生成式人工智能对欧洲教育的影响及对中国的启示  

Generative AI Based on Large Language Models Disrupts European Education

作　　者：托雷·霍尔 李思琦 孙波[4] Tore HOEL;Siqi LI;Bo SUN(College of Education for the Future,Beijing Normal University,Zhuhai 519087,Guangdong;Media and Academic Research Unit,Oslo Metropolitan University,Oslo NO-0130,Norway)

机构地区：[1]北京师范大学未来教育学院 [2]奥斯陆城市大学,挪威奥斯陆NO-0130 [3]北京师范大学未来教育学院,广东珠海519087 [4]北京师范大学珠海校区,广东珠海519087

出　　处：《中国教育信息化》2023年第6期8-16,共9页Chinese Journal of ICT in Education

基　　金：2022年国家自然科学基金青年科学基金项目“区域经济社会文化画像下的教育资源精准分配研究--基于地理信息系统的大数据分析”(编号:62107004);2023年广东省教育科学“十四五”规划项目“核心素养导向的高等师范院校职前科学教师培养研究”(编号:2022GXJK420)。

摘　　要：近期,美国OpenAI公司推出的生成式人工智能工具(ChatGPT)在全球范围内引起了各行各业的兴趣,特别是关于大型语言模型的应用与教育的利害关系引发了格外关注。而随着生成式人工智能技术的推广,关于其影响的辩论在欧洲以及世界各地已持续了数月。在教育领域,有观点认为这种新技术的推广将引发颠覆性变革,也有观点认为技术创新可以影响教育,但却没有改变它最基本的模式和规则。因此,教育界亟需了解和掌握此次技术变革的总体趋势和发展方向,以应对未来教育的变革。通过总结欧洲及大多数西方国家在教育数字化转型中面临的挑战和应对之策,及其对生成式人工智能技术的理解,提出在教育领域引入大型语言模型的机会、风险和挑战。首先,大型语言模型提供了学习机会,可以帮助教师改善教学质量,而这些机会因学习者不同而存在差异;其次,大型语言模型有可能彻底改变教学方式,并辅助教学过程,特别是在个性化学习、教学设计、语言学习专业发展等领域;再次,需要从不同视角来探索大型语言模型在教育中的应用,如学习者视角、教师视角、行政及管理者视角、研究者视角、雇主视角等;最后,大型语言模型也为教育带来了风险和挑战,如版权问题、偏见和公平、学习者对该模型过分依赖、教师对该模型过分依赖、缺乏深度学习和专业知识、模型生成结果与学生作答难以区分等。The recent launch of generative AI technologies has prompted a worldwide interest in how large language models could enhance or disrupt education.With the promotion of generative artificial intelligence technology,the debate about its impact has been ongoing for several months in Europe and around the world.In the field of education,there is a view that the promotion of this new technology will trigger disruptive change,while another view regards that technological innovation can affect education,but it has not changed the most basic educational patterns and rules.Therefore,the education community urgently needs to understand and grasp the overall trend and development direction of this technological change in order to cope with future educational changes.By summarizing the challenges and countermeasures faced by Europe and most western countries in the Digital transformation,as well as their understanding of generative AI technology,suggestions are provided for solving the“pain points”in the field of education.Firstly,large-scale language models provide learning opportunities that can help teachers improve teaching quality,and these opportunities vary depending on learners;Secondly,large-scale language models may completely change teaching methods and assist the teaching process,especially in areas such as personalized learning,instructional design,and professional development of language learning;Thirdly,it is necessary to explore the application of large-scale language models in education from different perspectives,such as learner’s perspective,teacher’s perspective,administrative and managerial perspective,researcher’s perspective,employer’s perspective,etc;Finally,large-scale language models also bring risks and challenges to education,such as copyright issues,bias and fairness,learners’excessive reliance on models,teachers’excessive reliance on models,lack of deep learning and professional knowledge,and difficulty in distinguishing model generation results from student responses.

关 键 词：生成式人工智能 大型语言模型 ChatGPT 教育人工智能 教育评估 

分 类 号：G434[文化科学—教育学]