Ten years after his last visit to Ad Black Sea, Mihnea Gheorghiu returned to the festival with a confession. For almost four years, he had not given a talk like this. He had done plenty of interviews, though, and lately they all seemed to contain the same question: “What do you hate about advertising?”
Apparently, years of optimism had finally caught up with him. So, fittingly for an Ad Black Sea edition built around On the Edge, Gheorghiu decided to come back with something less polished: a collection of hot takes about an industry that desperately wants to make brilliant work and somehow keeps producing things nobody cares about.
His defense was simple: “It’s not hating if it’s venting.”
Nobody, he argued, wakes up wanting to make bad advertising. Creatives walk into agencies convinced that today they will “crush it,” make something culturally relevant and perhaps even sell everything in the process. Yet, in his deliberately generous estimation, around 95 percent of advertising still ends up mediocre. One reason is the industry’s obsession with making sure absolutely nobody is upset.
To explain it, Gheorghiu told the story of Eleonora, a German housewife who appeared in research for one of his campaigns. She disliked three things about the ad. Unfortunately, those happened to be “the only things that made that ad at least edgy or nice or funny.” The team went back to work. Dozens of people across agency, client and production spent days trying to solve the problem until Eleonora finally approved.
The imaginary ending is painfully familiar. The ad eventually reaches her feed and she responds: “Ew, advertising. Skip.”
For Gheorghiu, the story captures one of advertising’s strangest habits. Enormous amounts of time and money are spent sanding ideas down until no one can object to them. “There’s a big difference between everybody loves something and nobody has a problem with it,” he said. The industry often celebrates the latter as success.
Then AI arrived, and exposed the problem.
“We got to 2026 where we discover, huh, oh no, AI can do that mediocrity we do every day much faster.”
His point was less about whether AI is good or bad for creativity than about what exactly humans are trying to protect from it. If advertising is reduced to familiar images, safe emotions, predictable formats and work engineered to offend nobody, machines are already extremely capable competitors.
Gheorghiu used Coca-Cola’s AI-generated Christmas advertising as an example of the contradiction. The work tested well, he noted. Consumers responded to familiar festive cues, trucks, Christmas imagery and the warm emotions associated with the brand. Then people discovered AI had been involved, and the reaction changed. Suddenly, the work was condemned. The irony fascinated him: people could enjoy something until they knew how it had been made.
There is, however, one place where Gheorghiu still sees real room for creative disorder: social media. “It’s truly the last medium where we can truly experiment,” he said. On social, a brand is not simply competing against another brand’s beautifully art-directed campaign. It is competing with whatever a person might prefer to watch next, from an absurd trend to a completely random video.
That changes the rules. The potential audience is almost infinite, but so is the risk of getting something wrong. Gheorghiu thinks the industry should accept the trade-off. “We can finally, finally be pirates,” he said, comparing social to a kind of creative buccaneering era where many of the old rules are still negotiable. “Everything is up for grabs. Like, there’s no rules. Just go do weird, random shit. This is the only way to see what works and what doesn’t.”
The work he admires most often contains precisely that loss of control. He spoke about projects designed to escape the brand’s own channels and become stories people wanted to pass around themselves. Once something genuinely enters the internet, he argued, the brand cannot completely control who films it, how it travels or what happens next. That uncertainty is part of the value. “A good story will always travel faster than any amount of paid media you put behind it.”
That same skepticism toward control shaped his take on the creator economy. Gheorghiu likes what creators have done to content production. Anyone with a phone can participate, and creators can often move faster and more resourcefully than traditional agency teams. What he dislikes is the assumption that a large following automatically makes someone the right creative partner.
“If you wanna pick a creator, first of all, go for their tone of voice. Second, go for their fucking craft. Third, go for their engagement,” he said. Following should come much later. Otherwise, brands are often treating creators as little more than cheaper media channels.
His advice to creators was even clearer: say no more often.
Accepting every brand brief can push a creator into what he described as a “sausage-making machine,” where the final piece no longer sounds like the creator, the brand or anything the audience originally cared about. “Protect your craft and your tone of voice at all costs,” Gheorghiu said. “What put you there in the first place as a creator is your craft, not the contracts you have with brands.”
For agencies, the keyword should be collaboration. The agency brings an understanding of strategy, history and brand voice; the creator brings a different instinct, speed and closeness to an audience. Trying to make one behave like the other defeats the purpose.
And then there is “culture,” perhaps advertising’s favorite word and one of Gheorghiu’s favorite targets.
Brands endlessly claim that they want to “be part of culture,” but too often this means spotting something people are already doing and reproducing it. Gheorghiu compared that behavior to arriving after somebody else tells a joke, laughing along and then expecting credit for being funny.
“Culture is never the idea. It’s the brief.”
Being culturally relevant requires adding something to what is already happening, changing it, challenging it or taking it somewhere unexpected. Simply mirroring culture does not make a brand part of it.
That idea led to perhaps the most useful measure of success in the entire talk. Gheorghiu is skeptical of spectacular case-study numbers that claim billions of impressions while leaving almost no trace on the people behind those numbers. He would rather affect a very small group deeply than reach millions who feel nothing.
Advertising does need results, he acknowledged. “We’re not art.” But reach alone is a poor substitute for relevance. “Your only KPI is to be relevant,” he argued. Once people genuinely care about a brand, selling something to them becomes considerably easier.
After nearly forty minutes of talking about AI, creators, social platforms, cultural relevance and whatever new format the industry may become obsessed with next, Gheorghiu reduced the whole discussion to something much older.
“You gotta be a story. That’s literally the most powerful thing in the universe.”
AI will change. Platforms will change. There will be another new format, another trend, another technology everyone feels obliged to master. Without a story, none of them will rescue mediocre work. With one, the formats become secondary.
And perhaps that was where Gheorghiu’s rant aligned most clearly with On the Edge. He did not end by promising that creativity would save advertising. In fact, his conclusion was considerably less sentimental: “I don’t wanna do good work to save advertising. Fuck advertising. But doing good work will save the people who know how to do it.”
His final advice was to find those people.
In an industry moving through AI, collapsing conventions and an increasingly unpredictable culture, “always surround yourself with doers,” he told the audience. Not the doubters, the corner-cutters or “the person on LinkedIn who has a ChatGPT-written opinion about the next big thing.”
Whatever advertising becomes next, Gheorghiu’s wager is that the people willing to actually make interesting things will survive it.
And on the edge, that may be one of the safer bets left.
