创造力是推动科技创新和社会进步的核心驱动力。本研究首次对约1万名人类参与者与9种主流大语言模型在发散性创造力任务上进行了大规模系统性比较。研究发现：人类平均创造力水平略高于大语言模型，但顶尖人类表现大幅超越所有模型；提升模型创造性参数可增强表现，但超过阈值后输出失真；提示工程对模型表现改善有限甚至产生负面效果。研究为人机协作创新提供了科学依据，揭示了大语言模型可辅助基础创意任务，但尖端创造力仍需人类专家主导。相关成果已接收在《自然人类行为》（Nature Human Behaviour）。

Publication:

Nature Human Behaviour

https://doi.org/10.1038/s41562-025-02331-1

Author:

Dawei Wang

Faculty of Business and Economics, University of Hong Kong, Hong Kong, China. 

Northwestern Institute on Complex Systems, Northwestern University, Evanston, IL, USA. 

 e-mail: dwdw@hku.hk

Difang Huang

Academy of Mathematics and Systems Science, Chinese Academy of Sciences, Beijing, China. 

School of Economics and Management, University of Chinese Academy of Sciences, Beijing, China.

Haipeng Shen

Faculty of Business and Economics, University of Hong Kong, Hong Kong, China.

Brian Uzzi

Northwestern Institute on Complex Systems, Northwestern University, Evanston, IL, USA.

Department of Management and Organizations, Kellogg School of Management, Northwestern University, Evanston, IL, USA. 

Ryan Institute of Complexity, Kellogg School of Management, Northwestern University, Evanston, IL, USA.  

e-mail: uzzi@northwestern.edu