Everyone keeps repeating the same comforting line about AI content: just keep a “human in the loop” and you’ll be fine. I’ve nodded along to that advice plenty of times myself. Then I read a post that flat out rejects it, and I think the author has a point.
This expert spoke at the Content Marketing Summit (CMS) 2026 in Seoul, organized by John Park (박세정) and his team at DMK, and came away with one big takeaway: “human in the loop” is not enough. Being somewhere in the process doesn’t mean you’re steering it. The talk came down to one idea: directing is a human skill.
🎬 Productivity Is the Wrong Scoreboard
A lot of AI talk is about speed and volume. The author pushes back on that. Yes, you can direct a team of AI agents to get things done, but productivity itself isn’t the goal. The outcome is.
That changes the question you ask. Instead of “how much can AI make for me?”, it becomes “what am I trying to create, and is this getting me there?” Only a person can really answer that second one.
😊 The Content Smile Curve
To explain where humans still matter, the speaker shared the Content Smile Curve, an idea borrowed from PC manufacturing. Yes, the author’s view is that we’re in the content manufacturing industry now. Picture a smile: the value sits high at both ends and dips in the middle.
- Left end (human): ideas, originality and taste
- Middle (AI): production, where AI does the work, if you know how to use it
- Right end (human): feedback, ownership and delivery
This is what surprised me most. Plenty of people open a chatbot and ask for ideas first, which hands over one of the two most valuable parts of the whole process.
🧠 What Brands Said About AI and Ideas
The panel backs this up. The author moderated a discussion on how brands actually use AI in ideation, and two voices stood out:
- Pei Ling Ho (Duolingo): culture can never be prompted. Don’t use AI for ideas. Use it to check which ideas you should not use, and never take its default answer.
- Tomoka Takada (Hakuhodo): AI sees patterns in past data, but some insights never come from a large language model. They come from creative people with real cultural experience.
I love Pei Ling Ho’s flip. AI goes from idea machine to filter, and that’s a much smarter job for it. The panel’s conclusion explains a lot of what we scroll past every day:
So much content looks the same because people automate without directing. You need to be in the director’s chair.
🎥 From Idea to a 17-Minute Film
The strongest evidence came from the workshop, because this wasn’t only theory. In two hours, the group went from idea to final video. The creator’s lesson was that complex stories need planning, plus a set of AI agents and skills.
As a live demo, this AI professional turned three chapters of Journey to the West into a brief, then into video prompts. Back home, the author finished the project over a weekend: a 17-minute film. That’s what directing looks like in practice. The story decisions came first, and the tools followed.
🛠️ How to Get Into the Director’s Chair
Here’s how I’d put these ideas to work on your next project:
- Write your own ideas before you open any AI tool.
- Use AI as a filter. Try something like: “Here are my five campaign ideas. Which ones feel generic or overused, and why?”
- Never accept the first answer. Push back, ask for alternatives and apply your own taste.
- For bigger projects, write a brief first, then turn it into prompts for your AI tools.
- Keep feedback, ownership and delivery in your own hands.
The author’s closing warning stuck with me. If you always ask AI, you’ll get the same ideas, the same challenges and the same possibilities as everyone else.
Think different first. Then use AI to bring it to life.
Check out the full LinkedIn post to read the author’s original take from Seoul. It’s well worth a couple of minutes before your next prompt.