It is true that you can’t judge a book by its cover. But these days, you can’t judge whether a cover is AI-generated or not, unless the credit says as much. I bought this book, The MANIAC by Benjamín Labatut, because of its title and cover, which piqued my interest in morbidity. It was shelved in the fiction section, but as I began reading, I was convinced it was, in fact, nonfiction. Likewise, the cover image looked to me like a real nuclear explosion (the kind that haunted me during the Cuban Missile Crisis).
And then I noticed a brief explanation on the back cover: “The image on this cover was created by Bennett Miller, using OpenAI’s DALL-E 2 software,” it began. “He arrived at the final product by making extensive edits on variations of images using the following prompt: A vintage photograph of huge plumes of smoke coming from a huge UFO that crashed in the desert.”
It was the first time I’ve seen a description on a design that pointedly explained its origin in AI. My colleagues have been arguing for over a year about the best way to discuss generative AI in an editorial (or any design) composition. Do we follow the rule of thumb for collage/montage/Photoshop illustrations? (E.g., Photo illustration by so and so, as the publisher has done here?) Or do we credit the maker as Prompted by so and so/generated by Midjourney? What would be most logical, ethical, honest and appropriate? I’m still not certain.
How do we credit an image if it’s entirely composed of existing images remade (even with numerous edits) into a unique entity? What we call it is one issue, whether we adopt the practice is another. I’m of two minds … maybe more. With The MANIAC, initially I felt hoodwinked by the artificial origin of the image. Despite the explanation that a human was indeed the driver, editor or prompter, I was both concerned for the future of illustration and admirous of the quality of the result. Indeed, it was the perfect evocation of the tagline, “A work of dark, eerie and singular beauty.”
I ultimately concluded that this book was the perfect vehicle for AI illustration. The MANIAC of the title (an acronym for Mathematical Analyzer, Numerical Integrator and Computer) was a fictional machine “designed” by the real-life quantum physicist John von Neumann, who was involved—and this is the point of the cover choice and its AI generation—in creating the first atomic bomb. The novel posits that von Neumann fathered AI for dubious reasons. Labatut writes:
“The first goal that von Neumann set for the MANIAC was to destroy life as we know it.”
“The second goal that von Neumann set for the MANIAC was to create a new type of life.”
Both first and second goals apparently posit evil rationales behind the invention and development of AI as it has evolved over time. The second goal is the inevitable result of the first, regeneration of life and, for our purposes, the transformation of art and design. So, if you are in the mood for a book that relates directly to our current times and the shifts in all we know at the moment, Labatut’s narrative and Miller’s AI generative illustration say it all with pessimism and dread.
The post The Daily Heller: A Book Cover Where AI Says it All appeared first on PRINT Magazine.