After reading the essay above, discuss your personal concerns about AI in the visual arts and in visual culture in general.
My biggest concern about AI in the visual arts is copyright violation and unauthorized use of intellectual property in generative AI models. If a generative AI model generates an image visually very similar to an artist’s original artwork (to the point where tracing is at least in part, very obvious) that they did not consent to feeding to the AI model, that could be considered theft. I have to wonder—why didn’t generative AI models stick to just extracting data from art under Creative Commons or public domain? I imagine that could have prevented a lot of their current legal battles, and respecting copyrights and intellectual property rights is a policy students are expected to abide by in our own DTC program here at WSU Vancouver.
At least for me, part of the appeal of hand-drawn art is that it is hand-drawn, often requiring a complex set of skills the artist has developed over time. Generative AI lowers the barrier of entry for people so they are able to generate an end result (artwork) while bypassing the traditional, often time-consuming process of original creation that requires a level of skill that also takes time to develop. Is this fair or ethical? That is a question that is very difficult to definitively answer. I think AI has its uses and can be very helpful, but that should not take precedence over the rights of traditional artists.
One of the arguments detailed in the essay regarding photography is that it “automated image-making in ways that were free from human interpretation.” I more or less agree with this, and as such, photography is regarded as a medium of art separate from hand-drawn art. Generative AI also automates image-making, but is not free from human interpretation as the datasets it draws upon are largely human interpretation.
What can artists/creators do with AI tools to address or respond to these personal and collective concerns about machine creativity?
Just as photography is considered a separate category from hand-drawn art, it would be ideal if the same distinction could be made between AI art and hand-drawn art. That said, I have seen many people try to pass AI art as hand-drawn art, especially when selling it. There seems to be an implicit understanding that part of the economic value of art is the time, effort, and skill used to create it—to say to a potential customer/consumer of your art that you took time, effort, and skill as required in hand-drawn art and the art is priced accordingly, when in fact you did not take that time, effort, and skill is indeed deceptive.
I have also read about Glaze and Nightshade, two tools artists can use to combat the extraction of their art into a generative model’s dataset. Glaze protects against style mimicry, while Nightshade changes data within digital art/rendition that deceives the AI model into seeing something different from what the human eye sees.