What some viewers thought was an AI-generated meme in The Devil Wears Prada 2 was actually drawn by a human artist.
Subscribe to read this story without ads
Get unlimited access to ad-free articles and exclusive content.
The long-awaited sequel to the 2006 classic dominated the box office this weekend as the original cast returned 20 years later to tell the story of the fictional Runway magazine.
One clip from the new film shows a number of internet memes disparaging Editor-in-Chief Miranda Priestley, played by Meryl Streep, including one of Priestley as a fast-food employee, accompanied by the text, “Can I lie about it?”
The meme only appeared on screen for a split second, but it appeared to many viewers as an intentional depiction of AI failing, part of the film’s satire on the modern media landscape.
However, after the artist behind the meme, Alexis Franklin, shared online that the image was hand-drawn, many viewers expressed surprise and excitement that the meme was not at all AI-generated, despite its appearance.
“I don’t mean to disrespect Queen Meryl in any way, but this is something I’ve been painting in my free time, so it was just fun when I was asked to do it,” Franklin said in an Instagram post. She said “The Devil Wears Prada” director David Frankel asked her to do the film.
Franklin’s post, which included a time-lapse video of her artistic process, drew hundreds of comments praising her work and the film’s decision to use a human artist for the project.
One commenter wrote, “It was so refreshing to see that it wasn’t AI.”
“AI will replace artists 🙅🏻♀️Artists will replace AI 🙂↕️,” another wrote.
Franklin told NBC News that it took several days of intermittent work to complete the digital painting and that he received a “substantial reward.”
She wrote in an email that she was aiming for a “cheap, plastic look reminiscent of the Photoshopped aesthetic of 2010s memes.”
However, its plastic look has some commenters asking if she was intending to mimic the look of the AI, especially with certain style choices such as blurring the menu text. (Illegible text is a telltale sign of an AI-generated image, as the model struggles to reproduce detail.)
“While technically I was trying to make it look artificial, I wasn’t thinking about emulating AI when I drew it,” Franklin wrote. “In that respect, I feel like the power of suggestion has taken hold. It’s funny how people regularly point out tiny human mistakes in their work and claim they did it on purpose to nail all the ‘AI slant.'”
Still, Franklin, who has been a professional illustrator for nearly a decade, said some people still have a hard time believing the work is her own. She has been accused of faking Priestley’s paintings, despite sharing time-lapses and a public portfolio of her work from long before AI images became commonplace online.
This is part of a growing phenomenon that is worrying media professionals. As generative AI technologies become increasingly sophisticated, people will not only be more likely to believe that AI-generated images are real, but they will also be more likely to believe that the actual images are AI.
“This group’s hypervigilance is rampant because people don’t want to be fooled, and you see signs on walls that don’t really exist, or signs that have very simple and rational explanations,” Franklin wrote. “And it’s hard to know what the solution is.”
She added that she understands the skepticism, especially from people who want to support human artists. But it can also hurt artists, she says.
“AI is so pervasive now that I feel like people have forgotten how it became so good. AI has been studied. us” Franklin wrote. our! ”
