A galaxy of noodles: the AI idea that was dropped, not patched
What a failed concept in a fully AI-produced pitch film shows about where human judgment has to sit.
A pitch film for European wholesale buyers, produced entirely with AI. One concept came out wrong. Here is why it was dropped, and what replaced it.
The brief
Samyang needed to enter the EU wholesale market. There was no footage, no lifestyle material and no narrative. Only charts, data and product specs. The job: a pitch film for category managers at major European retail chains, made entirely with AI video generation and motion graphics. No live shoot.
The first idea
In ideation the idea went toward a galaxy of flavors rising from every noodle cup. It sounded big. The AI kept pulling it further toward abstraction.
What came out
In execution the images came out like bloody brains, almost zombie-like. Far from the brand's identity, which is crazy fun.
Dropped, not patched
The idea was wrong, so it was dropped, not patched. A better prompt, another model or a new grade would only have polished the wrong idea. The rule that came out of it: when the failure is in the idea, no tool can fix it. Only a person can say so.
What replaced it
A five-act, two-minute film for European buyers, built on market data, a new face called Peppo and the XL format.
Market momentum: the success Samyang already has in Europe, and a new need the data shows.
The blind spot: a white space in the market matrix, the XL cups as the answer, Peppo as a spark from inside the cup.
Symphony of cravings: extreme macro food shots prove the no-drain, microwave-ready preparation.
Craving madness: absurd real-life moments, then the value math, a fair price per gram.
Shelf synergy: the XL cups beside the small ones on a premium shelf, nothing replaced, and an invitation to partner.
The fun came back, now with a business argument.
Where the human sat
Every visual was created from scratch with AI-generated cinematics and motion design. The decisions that mattered were not generated: killing the galaxy, building the film on market data, and ending on an invitation to partner instead of a sales line.
Full case with the rejected frames: manidastgerdi.com/portfolio/buldak-xl