首页|放管结合:高等教育应对大规模语言模型冲击的立场及其展开

放管结合:高等教育应对大规模语言模型冲击的立场及其展开

扫码查看
大规模预训练语言模型以学生主动使用为渠道入局高等教育,引发知识、诚信、价值观多重危机.其本身巨大的应用前景和海量用户的既定事实,表明封禁排异已不可行;现有产品尚处于快速迭代期,加之缺乏成熟的治理方案,因而鼓励推广亦为时过早.对大模型的规制应树立"放管结合"的立场,坚持有所为、有所不为.一方面,公权力机关宜克制规制冲动、适当放任观望,仅对部分确有必要的事项精确管治;另一方面,鼓励教师、学生、企业等私主体探索自治举措,引导大模型正向赋能高等教育.未来"师—生—机"三位一体的教育模式中,教师的中心地位非但不会削弱,反将得到增强.
The Combination of Laissez-Faire and Regulation on Higher Education Response to the Impact of Large-ScaleLanguage Models and Its Development
Large-scale pre-training language models invade higher education through the active use of students,leading to a triple crisis of knowledge,integrity and values.Its huge application prospect and the established fact of a large number of users indicate that the ban is not feasible.The existing prod-ucts are still in the rapid iteration period,coupled with the lack of mature governance schemes,so it is too early to encourage promotion.The regulation of the large model should establish the position ofcombining laissez-faire and regulation,insisting on doing something and not doing something.On the one hand,the public authority should restrain the impulse of regulation,properly indulge in wait-and-see,and precisely regulate only some necessary matters;On the other hand,private subjects such as teachers,students and enterprises are encouraged to explore autonomous measures and guide large models to positively empower higher education.In the future"teacher-student-machine"trinity educa-tion model,the central position of teachers will not be weakened,but will be enhanced.

large-scale pre-training language modelhigher educationChatGPTartificial intelligence

陈心哲

展开 >

安徽财经大学法学院 安徽 蚌埠 233030

大规模预训练语言模型 高等教育 ChatGPT 人工智能

安徽省高校科研重点项目

2023AH050214

2024

滁州学院学报
滁州学院

滁州学院学报

影响因子:0.235
ISSN:1673-1794
年,卷(期):2024.26(3)
  • 4