Seeing Popularity Reduces Innovation

Gautheron et al. (2026) examined how popularity ratings affect cultural dynamics in an experiment where participants select images from evolving markets and produce their own modifications. When popularity ratings were visible, cultural diversity dropped and the pace of innovation slowed.

This was driven by changes to both selection and creation. On the selection side, participants were more inclined to build on popular images, reducing diversity. On the creation side, participants didn’t just make smaller modifications, but changed the kinds of modifications they made. Human raters classified edits into five categories (disruption, addition, pattern-growth, refinement, removal), and popularity feedback specifically reduced disruption while increasing pattern-growth (extending existing visual features rather than introduce something new).

However, while popularity ratings initially produced less aesthetic images, the disadvantage shrank over time. The intuition: when images are far from optimal, bold mutations help because most changes are improvements. But as quality rises, large changes increasingly destroy what works.

Experimental vindication of the avant-garde? The sorry state of contemporary mainstream cinema seems to provide an example of this dynamic.