全球教育展望

以ChatGPT为代表的生成式人工智能已基本具备跨领域解决问题的文本生成能力，逐步接近通用人工智能。本研究对ChatGPT进行各项能力测评，发现其具有较强的逻辑分析能力与批判性思维水平，但在创造力方面，ChatGPT没有明显的创造力倾向偏好。在教学能力方面，ChatGPT已经能够通过中国高中语文教师资格证考试的笔试部分。可以预测，此类生成式人工智能具备四方面应用潜能，有望成为教师准备教育资源的助手、学生开展自主自学的助手、课堂增强学习互动的助手、课外作业自动批改的助手。展望未来，生成式人工智能对教育的影响包括：形成多元协同的“师—生—机”关系；推进“人机融合”的教师数字素养提升；重塑课程体系结构与学习科学研究；全面关注“智能鸿沟”以促进教育公平。

Generative artificial intelligence, exemplified by ChatGPT, has predominantly attained the capability to generate text for addressing problems across a wide array of domains, progressively approximating general artificial intelligence. In this study, multiple assessments of ChatGPT's abilities reveal that it has strong logical analysis and critical thinking competencies. Nevertheless, regarding creativity, ChatGPT does not demonstrate a clear preference for creativity. ChatGPT has successfully passed the written portion of the Chinese High School Language Teacher Qualification Examination pertaining to its pedagogical aptitude. This type of generative artificial intelligence is anticipated to have the potential for four crucial applications: facilitating teachers in the preparation of educational resources, supporting students in their pursuit of self-directed learning, enhancing classroom interaction, and automating the correction of assignments beyond the classroom. Future educational implications of generative artificial intelligence include the formation of a diverse and synergistic “teacher-student-machine” relationship, the advancement of teacher digital literacy through “human-machine integration, ” the reshaping of curriculum structures and learning science research, as well as a comprehensive focus on addressing the “intelligence gap” in order to promote educational equity.