Artificial intelligence is going to cure cancer, solve life’s mysteries and liberate humankind from drudgery. But maybe, when all is said and done, it will give you just the right pasta recipe to impress someone you have a crush on.
That seems to be the main message of an extensive advertising campaign for ChatGPT, the chatbot and virtual assistant developed by the company OpenAI.
The campaign’s 30-second commercials, created by the New York City creative agency Isle of Any, have a pronounced retro aesthetic, bringing to mind the days before everyone carried smartphones. Each ad emphasizes human beings and downplays the tech.
In a TV spot titled “Dish,” a young woman taste-tests a home-cooked meal prepared by a young man in his city apartment. As she takes the first bite, the expression on her face goes from neutral, perhaps even a little skeptical, to approving.
“Dish” is steeped in references to the old, analog world. The apartment is decorated with stacks of books, a little white gas stove that looks like a holdover from the 1950s and a glowing midcentury-modern light fixture. The song in the background, “Fool” by the indie artist Perfume Genius, is a 2014 throwback to 1970s soul.
White letters appear onscreen: “I need a recipe that says, ‘I like you, but want to play it cool.’” This is the prompt the young man has fed into ChatGPT. The camera pulls back to reveal the old building’s brick exterior. More text appears. It’s the chatbot’s reply, in the style of rolling film credits. “Here’s the move: Lemon-Garlic Butter Pasta with Cherry Tomatoes,” it begins. After laying out the recipe, the bot comes through with the encouraging words of a coach or buddy: “Above all, don’t sweat it. You got this.”
The ads have arrived at a time when many people are grappling uneasily with artificial intelligence. In a poll conducted by NBC News, 57 percent of registered voters in the United States said they believed it would do more harm than good. A Pew Research survey found that about half of Americans believed that A.I. would “worsen people’s ability to think creatively and form meaningful relationships.”
Isle of Any, which has done work for The New York Times, has so far created seven 30-second commercials for ChatGPT. In addition, several billboards created by OpenAI’s in-house team have appeared in Los Angeles, New York and other large cities. The campaign started last fall and is still going.
Each TV spot is meant to drop viewers into what feels like a movie scene, according to Laurie Howell and Toby Treyer-Evans, the founders of Isle of Any. The commercials were shot on custom Kodak film stock, and the production design was the work of Florencia Martin, who helped create the look of the Oscar-winning film “One Battle After Another.”
“Our focus was to find a cinematic, emotional way to celebrate the positive, enabling powers of ChatGPT,” Howell and Treyer-Evans wrote in an email. “We wanted this work to speak to how ChatGPT is a tool — a tool that works ‘with us, not for us.’” They added that “no part of the campaign was created by A.I.”
The ads feature young people exclusively. The casting was an attempt to target “the young cohort of people who were beginning to use it in their everyday lives,” Howell and Treyer-Evans said.
Ken Goffman, a tech writer who goes by the moniker R.U. Sirius, seemed puzzled that the campaign did not try to convey the full power and potential of artificial intelligence. “It’s a very cautiously denatured version of what ChatGPT and other A.I.’s are bringing to the world,” he said. “They seem to be reaching out particularly to dull young people.”
In many ad campaigns, the message is contrary to the product being sold. Insurance companies, which necessarily deal in car crashes and death, present a comic face to the world: Progressive goes heavy on the shenanigans of the character Flo and her pals, and Geico features a lovable lizard. Similarly, many commercials for prescription medications show people cavorting in idyllic settings, rather than the maladies that the drugs treat.
In the ChatGPT ads, a revolutionary technology — one that could be the most dangerous invention in human history, as Sam Altman, the founder of OpenAI, has warned — is presented as a helpful digital pal.
“There’s this scary, scary cultural backdrop about this Frankenstein monster technology that’s going to come and destroy us all,” said Randall Rothenberg, a longtime marketing executive and writer and the former chief executive of the Interactive Advertising Bureau, a trade association. “What’s the best way to create comfort around that? It’s to say, ‘No, it’s just an extension of you. It allows you to bring your full humanity out, to plan vacations with siblings.”
Rothenberg was referring to “Trip,” another retro-tinged ad in the ChatGPT campaign. It is focused on a Gen Z brother and sister who embark on a road trip in a Volvo 240 station wagon, a car that stopped being manufactured in 1993. The boxy vehicle recedes into a lush rural landscape to the strains of “Brother Love’s Travelling Salvation Show,” a minor hit for Neil Diamond in 1969.
The ChatGPT billboards are no less throwback. One shows a pair of young dudes who look like hipsters from a decade ago. They are sitting in front of a cluttered garage, apparently taking a break from their start-up project. It’s a restful image suggesting that the world hasn’t changed all that much since the 2017 publication of “Attention Is All You Need,” the landmark scientific paper that introduced the fundamental architecture for most machine-learning systems.
That touch of comforting nostalgia is what makes the ads effective, in the view of Robert Cornish, the chief executive of Richter, a corporate consulting and communications firm. “New technology often introduces confusion,” said Cornish, who wrote a blog post in praise of the campaign. “And confusion blocks sales and adoption.”
Rather than showing off the wizardry of ChatGPT, OpenAI has “ripped the page out of the Steve Jobs playbook,” Cornish added. He cited the successful “Get a Mac” campaign from the mid-2000s created for Apple by the ad agency TBWA Worldwide. Those ads show two actors — one cool, one square — meant to personify the Mac and other PCs. They stand against a white background discussing tech stuff in funny, everyday terms.
“If you want something to be adopted, take it down to a third-grade level,” Cornish said. “The ChatGPT ads help counter the fear around A.I. and make it safe for regular people. They need it to be relatable.”
The absence of authority figures in the campaign may send another signal, according to the political analyst and historian Thomas Frank.
“ChatGPT has definitely made parents redundant,” Frank said in an email. “In each installment of the series, we see young people alone in the world with ChatGPT guiding them along, always there for them, telling them how to study in college, how to exercise, how to cook, how to grow up, basically. It is the parent we all wish we had, who is omni-competent and omnipresent, totally forgiving, totally understanding.”
A.I. executives have pitched the technology as a kind of superbrain able to solve the knottiest problems. In a 2024 blog post, Altman promised “astounding triumphs” that would soon become “commonplace” as a result of A.I., including “fixing the climate, establishing a space colony and the discovery of all of physics.” Yet the ChatGPT ads cast A.I. as an improved search engine or virtual assistant.
That, apparently, is the point, Rothenberg said. OpenAI’s business model, like Google’s, relies on selling ads. And in the nearly three years since Google unveiled its own A.I. assistant, Gemini, the companies have been locked in a battle for attention. The ChatGPT ad campaign, in other words, has arrived at what could be a crucial moment for OpenAI.
Founded in 2015, the company is said to be preparing to go public even as it faces negative publicity. Recent news reports have described its profligate spending, missed revenue targets and lagging position in the A.I. race. In April, The New Yorker published a 16,000-word investigative article detailing the power struggles within OpenAI and portraying Altman as a congenital liar. The company also has a powerful enemy in Elon Musk, who has sued unsuccessfully in an apparent effort to drive it out of business.
Anthropic, the $900 billion company behind the A.I. model Claude, has taken an advertising route much different from OpenAI’s. During the last Super Bowl broadcast, it ran a punchy commercial that showed a young man seeking an A.I.’s counsel on how to improve his relationship with his mother. He grows suddenly disappointed when the chatbot’s advice morphs into an ad for a dating site meant to match young men with older women. “Ads are coming to A.I.,” goes the tagline. “But not to Claude.”
As OpenAI’s market share has slipped in recent months, Anthropic has generated media buzz for its supposedly “too powerful to release” Claude Mythos system and its decision to file for its own initial public offering. If OpenAI proves unable to convert new users in droves, it may find itself becoming to A.I. what Netscape was to internet browsers or Myspace was to social media.
“That gets us back to the ad campaign,” Rothenberg said. “OpenAI created the market and is most in danger of losing the market it created. They’re saying: ‘Hey, we’re not scary. Come to us.’”
