聚焦教育智能体的三阶进化,包括工具嵌入、主体共生和基础设施化阶段,基于此构建教育智能体进化的理论框架并深入探讨其范式跃迁的内涵及动力机制。范式跃迁的动力源于技术迭代、教育需求升级与制度创新的协同作用,不同范式间的过渡矛盾是技术逻辑与教育规律博弈的显性化。最终,教育智能体的终极形态将实现教育系统从技术赋能向智能内生的生态革命,推动教育公平从资源均衡到能力平权的质变,其发展终局将模糊技术与文明的边界,成为人类认知进化的加速器。
Abstract
This paper focuses on the three-order evolution of educational agents, including tool embedding, agent symbiosis, and infrastructure, and constructs a theoretical framework for the evolution of educational agents, and deeply explores the connotation and dynamic mechanism of its paradigm transition. The driving force of paradigm transition comes from the synergistic effect of technological iteration, educational demand upgrading and institutional innovation, and the transitional contradiction between different paradigms is the explicit game between technical logic and educational law. In the end, the ultimate form of educational agents will realize the ecological revolution of the education system from technology empowerment to intelligent endogenous, promote the qualitative change of educational equity from resource balance to ability equality, and its development will eventually blur the boundary between technology and civilization, and become an accelerator for human cognitive evolution.
关键词
教育智能体 /
三阶进化 /
主体共生 /
基础设施化 /
范式跃迁 /
教育公平
Key words
educational agents /
third - order evolution /
subject symbiosis /
infrastructure - based paradigms /
paradigm shifts /
educational equity
{{custom_sec.title}}
{{custom_sec.title}}
{{custom_sec.content}}
参考文献
[1] 朱珂,张斌辉,宋晔.技术生态位视阈下“人工智能+教育”的融合逻辑与模型构建[J].电化教育研究,2023,44(1):13-19.
[2] 北京大学.北大通知:开放多项AI应用[EB/OL].(2025-3-13)[2025-6-2].https://mp.weixin.qq.com/s/sV5MegXMJhQP7QSndv1hqA.
[3] 黄昌勤,钟益华,王希哲,等.从单智能体到多智能体:大模型智能体支持下的激励型学习活动设计与实证研究[J].华东师范大学学报(教育科学版),2025,43(5):44-56.
[4] SONG T,LUO M,ZHANG X,等.A Multiagent-Driven Robotic AI Chemist Enabling Autonomous Chemical Research On Demand[J/OL].Journal of the American Chemical Society,2025,147(15):12534-12545.
[5] 王雪,孙明琳,杨洁,等.教育智能体如何提供更有效的支持?——基于EEG信号的脑机制与优化策略探究[J].电化教育研究,2025,46(2):49-56.
[6] 施聪莺,李晶.FaceReader 7.0对国内常见表情图片库识别的有效性研究[J].心理技术与应用,2018,6(2):100-108.
[7] NOLDUS L.FaceReader reference manual:Version 4[M].Wageningen,Netherlands:Noldus Information Technology,2011.
[8] BIJLSTRA G,DOTSCH R.FaceReader 4 emotion classification performance on images from the Radboud Faces Database[Z].Nijmegen,The Netherlands:Department of Social and Cultural Psychology, Radboud University Nijmegen,2011.
[9] 张涛,孙国玺,马远佳.基于IEET工程认证的嵌入式系统教学改革与探讨[C]//AEIC Academic Exchange Information Center(China).Proceedings of 4th International Conference on Modern Management,Education Technology and Social Science(MMETSS 2019)(Advances in Social Science,Education and Humanities Research,VOL.351).2019:838-842.
[10] Xu Hao Cheng Qingxuan Dong Jing Wu Dan Tian Li.Generation AI in Human-AI Interaction: Origins,Characteristics and Future Prospects[J].Journal of Information Resources Management,2025,15(1): 13-20.
[11] 刘革平,王星,高楠,等.从虚拟现实到元宇宙:在线教育的新方向[J].现代远程教育研究,2021,33(6):12-22.
[12] 上海市教育委员会.【第一线】点赞！教育部推荐宝山基础教育综合改革典型案例[EB/OL].(2022-12-9)[2025-6-2] https://mp.weixin.qq.com/s/G4m_El1-Ujezu3HW-6Y94w,2025.
[13] 吴砥,郭庆.智能技术赋能教学的伦理挑战:表征、溯因与纾解[J].开放教育研究,2024,30(4):20-27.
[14] 吴永和,姜元昊,陈圆圆,等.大语言模型支持的多智能体:技术路径、教育应用与未来展望[J].开放教育研究,2024,30(5):63-75.
