RESEARCH FILM

Distributed Creative Search on Digital Platforms

Read the full transcript

The same dish. The same creator. But a different camera angle, rhythm, or performance. When a creator tries something new, what actually changes?

Think of a video as three layers of building blocks. Conceptual content is what the work is about: the dish, the ingredients, and the theme.

Audiovisual production is how it is made: camera angles, color, sound, and editing. The same dish can be presented in a different way.

Embodied enactment is how the creator appears and performs: their face, movements, and posture. A familiar subject can be performed differently.

Now imagine many creators building at once. A new combination for one person may already exist on someone else's shelf. How does ongoing creative search change the group, and when does it go beyond what the group has made before?

We studied twenty-nine food and cooking creators on TikTok, using four thousand, six hundred and thirty-five videos. Three connected studies follow individual change, differences across creators, and departures from earlier group work.

We turn each video into three numerical representations. Captions, speech and on-screen text describe conceptual content. Images, sound and editing describe production. Face and body sequences describe performance. These representations let us compare works across all three dimensions.

First, what changes within a creator? Conceptual content accounts for about fifty-five percent of measured change, audiovisual production thirty-five percent, and embodied enactment ten percent. Creators most often change the dish, ingredients or theme.

The changes are usually uneven. In about seventy-seven percent of half-year comparisons, one dimension accounts for at least half the movement. The leading dimension also switches in about twenty-seven percent of the sequences we can follow.

But changing most is different from carrying forward. Between consecutive videos, audiovisual production shows the strongest association. A creator can try a different dish while keeping a familiar way of filming and editing.

We also compare successive works with recent personal history. When each work uses a separate historical reference, estimated continuation is weak across all three dimensions. Familiar production features can recur, even when departures from recent practice are not consistently carried forward.

Now pull back to the group. Think of each point as a creator's typical work in a half-year. Nearby points mean more similar works. As creators change, the distances between them can grow or shrink.

The differences do not simply keep growing. Across twelve comparisons, five widen and seven narrow. Audiovisual differences contract in early twenty twenty-two, while conceptual differences expand in early twenty twenty-five. The pattern is opening and closing, rather than a steady march outward.

Where creators start matters. If two creators move toward one another, their work becomes more similar. The same-sized steps in opposite directions make them more different. Collective change depends on how individual movements fit together.

Earlier shifts often remain visible. In nine of eleven comparisons, the next period stays on the same side of the original level. A group can partly move back and still remain more spread out, or more closely gathered, than before.

But being different from other creators today does not mean doing something the group has never done. We compare each eligible work with earlier work by its creator, then include the earlier work of peers. The question is whether a close precedent already exists.

Sixty-two works depart from personal history. Half of them resemble something a peer produced earlier. Like building a new shape at your own table, then discovering a similar shape on the shared shelf.

Only thirty-one works lack a close precedent in the observed group history: three point three percent of the nine hundred and forty-six works eligible for this comparison. All involve large personal changes. But most large changes still resemble earlier group work.

Looking at each dimension separately, departures from group history appear more often in conceptual content and audiovisual production than in embodied enactment. These counts can overlap: one video may depart in more than one dimension.

What happens next? All twenty-two candidates with available follow-ups show at least one return toward earlier group work. Nine have no observed follow-up. So an unusual work does not necessarily begin a lasting move beyond the earlier repertoire.

Return to the building blocks. Creators keep trying new arrangements, and those movements reshape how different their works are. Much of this experimentation happens among possibilities already present on the shared shelf. Distributed creative search can reorganize collective differences without continually expanding the historical range.

Footage credits

The following clips illustrate the research; they are not evidence of identified change events.