ISSN 1671-3710
CN 11-4766/R
主办:中国科学院心理研究所
出版:科学出版社

• •    

人机协同创造力:认知增强还是认知依赖?

赵源, 卢凤, 沈汪兵   

  1. 山东政法学院警官学院, 250014
    泰州学院教育科学学院,
    河海大学公共管理学院,
  • 收稿日期:2026-08-27 修回日期:2026-10-18 接受日期:2026-11-12
  • 基金资助:
    2026年山东省社会科学规划年度项目(青年项目)(26DJYJ13)

Human–AI Collaborative Creativity: Cognitive Enhancement or Cognitive Dependence?

  1. , 250014,
    , ,
  • Received:2026-08-27 Revised:2026-10-18 Accepted:2026-11-12

摘要: 人机协同创造力是人类与生成式人工智能(generative artificial intelligence, GenAI)通过认知分工与动态交互共同产生新颖且有价值成果的创造形式。GenAI介入人类创造活动可能产生不同的创造结果,本文从创造认知功能在人与GenAI之间的分配视角出发,整合扩展认知、创造力双加工和创造力成分理论,提出人机协同创造力动态双路径模型:人类保持核心认知功能时,GenAI发挥认知扩展作用,促进创造增强;核心认知功能逐渐外部化时,则可能形成认知替代与创造依赖。知识表征、认知控制和元认知调节共同影响双路径的形成及动态转换。本研究为理解GenAI对创造力的双重影响及优化人机协同方式提供理论框架,未来需进一步检验相关认知机制与路径转换条件。

关键词: 人机协同, 创造力, 认知增强, 认知依赖, 动态双路径模型

Abstract: Human–AI collaborative creativity refers to a form of creativity in which humans and generative artificial intelligence (GenAI) jointly produce novel and valuable outcomes through cognitive division of labor and dynamic interaction. The involvement of GenAI in human creative activities may lead to different creative outcomes. From the perspective of how creative cognitive functions are distributed between humans and GenAI, this article integrates extended cognition theory, dual-process theories of creativity, and the componential theory of creativity to propose a dynamic dual-pathway model of human–AI collaborative creativity. When humans retain core cognitive functions, GenAI serves as a cognitive extension that enhances creativity; when these functions are progressively externalized to GenAI, cognitive substitution and creativity dependence may emerge. Knowledge representation, cognitive control, and metacognitive regulation jointly shape the formation and dynamic transitions of the two pathways. This study provides a theoretical framework for understanding the dual effects of GenAI on creativity and optimizing human–AI collaboration. Future research should further examine the underlying cognitive mechanisms and the conditions governing transitions between the two pathways.

Key words: human AI collaboration, creativity, cognitive enhancement, cognitive dependence, dynamic two pathway model