In response to Adapt or Die, Ben Zhao, a professor of computer science at the University of Chicago, who was interviewed for the story and whose SAND Lab (Security, Algorithms, Networking and Data) developed the anti-mimicry tool Glaze, wrote the following:
I don’t usually give feedback after these articles come out, but I feel compelled to respond to this article.
I am frankly, beyond disappointed.
You represent a school of design … you educate artists and aspiring artists. But this article fails them.
Generative AI in its current form, literally cannibalizes the work and livelihood of multiple generations of active artists in the field today. Training on their data, without consent, without compensation, so that others might bypass them entirely and instead reduce them down to a prompt at a Midjourney console. Is this what you want for your future graduates? There are artists and artist advocates fighting for the rights of your future graduates, in the courts, on Capitol Hill, on social media, on picket lines, and in research communities. If I were a graduate alumni or parent of a student at ArtCenter, I would be deeply disappointed at this article.
The people who chant "adapt or die" have a fundamental misunderstanding of how AI works. AI requires human artists to make it better. This is not a one time transformation tool that’s built, and everyone just uses it. It is a pipeline that even today, actively siphons off of real human artists to fund the hype and profit of a small group of companies. Even as you take a positive spin on genAI image models, they are literally driving hundreds/thousands of human artists out of their jobs and homes. And as human artists exit, AI models will lose their ability to evolve and stagnate into the glossy, cheap look that is becoming their trademark. Every time a student uses a tool from Midjourney or Stable Diffusion or their many derivatives, they are unknowningly exploiting the same artists who likely inspired them to enter the profession. I can understand many art students not understanding that, but as ArtCenter, it should have been your responsibility to educate them to the complex issues involved, instead of glossing over the ethical issues with stories of how great it was to use these tools.
I am not an artist by trade, but even I can observe the concrete harms these tools and companies are inflicting on some of the most creative and skilled people I know. They, and your students, deserve better.
Ben Y. Zhao
Neubauer Professor of Computer Science
Director of Graduate Studies, UChicago CS
July 15, 2023
In response to Zhao's email, as well as to the larger online discourse surrounding the article, the author of the piece offers the following:
As the editor of ArtCenter’s Dot magazine and author of this story, I appreciate the discussion prompted by the piece and would like to clarify a few important points.
First, this article should not be construed as a statement or official position on generative AI by the College.
Secondly, my goal was not to take a position on AI, but instead to illustrate the dialog happening at ArtCenter and in the larger world over its use.
As I tried to convey in the piece, generative AI brings with it many legal and ethical considerations. Some artists and designers are embracing it. Some are fighting it in the courts. Others are seeking technological means to combat it in its current form. Whatever stance they’re taking, they are all adapting to meet this moment.
As a writer, I am not immune to the potential impacts of generative AI. I find myself asking questions like, “Will my job be replaced by AI in the future?” “If my writing is no longer needed, what is my value to an organization?” and “What can I bring to this story that AI can’t deliver?” These uncomfortable and hard-to-answer questions are a reason I was drawn to writing this piece.
This past Friday, the College held a symposium, “AI in Teaching and Learning at ArtCenter,” in which faculty presented their thoughts on the use of generative AI in the classroom. The opinions varied widely. The discussion over generative AI is just beginning, and I look forward to covering the topic as it continues to evolve.
Mike R. Winder
July 18, 2023