Human diversity fuels collective creativity that large language models cannot simulate or sustain

2026-07-29

AI Summary

This study investigates the impact of generative AI on creative diversity in human groups, with a particular focus on whether AI induces idea homogenization and diminishes the unique contributions of non-native speakers. Through a preregistered metaphor-generation experiment, the authors compare native and non-native writers under three conditions: no AI, AI-generated ideas, and AI-refined ideas, while also simulating human collectives using AI personas grounded in real demographic profiles. The findings provide the first empirical evidence that non-native speakers significantly enhance collective creative diversity—an advantage attenuated when using AI-generated content but preserved under AI refinement. All AI-simulated groups exhibited lower diversity than real human groups. Although AI improved individual-level evaluations, it generally undermined collective diversity, except when non-native speakers used their native language, yielding simultaneous gains in both individual quality and group diversity.

Abstract

Diverse human groups produce diverse ideas, the raw material of innovation. Generative AI challenges this engine twice over: everyday AI assistance may homogenize what diverse people create, and AI-simulated diversity may replace the people altogether. We tested both challenges in a preregistered creative metaphor experiment with native (L1) and non-native (L2) English writers, who wrote without AI, with AI-generated ideas (AI ideation), or with AI refining their own ideas (AI refinement). L2 writers contributed more collective diversity than L1 writers, with native-language ideation showing the most diverse pools. AI ideation compressed collective diversity for everyone and left the L2 advantage undetectable, whereas AI refinement preserved both. We then simulated the entire writer pool using personas built from participants' real backgrounds, three model families, native-language prompting, and elevated sampling temperatures. Every simulated pool fell below every human pool, and pushing models further induced diversity only through degenerate text. However, at the individual level, AI ideation raised writers' ratings, pitting private incentives against the collective good, except when L2 writers used their native language, which benefited both. Human diversity remains a valuable creative resource that current AI cannot simulate or sustain; the design of human-AI collaborative workflows determines whether it survives.

Problem

• human diversity
• collective creativity
• generative AI
• creative homogenization
• AI-simulated diversity

Innovation

• collective creativity
• human diversity
• AI ideation
• AI refinement
• language diversity