SELLING FEELING IN THE AGE OF AI: HOW BRANDS LIKE KITKAT AND FLIPKART ARE QUIETLY REDEFINING WHAT 'CREATIVE' EVEN MEANS

Every week, the advertising veteran behind Bhatnaturally publishes a roundup of the campaigns that caught his eye — a curated sweep of what's clever, what's fresh, and what deserves a closer look. The format is deliberately casual: a few paragraphs per ad, a personal take on the craft, and a move to the next entry. No grand theory. No manifestos. Just practiced taste, applied in public. And it's precisely that casualness that makes the May 2026 edition so quietly extraordinary.

In a single post, three campaigns share the spotlight. KitKat's "Break Mode" is a physical product innovation — a limited-edition wrapper embedded with NFC technology that silences phone notifications when you peel it open, an idea rooted in decades of brand philosophy, real engineering, and a cultural insight about digital overwhelm. Flipkart's "SASA LELE" sequel rides on something equally rare: a coined nonsense phrase from an earlier campaign that escaped the ad and entered the vernacular, the kind of cultural penetration no brief can guarantee and no algorithm can reverse-engineer. And then, placed right alongside them, there's ChatGPT's India campaign — "Bala vs Waterfall" and its follow-ups — praised for depicting realistic use cases and bringing a smile with each solution. The blog notes the ads were crafted by agency Hue & Why, and that's about it. No asterisk. No caveat about the advertiser being an AI company marketing itself. No meditation on what it means for a large language model to be selling feeling through the same thirty-second format that KitKat and Flipkart use to sell chocolate and e-commerce.

The three campaigns are treated as creative equals. And that editorial equivalence — that shrug — is the most revealing gesture the advertising industry has made in months.

Consider what's actually being flattened here. KitKat's wrapper required material science, supply chain coordination, and a brand idea so durable it's survived over sixty years of reinterpretation. Flipkart's sequel demanded something arguably harder: the restraint to build on a viral moment without strip-mining it. ChatGPT's spots, meanwhile, are marketing for a tool whose stated commercial ambition — at least through its platform partners — is to eventually generate all ad creative without any human involvement. These are not the same species of work. They emerge from fundamentally different creative traditions, serve fundamentally different strategic purposes, and imply fundamentally different futures for the people who make ads for a living. Yet they sit in the same listicle, evaluated by the same standard: Did the ad land?

This is how cultural shifts actually happen — not through debate, but through absorption. As the CopyHackers blog has warned, when organizations remove the human element, they miss the subtleties that shape effective content — nuance, humor, timing. But that argument assumes the audience is still looking for the seams. What Bhatnaturally's roundup suggests is something more unsettling: the seams have already been smoothed over, not by technological perfection, but by editorial habit. When a respected critic with decades of fluency in what makes advertising work treats an AI brand's campaign as simply another well-executed spot, the rest of the industry receives a quiet permission slip.

The question this article sets out to explore isn't whether AI can produce a passable ad. That debate is over. The real question is what we lose when we stop interrogating the difference between a campaign born from human culture and one designed to replace the humans who make it — and what it means that we may have already stopped.

Start with KitKat, because the layers stack so neatly. The "Break Mode" wrapper is, on its surface, a packaging gimmick — an oversized KitKat sleeve that happens to double as a Faraday cage, blocking your phone's electromagnetic signal the moment you slide the device inside. But trace the creative logic backward and you find at least four distinct leaps, none of which follow from the last in any linear way. Someone had to begin with a brand platform — "have a break" — that has been running for decades and recognize that its meaning had quietly shifted. In 1958, a break meant putting down your work. In 2026, it means escaping your screen. That's the first leap: reading cultural drift against brand archaeology. The second leap is the diagnosis — digital addiction isn't just a talking point; it's a universal consumer truth of seeking digital detox that people feel physically, in the phantom buzz of a notification that never came. The third leap is the wildest: somebody in the room connected that anxiety to a piece of nineteenth-century physics, the Faraday cage, a conductive enclosure designed to block electromagnetic fields, most commonly found in medical labs and data-security facilities. And the fourth leap was deciding that the right format wasn't a film or a social activation but the packaging itself — turning the product's own wrapper into the mechanism of relief. Each jump crosses a different domain — brand history, cultural psychology, applied physics, industrial design — and the chain only holds because a human mind saw the thread between all four.

Now consider Flipkart's "SASA LELE" campaign, which required an entirely different kind of creative intuition. The original 2025 spot coined a nonsense term born from a visual pun on the word "sale," then built a jingle around absurdist product pairings — the kind of approach that, as the Bhatnaturally roundup noted, broke the mould in a category where most sale ads are merely catalogue displays. The real creative bet, though, came in the sequel. The 2026 version had to deepen the absurdity — "laptop crop-top" and the like — without losing the commercial function of driving clicks to an actual sale event. That requires judgment about tone thresholds: how much silliness can a brand sustain before the joke eclipses the offer? How do you build a recurring sonic and narrative property from a word that means nothing? These are calibrations no prompt can specify, because they depend on an intuitive grasp of audience patience, comedic rhythm, and brand permission that shifts with every viewing.

What ties the two cases together is the nature of the creative act itself. Neither campaign is defined by production polish or visual spectacle. They're defined by leaps — lateral connections between domains that share no obvious surface similarity. This is precisely the quality that Contently identifies when it describes the ideas that survive translation, ideas with enough dimensionality and cultural rootedness that they can't be flattened into a tidy AI-generated summary without losing the thing that made them work. A generative model can remix what it has seen; it can even surprise with novel adjacencies. But it cannot yet perform the deliberate, meaning-laden act of connecting a chocolate bar to Michael Faraday, or a nonsense syllable to a billion-dollar sales event, and know — without data, without A/B testing — that the connection will land. That knowing is the last human monopoly in the creative chain, and it is what both campaigns quietly put on display.

Google would like you to know that it has thought about this. When Ginny Marvin, Google's Ads Liaison, raised the specter of a "sea of sameness" during a recent Ads Decoded episode, she was giving the company's own product team a chance to address the anxiety head-on. And Charles Boyd, Google's Groupe Product Manager for Creative, delivered exactly the reassurance you'd expect: AI creative tools aren't homogenizing anything. They're infrastructure. They help advertisers produce more combinations, run more tests, and tailor messaging to more audience segments — all while the brand stays in control. Boyd described the value as "the ability to quickly create different creative styles and iterations at scale," and Google leaned heavily on the phrase "advertiser-in-the-loop" to signal that humans remain the strategists, the taste-makers, the ones steering the ship.

It's a comforting framework. It's also one that collapses a critical distinction: the difference between variation and differentiation.

Variation is dimensional. It's the same ad in sixteen aspect ratios, forty headline permutations, and three tonal registers calibrated to different cohorts. It's useful, even necessary, in a media landscape that demands creative volume at a pace no human team can sustain alone. But variation is not the same as having something genuinely novel to say. A Faraday-cage KitKat wrapper doesn't emerge from running more A/B splits. Flipkart's moment of anti-consumerist honesty during its own sale event doesn't surface from testing headline combinations. Those ideas required lateral thinking — the kind that defies optimization because it can't be reverse-engineered from performance data.

Google's position assumes that originality already exists inside the advertiser's strategy and that AI merely amplifies it. But the evidence from the market suggests something less flattering. A report released by Callan Consulting in April 2026 found that two-thirds of senior marketing specialists now say AI has a "strong" or "very strong" impact on their teams — double the level from a year earlier — and that half of organizations have already restructured their marketing functions around AI. The same report warns that overreliance on AI-generated content is producing "copies of copies," a phrase that captures the entropy perfectly: each generation of output slightly less distinct than the one before, quality and originality degrading in quiet increments.

Billo CEO Donatas Smailys put the implication bluntly. "AI is no longer a differentiator," he told the World Branding Forum. "Now everyone uses it, so the opposite is happening: human creators and real creativity are becoming premium." What Smailys is describing is the exact sameness that Google's tools are theoretically designed to prevent. If everyone has access to the same generative models, the same prompt architectures, and the same optimization loops, then the output converges — not because any single brand lacks a strategy, but because the underlying engine flattens idiosyncrasy by design. A megaphone makes a voice louder. It doesn't give you something worth saying.

The quiet sleight of hand in Google's framing is that it redefines "differentiation" downward, making it synonymous with "personalization at scale." And the industry, hungry for efficiency and under pressure to produce volume, is largely accepting the substitution. More formats, more sizes, more micro-targeted variants — all of it counts as creative output in the dashboard. But none of it answers the question that actually matters: does anyone remember the ad? Google's tools are solving for reach and relevance. What they're not solving for — what they structurally cannot solve for — is memorability. And that gap is where the real creative crisis lives.

When Donatas Smailys, CEO of the creator marketing platform Billo, declares that "human creators and real creativity are becoming premium," he is describing something more consequential than a pendulum swing in taste. He is describing the emergence of a new class system in advertising — one where the ability to reject AI becomes its own kind of luxury good.

Consider what it actually takes to stamp "No AI" on your campaign. You need the budget for a top-tier agency with human strategists, writers, and art directors. You need bespoke production: location scouts, cinematographers, editors who charge by the hour rather than the token. You need timelines measured in weeks, not the minutes it takes a generative model to spit out a hundred variations. In other words, you need resources. "No AI" as a brand signal risks becoming the advertising equivalent of "organic" or "artisanal" — a premium label that communicates affluence as much as philosophy. The brands most loudly championing human-only creation are, almost by definition, the ones that can afford to.

Below them sits a pragmatic middle tier, and its contours are becoming clearer. The trade publication Dash Two has framed AI's proper role in media buying as that of a "co-pilot, not driver," a formulation that captures how many agencies and in-house teams actually operate: using generative tools to accelerate brainstorming, automate asset resizing, or draft first-pass copy, while keeping strategic and editorial control firmly in human hands. This middle ground is functional and arguably honest, but it rarely makes for a compelling brand story. Nobody prints "Partially Assisted by AI" on a billboard.

Then there is the bottom of the hierarchy — the small businesses and solo operators for whom AI-generated creative is the entire operation. As platforms like Meta and Google embed generative tools directly into their ad infrastructure, a future is materializing in which a business can enter a product URL and let the platform generate all the creative without any human involvement. For a bootstrapped DTC brand or a local service provider with no design staff, this is genuinely liberating. It is also, in the context of the "No AI" premium, a marker of exactly the kind of operation that cannot afford the alternative.

The irony is sharp. AI was supposed to democratize creativity — to give the garage startup the same visual polish as the Fortune 500 brand. And in a narrow, technical sense, it has. But the cultural reaction is already clawing back that parity. If audiences learn to distrust or devalue AI-generated content, and if "human-made" becomes shorthand for quality in the same way "handcrafted" does on an Etsy listing, then the brands with the deepest pockets will simply convert their existing advantages into a new vocabulary of authenticity. The playing field doesn't level; it just gets relabeled.

The real question, then, is not human versus machine. It is who gets to choose. The luxury brands and global conglomerates will choose human-only creation when it serves the narrative, and deploy AI internally when it doesn't. The mid-market agencies will blend both and say little about it publicly. And the smallest players will use whatever they can access, increasingly aware that the tools saving them money may also be coding them as low-status. What began as a technological revolution is quietly hardening into a creative caste system — one where the line between "authentic" and "automated" maps suspiciously well onto the line between rich and everyone else.

The content marketers at Contently have identified what may be the defining structural tension of working in an AI-saturated landscape: you are now, whether you like it or not, writing for two audiences simultaneously. One audience is human — emotional, distractible, hungry for narrative. The other is algorithmic — a fleet of AI systems that "read, compress, and represent your content, often before a human sees your page." The strategic imperative, in this framing, is to craft content that satisfies both: stories vivid enough to move a person, structured cleanly enough for a machine to extract and redistribute without mangling the point.

It's a genuinely useful framework for anyone producing articles, white papers, or SEO-driven blog posts. But it also reveals its own limits the moment you step outside the content-marketing feed and into the broader landscape of brand advertising. The two-audience trap assumes your work will be consumed in a digital pipeline — crawled, indexed, summarized, served. It assumes the battlefield is the search result, the AI Overview, the Perplexity citation. And for a huge swath of marketing output, that assumption holds. But the most interesting creative work being done right now doesn't just survive algorithmic translation. It sidesteps the pipeline entirely.

Consider KitKat's "Break Mode" packaging, which Bhatnaturally described as a Faraday cage disguised as an oversized wrapper — a conductive enclosure that blocks electromagnetic fields and renders your phone unusable the moment you slip it inside. This is not content designed to be extracted by a large language model. It cannot be compressed into an AI Overview. It is a physical object that enacts the brand's proposition — "have a break" — by literally severing your connection to the digital infrastructure that would otherwise mediate your experience of it. The wrapper doesn't need to survive translation because it was never meant to be translated. It was meant to be held.

This is the escape hatch the two-audience framework doesn't quite account for. When machines become the primary gatekeepers of digital content, the most powerful creative move available to a brand is to produce something that exists outside the gate. That could be a physical product innovation like KitKat's wrapper. It could also be a piece of advertising so rooted in cultural specificity and sheer weirdness that no compression algorithm can reproduce the experience of encountering it — something like Flipkart's "SASA LELE" campaign, which stacks absurd product pairings like "laptop crop-top" atop a nonsense-word jingle that lodges in memory precisely because it resists easy paraphrase.

The danger of the two-audience model, taken to its logical extreme, is that it optimizes for extractability at the cost of irreducibility. You produce content so well-structured that a machine can summarize it in a sentence — and then a machine does, and nobody needs to visit your page. Feeling, by contrast, is irreducible. You cannot extract the experience of sliding your phone into a signal-blocking pouch and watching it go dark. You cannot compress the giddy nonsense of "SASA LELE" into a citation. The brands that understand this aren't just optimizing for two audiences. They are building creative work that refuses to be an audience for the machines at all — work that only makes sense when a human body is in the room, holding the wrapper, hearing the jingle, feeling the break.