificial reproduction. And in a sense that’s also what I am doing in my project, I further contort the contorted. I don’t display the A.I. in its “raw” form because that doesn’t exist, at least to me it doesn’t.”
Student Reflections
“What I would soon find out, was the extent in which the A.I model would perpetuate the disturbing stereotypes and cliches contained within the data the model had been trained on. As my project contained delicate topics such as incarceration, there was the need to produce images of incarcerated prisoners, to then contrast alongside scholars and workmen. Without mentioning race once in my prompts, the results were overwhelmingly racist. Of the thirteen times I asked the software to create the image of a prisoner, ten of the results came back depicting black prisoners. Of the ten times I asked the software to create the image of a business man in a suit, eight of the results came back depicting white men.”
— Wilf M.
“Artistic appropriation differs from AI appropriation, in my opinion, primarily because of the intent behind the appropriation. When appropriating images, artists are often particular about what they appropriate.… Aesthetically, many AI-appropriated pieces of art are not pleasing to the eye, but beyond that, the lack of intentionality around the selection and curation that goes into a piece of appropriated art made by AI removes the emotional and intellectual value behind the piece, as it has been created from an algorithm, not an idea.”
— Chris W.
“Over my time at the university, I have watched my professors’ stance on AI change drastically. As a freshman, it either wasn’t on teachers’ radars or they vehemently discouraged the use of it. Now, my professors are not only encouraging it, they are requiring me to use it.”
— Lily-Rose V.