A contest built around the brief
Mondelez Japan announced the Clorets Mint Tab comparison on June 6, 2016. Its web campaign invited viewers to watch two commercials and vote on which communicated the product proposition better. The opposing creative directors were AI-CD β and Mitsuru Kuramoto. These web films should not be confused with the separate television commercial also mentioned in the announcement.
The setup made process part of the entertainment. Rather than ask people to inspect an invisible production tool, it gave them a comparison to discuss. That is a communications idea in its own right, whatever the eventual preference between the films.
Direction is not production
McCann’s March 29 agency announcement describes a database of structurally analyzed, tagged commercials, including ten years of ACC award winners. Entering a brief produced creative direction. It does not disclose a modern generative model or establish autonomous end-to-end filmmaking.
A useful reading therefore separates recommendation from execution. An instruction can propose a tone or narrative device without deciding every performance, camera position or edit. The comparison should not erase the people needed to turn direction into a viewable advertisement. Nor should the robot’s memorable physical form be mistaken for a complete explanation of the software behind it.
What a public vote can tell you
A campaign vote can be an invitation to participate without functioning as a controlled experiment. To interpret one as evidence of superior creative ability, a researcher would want to know who voted, what they were told, how the alternatives were presented and what exactly the question measured. Preference, accurate product recall and willingness to buy are different outcomes.
There is also a distinction between judging two finished advertisements and judging two production systems. One strong execution cannot establish that a process will outperform across future briefs. These are limits of the inference, not allegations about the conduct of the promotion. This entry does not report an unverified vote result.
The archive lesson: keep the handoff visible
The most transferable idea is the handoff between a structured reference collection and a working creative team. A useful case archive would retain the original brief, the system’s recommendation and the finished execution separately. Readers could then see which ideas traveled intact and which depended on interpretation.
That documentation is more instructive than assigning a single authorship label to an entire film. It also makes disagreement productive: a team can reject a recommendation without declaring the whole technology useless. Clorets belongs in this history as a public test of delegated creative direction, not as proof that software replaced all the work involved in making advertising.
Sources & limits
Contemporaneous brand release; campaign mechanics verified, no winner or effectiveness claim adopted
- Clorets Mint Tab commissions AI-CD β for a human-versus-AI commercial campaign — Mondelez Japan — 6 June 2016.
- AI-CD β joins McCann Erickson — agency announcement — 29 March 2016.
Send a correction with the passage and supporting source.