摘要
生成式人工智能的出现改变了人工智能与教育的关系,使其从工具型应用跨越到互构式发展。然而,生成人工智能真正走向与教育的融合发展依然面临诸多挑战,集中表现在生成式模型固有的低认知性使其难以与教育垂直领域求真求实的底色相符。因此,从以大语料和高参数为特征的生成式模型向以认知推理学习为基础的世界模型跨越成为人工智能发展的必然方向,教育的基本理论将成为推动这次跨越的重要力量。基于该视角,借助以世界模型为基础逻辑构建的契机,研究从知识、实践、态度和认知四个维度充分探讨教育对人工智能的反哺价值。此外,本研究还从技术实现和理念转变两个角度提出了相应的挑战。在当前生成式模型井喷式发展的节点,研究提出人工智能模型与教育的互构作用,对探索未来人工智能教育生态新高地有较为重要理论价值。
The emergence of generative artificial intelligence has changed the relationship between artificial intelligence and education,transition from tool based applications to interdependent development.However,the true integration of generative artificial intelligence and education still faces many challenges,mainly manifested in the inherent low recognition of generative models,which makes it difficult to match the background of seeking truth and reality in the vertical field of education.Therefore,the leap from generative models characterized by large corpora and high parameters to world models based on cognitive reasoning learning has become an inevitable trend in the development of artificial intelligence,and the basic theories of education will become an important force driving this leap.Based on this perspective and the opportunity of constructing logic based on the world model,this study fully explores the value of education in nurturing artificial intelligence from four dimensions:knowledge,practice,attitude,and cognition.In addition,this study presents corresponding challenges from the perspectives of technological implementation and conceptual transformation.At the current stage of explosive development of generative models,this study proposes the interaction between artificial intelligence models and education,which has significant theoretical value in exploring new heights of artificial intelligence education ecology in the future.
作者
翟雪松
吴庭辉
袁婧
李艳
Zhai Xuesong;Wu Tinghui;Yuan Jing;Li Yan
出处
《远程教育杂志》
CSSCI
北大核心
2023年第6期34-41,共8页
Journal of Distance Education
基金
国家自然科学基金青年项目“融合视觉健康的在线学习资源自适应表征及关键技术”(项目编号:62177042)
2024浙江省自然科学基金项目“基于双维眼动融合分析的在线学习者情感计算研究”(项目编号:Y24F020039)、2023年度浙江省教育信息化综合评价项目(ZZCG2023P-GK-122)的研究成果。
关键词
生成式人工智能
世界模型
教育反哺
Generative Artificial Intelligence
World Model
Educational Back-feeding
作者简介
翟雪松,特聘研究员,博士生导师,浙江大学教育学院(浙江杭州310058)、杭州国际城市学研究中心浙江省城市治理研究中心兼职研究员(浙江杭州311121);吴庭辉,在读硕士研究生,浙江大学教育学院(浙江杭州310058);袁婧,在读博士研究生,浙江大学教育学院(浙江杭州310058)、安徽新华学院大学生素质教育研究中心(安徽合肥230088);通讯作者:李艳,博士生导师,教授,浙江大学教育学院(浙江杭州310058)。
