On Who Gets Credit for AI-Generated Art?

A short response to Science Direct article Who Gets Credit for AI-Generated Art?, written after making i feel so alone in this house (2021) Questioning whether AI-generated artwork is a product of the artist behind the computer or of the computer itself inherently humanizes the AI, creating a flawed comparison from the very beginning. The article explains that “anthropomorphizing” AI systems can create a false sense of trust in them, as well as prevent people from holding the technology and its creators accountable for wrongdoings. From a more creative perspective, however, I think humanizing artificial intelligence creates a different kind of problem. Setting up the scenario of artist vs. computer puts them both on the same level, suggesting that both are capable of expressing imagination. Factually, this isn’t the case: the artificial intelligence currently used to create imagery is the product of sophisticated code that can recognize and reproduce visual patterns based on their vast image databases. Although the images they create look like artistic renditions of our world, they are, in reality, just complex visualizations of data and code. Professor Aaron Hertzmann argues that AI is simply another tool for creating art and, like cameras not receiving credit for films, it should not receive credit for the creation of an image. Artist and researcher Jon McCormack similarly concluded that “AI systems are not broadly accepted authors” but instead “the creator of the software and the person who trained and modified the parameters to produce the work” can be. Although irrelevant to this debate, one could argue that the human imagination is no different, only with complex biological pathways in our brain instead of code. Rather than pulling from image databases to imagine, for example, a blue tiger, we reach into our memory to combine tigers we’ve seen before with the color blue. Nonetheless, the conclusion reached within the article leans more towards regarding AI as an artistic tool, rather than a conscious creator. I’ve come to a similar conclusion myself when it comes to determining the role of artificial intelligence within my artwork. Although the visual material in i feel so alone in this house is entirely AI-generated, I still had to go through the process of selecting which outcomes I used, weave them into a narrative, and edit them together into a final artwork. When I think about generating images with AI, I like to compare it to the process behind some forms of abstract painting. Painters like Jackson Pollock or Damien Hirst aren’t in control of how the paint lands, drips, and flows on the canvas, yet they are still the sole artists of their paintings. Suppose one questions their art in the same way people often question AI-generated art. In that case, one could argue that the force of gravity, the paint's viscosity, and the canvas's angle should be considered authors of the final painting. In the end, their process of selecting the colors and methods of applying the paint as well as selecting the best random arrangements of paint on canvas ultimately grants them their role as the artist.