A Lo-Fi Rebellion Against A.I.
.jpg)
Two and a half years ago, Christine Tyler Hill, a designer and artist in Burlington, Vermont, began working as a crossing guard in her neighborhood. The city paid her twenty dollars an hour, but the real draw was the chance to get to know local families and “be more enmeshed with my very immediate, outside-my-door community,” she told me recently. She was tired of staring at a screen doing design work, and new clients were getting harder to come by, in part, she surmised, because of the rise of generative artificial intelligence. She began documenting her crossing-guard shifts on Instagram, posting mini comics about the frigid weather, the charming habits of commuting children, and the beauty of an overflowing trash can. Those dispatches were met with such enthusiasm that, this past January, Hill turned them into a monthly mailed zine called The Cloud Report, filled with off-the-cuff notes and sketches of her surroundings. “It can appear quite rough or hand-drawn, but I’m doing this intentionally—trying to make something that really feels like it was made by a human being,” she said. She expected that a few local fans would enjoy her project, but one viral video on TikTok led a thousand paying subscribers to sign up within twenty-four hours. She now prints and mails the publication to more than three thousand people, some overseas. Hill sees the appetite for her work as part of a broader craving for the handmade in a moment of A.I. automation. “I’ve been seeing artists talk about leaving typos in their work, just to sort of signal a person wrote this,” she said. “We are moving into an age where people are going to relish any little detail like that that signals there was room for error.”
Visual perfection is easy to come by these days: just type whatever you’re envisioning into a text box on ChatGPT, Gemini, or Midjourney, and back comes a glossy, detailed digital rendering of, say, your living room with fancier furniture, an aspirational outfit, or a new company logo replicating mid-century-modern style. This instantaneous slickness—which the trend forecaster Emily Segal has labelled “tasteslop”—is easy and abundant, and the appearance of “good taste” is robotically ubiquitous. So it’s no wonder that the artists and designers whose livelihoods are threatened by it have begun moving aggressively in the opposite direction, toward the scrawled, the sloppy, the seemingly mistake-riddled. The poster for an upcoming Weezer festival at a Los Angeles arena looks like it was drawn in marker by a ten-year-old boy on a school desk, with misaligned handwriting, doodles of guitars, and a few iterations of the “cool S.” The publisher Picador recently released a line of reissued books by Roberto Bolaño with covers reminiscent of a prison-tattoo collection, with amateurish sketches of femmes fatales, knives, snakes, and sigils. A professional wrestler named Orange Cassidy walks into the ring to an enormous, heavily pixelated video projection of his slogan, “freshly squeezed,” with its plain type blurrily surrounded by fist icons that call to my mind the nineties digital-art software Kid Pix. Charli XCX just announced a new event in New York with what looks like a handwritten note dashed off on stationery. What these creations have in common are elements of deliberate casualness, accident, even confusion—qualities that an A.I. tool, trying to satisfy users with clean efficiency, would typically avoid.
Shaun Singh, the founder and C.E.O. of Death to Stock, a provider of culturally savvy, stylishly haphazard stock photos, told me that this new wave “strips away any optimization” and “removes expectations of polish.” A.I. has made super-high-resolution images so standard, he added, that they have lost all currency as signals of quality. Distinguishing oneself aesthetically requires letting conspicuous human effort show. For a recent Death to Stock campaign, Singh’s team lugged suits of armor and swords to a castle in Berlin and hired a cameraperson to document behind-the-scenes footage of the stunt for social media—proof that the fantastical photos were not just generated by a machine. “We are constantly battling A.I. visuals,” Singh said. Even Apple, usually a secretive company, recently released a similar behind-the-scenes glimpse into the making of a MacBook Neo commercial, showing real people building props and filming hand gestures, as if to prove the ad’s authenticity. The video highlighted the ad’s “tactile and handmade production methods,” Tor Myhren, Apple’s vice-president of marketing communications, told me, emphasizing “human creativity.”
The shift is especially apparent in the realm of graphic design. In January, the clothing label Guess launched a rebrand campaign composed of lowercase copy in acid-yellow type over grainy selfie-style photos; the Guess logo is printed over patchy white highlights, as if someone had hastily scratched out the surface beneath the word in Photoshop. The sense of raw-edged risk made the campaign a runaway hit. It was the work of Travel Agency, a small New York-based creative agency in collaboration with Richard Turley, the designer famous for giving Businessweek a brash, discordant house style. The intention was to embrace “randomness that comes from people,” Zack Cohn, the co-founder of Travel Agency, told me. Such “bad design” “actually performs better than ‘good design,’ ” he added, by standing out in the too-clean crowd. This strategy has been adopted even among A.I. companies themselves. Granola, an A.I. note-taking service for work meetings, recently rebranded with a rough-hewn spiral logo in a soft, organic green. Max Ottignon, the co-founder of Ragged Edge, the branding agency that did the new logo, told me that the style was “janky in places, by design.” (It provides a contrast with OpenAI’s branding, which is “very precise” and “could be the identity for a defense firm,” Ottignon said.) Anything that appears too smooth these days is suspicious; the more refined-looking, the greater the chance it was machine-generated. As Ottignon said, “A.I. is a force multiplier of mediocrity.”
Embracing a visual vocabulary of the lowbrow and the rudimentary is a tried-and-true method of rebelling against a culture that feels vapid or corporatized. Painters such as Jean-Michel Basquiat and Rose Wylie did as much when they intentionally incorporated shakily drawn lines and snatches of text into their work, as if to invite arguments that a child could have painted their canvases. The strategic weakness in the current vogue for the unpolished and the hand-drawn, of course, is that A.I. can also be childlike if you ask it to. Recently, a viral prompt on social media asked ChatGPT’s image generator to turn photographs into drawings done “in the most clumsy, scribbly, and utterly pathetic way possible,” with a “low-quality pixel-by-pixel feel.” The result was an avalanche of rudimentary MS Paint-style selfies, memes, celebrity portraits, and tech logos that could have been the work of a Reddit-obsessed eight-year-old equipped with a vintage computer. Whatever their semblance of lo-fi charm, though, the drawings still betray certain hallmarks of robotic sameness: the brushstrokes are all of equal width, and the images are all composed with a centered focal point—signs of an A.I. stuck in its default aesthetic modes. Humans will have to determine what degree of machine-made messiness remains on the right side of cool. ♦