There’s a familiar anxiety running through every industry that touches creative work right now: writing, design, music, advertising, film. The assumption underneath it is simple. If a machine can produce the artifact, it can replace the person who made it.
Benny Lee has spent his career at the intersection of design, emerging technology, and enterprise AI, most recently leading AI design strategy at Newell Brands after twelve years building creative and immersive experiences at The Coca-Cola Company. When we talked, he offered a different read on that anxiety.
“If you think AI is gonna replace you because it’s going to be able to replicate your work… then you probably already had a problem to begin with,” he told me.
He followed that with the line that gives the whole statement its bite: “It’s just a pixel, it’s making pixels faster, right? It’s not really replacing the value and the thinking behind it.”
AI is making it dramatically cheaper and faster to produce creative output. That’s real, and it’s not slowing down. But speed of production and value of creation are two different things, and the gap between them is where this conversation actually needs to happen.
We’ve confused making something with creating something
Generative AI can produce an image, draft copy, generate a song, cut a video, and iterate faster than any human team ever could. Judged purely on output, that looks like a replacement.
Benny’s distinction points somewhere else entirely. Production is one layer. The decisions that come before production are a different layer, and they’re the one a prompt can’t originate on its own.
Someone still has to decide what the piece is even about. Someone still has to decide why it needs to exist, who it’s for, what they should feel, what stays, and what gets cut.
Benny made this point with an example closer to his own work. Coca-Cola’s association with the modern image of Santa Claus goes back to the 1930s, so when the brand decided to bring generative AI into that legacy through the Create Real Magic AI Santa campaign, the technology was never the starting point. Someone had to decide the campaign needed to exist. Someone had to decide it should build on decades of Santa imagery instead of reinventing it, and someone had to decide what the AI-generated content needed to feel like to still read as authentically Coca-Cola rather than a novelty. AI didn’t originate any of those calls. A person made them, and everything the technology touched afterward depended on it.
Benny put it simply: the decision belongs to the person, not the tool.
Creation starts with intention. Generating is what happens after someone already knows what they want. Those are not interchangeable skills, even when the finished product looks the same.
When execution gets easy, judgment gets valuable
Producing something polished used to require time, technical training, and often real money. Those barriers are falling fast. More people can now put together a professional-looking image, deck, video, or campaign in an afternoon.
That changes what actually sets people apart. “I can make this” stops being the differentiator it used to be. What takes its place is taste: knowing what’s actually good, having a point of view, recognizing what an audience will feel and why.
Benny pointed out a real risk that runs alongside this shift. AI can hand you something that looks completely right even when it isn’t. “It’s presenting to you a solution that’s so confident,” he said. The same thing happens creatively. Work can look finished and still say nothing. Polish and insight are not the same thing, and AI is very good at manufacturing the former.
More options, faster, is a genuine gift. Someone still has to recognize which of those options is worth keeping.
Removing the human doesn’t prove AI is creative. It proves nobody was directing it.
Benny described watching people try to get AI to run an entire creative process off a single prompt, and pointed to filmmaking as the clearest example. “People are trying one prompt, the entire movie comes out. No,” he said. “You gotta chop it up by scenes, by lines, by angles, by shots.”
His approach breaks work into pieces and matches each one to the right tool. “Use the best tool for the best time, for the best place,” he said. Then someone still has to assemble those pieces into something coherent. That reframes a lot of what people call AI slop. The problem usually isn’t that AI touched the work. The problem is that nobody exercised judgment over what it produced.
Benny’s own workflow bears this out. AI helps him brainstorm and build a rough foundation. “I found myself using AI more for brainstorming and helping me build a solid foundation,” he said. But he still owns what comes out the other side. “I’m still telling the story,” he said, and, “I still have the decision to pick.”
A human sets the direction, AI compresses the work in the middle, and a human decides what survives. Benny put it plainly: “There’s always a human loop at the beginning and the end. What we’re doing is just compacting that middle part… from ten steps to a step.”
He offered a good example of what that compression actually looks like in practice, and where it goes wrong. We talked about a design director using AI to generate 40 different options for a project, rather than doing the harder work of narrowing toward the strongest few. Benny’s point was that this isn’t new territory for a design director. Before AI, that same director’s job was managing a room of junior designers: giving them direction, reviewing what came back, and pushing them toward the strongest ideas rather than keeping all of them. AI hasn’t changed that job. It’s replaced the junior designers generating the raw material, but the director’s real work, choosing what’s actually good and cutting what isn’t, is exactly the same skill it always was.
The opportunity here isn’t automating creativity out of existence. It’s stripping away enough friction that people spend more of their time on the parts only they can do.
Creativity includes knowing what shouldn’t change
Benny’s years at Coca-Cola gave him a useful way to think about what’s worth protecting when everything else is moving fast. “The beverage hasn’t changed for 140 years, it probably won’t change for another century,” he said. “If the product doesn’t change, we change everything around it.”
Technologies come and go around that constant. XR was useful for a while. AI is useful now. Something else will replace both eventually. “It’s just a tool to tell a story of the Coke brand,” he said, and that line applies well beyond soft drinks.
Every organization bringing AI into its creative work needs an honest answer to one question: what is the core that the technology is supposed to amplify, not replace? For a brand, that might be trust. For a creative professional, it’s usually perspective. For a leader, it’s judgment.
Real innovation isn’t changing everything just because the technology allows it. It includes deciding what deserves to stay exactly as it is.
The leadership job is building better thinkers, not better prompters
If production keeps getting more automated, the skill organizations actually need to build isn’t tool proficiency. Tools will keep changing. What holds up is the ability to ask a sharper question, connect two ideas that don’t obviously belong together, push back on mediocre output, and read what will actually land with a human audience.
Benny’s own definition of AI literacy gets at this directly: “Helping people understand where AI lives and where there’s a time for AI and where there’s a time not to use AI.” He pushes back hard on the instinct to reach for AI as a default. “I always ask them, should we even be doing AI?” he said, describing a pattern he sees constantly inside organizations: teams reach for AI even on small, routine tasks that don’t need it. His example was something as simple as placing an order. The instinct becomes “AI could do that for you,” whether or not AI is actually the right tool for the job.
That instinct to apply AI everywhere runs into a second problem: most people don’t have an accurate picture of what the technology can actually do right now, as opposed to what it might eventually do. Benny is blunt about the gap between the sales pitch and the current reality. “Understanding expectations of what it can do today,” he said, is half the battle. He’s watched people get sold on capabilities that simply aren’t there yet. “Maybe in ten years,” he said, “but today” the technology can still deliver plenty, just not the version some vendors are selling.
That’s the harder, more durable skill leaders need to develop in their teams. Access to the technology was never the hard part.
The bar just got higher
AI will replace some creative tasks outright. It will make plenty of others much faster. It will let people without years of technical training make things they couldn’t have made before. None of that makes human creativity less essential. If anything, it makes it harder to fake.
It just makes shallow creativity a lot easier to spot.
When producing something polished gets cheap and nearly instant, the question stops being can you make it and becomes do you have something worth saying. AI can generate the pixels, the words, the variations. Someone still has to decide what deserves to exist.
As Benny Lee put it, someone still has to be holding the steering wheel.
