Recently, 174 Peking University students took part in a unique ‘special exam’—competing head-to-head with artificial intelligence (AI). Organized jointly by Peking University and several research institutions, the exam aimed to explore the differences and potential complementarity between human cognition and AI capabilities in complex tasks. The test covered multiple dimensions, including logical reasoning, language comprehension, and mathematical modeling. Students completed the questions under time constraints, and their answers were evaluated against those generated by leading large language models—such as GPT and ERNIE Bot—using identical grading criteria.Results showed that while AI excelled in information retrieval and standardized responses, human students demonstrated clear advantages in creative thinking, interdisciplinary integration, and deep contextual understanding. Some participants noted that competing against AI was not only an academic challenge but also prompted them to reflect on their learning approaches and future competitiveness. The experiment has sparked broad discussion in educational circles: as AI advances rapidly, how should higher education adapt its goals to strengthen uniquely human high-order cognitive skills?Peking University stated that such human-AI collaborative research will continue and may eventually be integrated into the curriculum, helping students define their roles and cultivate irreplaceable core competencies in an AI-augmented world.
近日，174名北京大学学生参与了一场别开生面的‘特殊考试’——与人工智能（AI）同场竞技。这场考试由北大联合多家科研机构发起，旨在探索人类在复杂认知任务中与AI的差异与互补性。考试内容涵盖逻辑推理、语言理解、数学建模等多个维度，学生需在限定时间内完成题目，其答案将与主流大模型（如GPT、文心一言等）在同一标准下进行评分对比。结果显示，尽管AI在信息检索和标准化答题方面表现优异，但在创造性思维、跨学科整合及语境深度理解上，人类学生仍具明显优势。部分学生表示，与AI‘对战’不仅是一次学术挑战，更促使他们反思学习方式与未来竞争力。此次实验也引发教育界广泛讨论：在AI快速发展的时代，高等教育应如何调整培养目标，以强化人类独有的高阶思维能力？北大方面表示，此类人机协同研究将持续推进，未来或纳入课程体系，帮助学生在与AI共处的新生态中找准定位，提升不可替代的核心素养。
原创文章，作者：admin，如若转载，请注明出处：https://avine.cn/5702.html