【作 者】：
吴 虑， 杨 磊
【关 键 词 】：
ChatGPT； 技术赋能； 学习发生； 促进学习
【栏 目】：
理论探讨
【中文摘要】：
ChatGPT凭借类人的认知、生成和创造能力引发了人工智能领域的思维革命，一种不同于以往的认知框架正在建立。对学习而言，ChatGPT将通过高密度信息的即时输出，强调聚焦问题的自组织学习，关注学习过程中的知识迁移和理解生成来实现对学习起点、学习形态和学习路径的重塑。同时，该过程中潜在的逻辑张力也给学习带来了不确定性，集中体现为知识高效生产与学习者思维替代、深度学习生成与学术伦理越界、海量数据训练与特定意识形态偏向三重矛盾。为更好地规避风险，加速释放ChatGPT在教育数字化转型中的技术红利，使ChatGPT真正赋能学习，需要从三个方面着力：首先，要将ChatGPT技术教育化，开发出具备“教育品性”的EduGPT并研制使用规范，以确保学习者与其理性健康互动；其次，要从为学习者搭建脚手架和促进知识迁移的角度优化应用设计，促使学习者由“利用ChatGPT学习”转向“在ChatGPT中学习”；最后，要从关注学生如何学习、学习如何有效发生的立场重设教育评价，从而驱动有意义的学习体验。
【英文摘要】：
ChatGPT has triggered a thinking revolution in the field of artificial intelligence by virtue of its human-like abilities to recognize, generate and create, and a different cognitive framework is being established. For learning, ChatGPT will reshape the learning starting point, learning form and learning path through the instant output of high-density information, the emphasis on problem-focused self-organized learning, and the attention to knowledge transfer and understanding generation during the learning process. Meanwhile, the potential logical tension in this process also brings uncertainty to learning, which is mainly reflected in the triple contradictions between efficient knowledge production and learner's thinking substitution, deep learning generation and academic ethics transgression, massive data training and specific ideological bias. In order to better avoid risks, to accelerate the release of ChatGPT's technical dividend in the digital transformation of education, and to make ChatGPT truly empower learning, this paper focuses on three aspects. Firstly, ChatGPT should be educated. EduGPT with "educational character" should be developed and the usage specification should be formulated to ensure the rational and healthy interaction between learners and EDUGPT. Secondly, the application design should be optimized from the perspective of building scaffolding for learners and facilitating knowledge transfer, so as to encourage learners to shift from "learning with ChatGPT" to "learning in ChatGPT". Finally, educational evaluation should be reset from the perspective of focusing on how students learn and how learning occurs effectively, so as to drive meaningful learning experiences.