If You're a Creative Who Hates AI, Shut Up and Watch This

TLDR: Dean Clark ran a six-figure creative studio for thirty years until AI eliminated his client base in under a year. His response, and the mindset behind it, is the most honest account of creative adaptation I've heard. The lesson is not about the tools. It's about moving up a level.

Summary: Picture a teenager in the back of a Hollywood car in the mid-80s, spotting a pink car that matches Axl Rose's jacket, grabbing a film camera, and capturing a shot that ends up in Rolling Stone and on an album cover decades later. That was Dean Clark. He spent thirty years running a premium creative studio, doing work for A-list brands that no one else could do at his level. Then AI arrived and, in his words, "said hold my beer." Within a year, the clients stopped calling. Work he used to charge six figures for could now be faked by a non-artist with a prompt in an afternoon. His industry didn't shrink. It vanished.

The part of Dean's story that sticks with me is his framing of what AI actually does to markets. It doesn't replace skill at the top. It destroys the middle. Everyone can now produce something average, a passable logo, a serviceable paragraph, a decent video. That sounds like democratization until you notice the side effect: average is now free, so average is worthless. The value didn't disappear. It migrated. It moved from the ability to make things to the ability to judge which things are any good. Dean's phrase for this is "taste is the new skill," and I think that's one of the cleaner formulations of the current moment that I've come across.

He gave a concrete example that illustrates the shift better than any abstract argument could. A major brand was relaunching a sneaker with a basketball superstar. The plan was to pay the city of Chicago a million dollars to halt a subway train, light a tunnel, and shoot a three-hour window. Six months of planning. It fell apart days before launch. Dean's team got an emergency call and built a shot-by-shot proof of concept in ninety minutes using AI. The executives approved it. They shot the athlete on a white soundstage. The shoe sold out in two hours. That's not a story about AI replacing humans. That's a story about what a human with taste and judgment can do when the tools amplify their speed.

The honest version of that story, which Dean tells himself, is that the commercial wasn't "100% AI." He went frame by frame through thousands of frames, re-prompting, fixing, pulling pieces into Photoshop, bending them manually. Animators, sound designers, and editors still did the work that only humans could do. His operating principle became "use AI for what only it can do, but if people can still do the job and there's budget, still employ people." That's not a platitude. It's a production philosophy that kept the work good and kept humans in the room. The brands that get burned are the ones who let the tool make the taste decisions. The ones who win keep a human with judgment sitting between the model and the market.

Dean's personal life during this transition is worth knowing because it puts the professional shift in context. About a year ago, almost simultaneously, he had a massive heart attack, his thirty-five-year marriage ended, and his business cratered. He calls it his "Job year" after the biblical figure, and also the best year of his life, because it forced him to climb. He's now a creative director at an AI ad studio, prepping for Cannes. The trajectory is not a comeback story in the conventional sense. It's more specific than that. The tools kept eating the floor he was standing on, and his response every time was to go up a level. That's the whole playbook.

Key takeaways:

AI raises the floor for everyone, which makes average output worthless and makes taste and judgment the scarce resource worth paying for.
The creative economy shifted from a creator economy to a curator economy. The best curator wins, not the fastest executor.
Effective AI-assisted production is still human-directed. Frame by frame, judgment call by judgment call, with AI doing the parts only it can do.

Why do I care: From where I sit in frontend architecture, this maps directly onto what's happening with code generation. The engineers who will struggle are the ones whose value was "I can write this boilerplate faster than you." That floor is gone. The engineers who will do fine are the ones whose value is "I know which architecture won't fall apart in eighteen months, and I can tell the difference between generated code that's subtly broken and code that will hold." Taste, judgment, and the ability to read the output critically. Dean arrived at the same conclusion from a completely different direction, which tells me the pattern is real and not domain-specific. Go up a level. Stop defending the floor.

If you're a creative who hates AI, shut up and watch this!