It was a few weeks before Mother’s Day, and Yagnesh Mehta wanted to boost food sales at his branch of PJ’s Coffee inside The Mall in Columbia. So, as Mehta so often does these days, he turned to his right-hand man with all the answers: ChatGPT.
The popular AI tool helped him come up with a promotion — $17.99 for beignets and two coffees — that offered customers a discount while maintaining some profit, “so we’re not giving away food.” Mehta also provided ChatGPT with an image of PJ’s beignets to help create a flyer, though the chatbot made the treats look “prettier” than the reference photo. The final ad featured a middle-aged mom and her grown daughter smiling at each other, holding iced coffees in front of a plate of the golden pastries.
“Make her feel loved,” reads the text. “Make it a PJ’s moment.”
But when Mehta posted the flyer on the 20,000-member Howard County Eats Facebook page, the response was prickly. “With an AI mom?” one commenter wrote. They vowed never to return to the business.
Though some Americans worry about AI’s impact on creativity, relationships, the environment and the job market, its use has exploded across industries. Increasingly, ordinary people are also using AI for daily tasks, from drafting emails to checking medical symptoms.
But AI adoption in the food and beverage world is fraught with controversy as customers and others speak out against it, despite some entrepreneurs touting the technology as a leg up in a competitive market.
Read More
“People are starting to revolt against the use of AI,” said food photographer Steve Vilnit, who’s noticed an uptick in restaurants promoting themselves with ads that use artificial intelligence. “People are getting upset by that — they feel like they’re getting scammed.”
Some of the most vocal objections center on food photography. While it can be hard to discern AI food photos from the real deal, the former often feature uniform lighting and backgrounds or slightly “off” patterns and textures. Researchers have hypothesized that the phony images trigger a visceral revulsion in people, similar to the “uncanny valley” effect when looking at a humanoid robot.
On the Howard County Eats page, commenters are quick to point out suspiciously perfect pizza. “Can we ban posts from restaurants that use AI images of food instead of the actual food that they make?“ someone asked in May.
Restaurateurs argue, though, that the chatbot can take the place of a consultant or a designer whom they otherwise wouldn’t be able to afford.
Mehta, who has a day job as a mortgage lender in addition to being a cafe owner, said ChatGPT makes him feel like he cloned himself. He used the technology to analyze sales data that helped him determine that PJ’s food sales were underperforming and decide what to do about it. To him, it’s about survival. “The small businesses of the world that don’t use AI will just be eaten up by the large corporations of the world that use AI,” he said.
For Mary Miles, who owns cheesesteak spot Dimples Bar and Grill in Locust Point, “it’s like having an assistant that doesn’t talk back.” She consults with ChatGPT all day every day, even calling it “him,” and addressing it like a person. “Good morning, Chat,” she types. “The damn thing calls me ‘Ms. Mary,’” she laughed.
Since April, when her grown children urged her to start using the technology, Miles has used the platform to throw together flyers for her social media. As she prepares to launch another branch of her restaurant in Fells Point, ChatGPT has helped her figure out how to scale her inventory up or down and set prices.
But Miles understands why customers might push back against AI in food photography. While her ChatGPT-generated flyers incorporate her own snapshots of the dishes she serves, the bot “sort of airbrushes the photo,” and sometimes takes it too far. If she uploads a photo of a cheesesteak, she said, “he’ll add green peppers.”
Commenters often gripe when Ed Reynoso posts AI-generated images on the Howard County Eats Facebook page to promote his Columbia restaurant, Celia’s Cuban Cuisine and Mojito Bar. One graphic features a photo Reynoso said he took of his own tamale against an AI-generated background that reads “the tamale hall of fame.”

Reynoso said he’s baffled by some of the AI backlash. Though critics often say AI is taking jobs away from people, when it comes to his own posts, “I’m like, what job? I just created a funny flyer.”
Reynoso said he’d be reluctant to consult with ChatGPT on the more serious work of running his restaurant. “My time, I think, is better spent in the kitchen than it is on the computer,” he said. “If the food’s not good, it’s not gonna matter regardless.”
Yet AI food photos are becoming so pervasive that one local restaurateur was surprised to see photos he didn’t recognize on his eatery’s website. An online ordering platform for Benny’s in Little Italy features some real images, such as a picture of the spot’s signature meatball dish: The bowl is off-center, the lighting is natural, and the table and wine glass in the background are identifiable. But the picture accompanying the Caesar salad, for instance, looks stock-photo perfect, albeit with an unusually speckled yet uniform texture. It’s AI, said co-owner Joe Gardella, though he hadn’t realized; his business partner had set up the online menu.
“I don’t like that,” Gardella said. “I want people to see our food.”





Comments
Welcome to The Banner's subscriber-only commenting community. Please review our community guidelines.