# What Should Universities Teach When AI Can Teach Too > As AI enters the classroom, a university's value lies less in delivering knowledge than in cultivating the abilities machines cannot replace. One scholar argues universities should move from knowledge factories to shapers of ability. - Source: Times of China (Times of China) - Canonical URL: https://timesofchina.cn.com/en/article/what-universities-teach-when-ai-teaches - Chinese edition: https://timesofchina.cn.com/article/what-universities-teach-when-ai-teaches - Author: 中国时报编辑部 - Section: Education - Published: 2026-07-27T09:00:00+08:00 - Updated: 2026-07-27T09:00:00+08:00 - Tags: 人工智能, 高等教育, 教学改革, AI幻觉, 能力培养 --- When artificial intelligence can generate explanations, essays and code in an instant, a question is landing on universities: if AI can teach too, what should a university still teach. In a commentary, Xia Wenbin, dean of the School of Marxism at Southern University of Science and Technology and a former president of the University of International Business and Economics, argues that universities need to rethink their purpose, shifting from being knowledge factories to shapers of ability. AI in the classroom is no longer hypothetical. Tsinghua University has introduced human-machine collaborative learning in chemical engineering, Northwest Agriculture and Forestry University uses AI-generated images to teach calculus, Huaqiao University draws on AI frameworks for digital video creation in a traditional Chinese style, and Xi'an Jiaotong-Liverpool University is exploring AI-empowered training of translators. Technology is reshaping how teaching and learning happen. ## Do not be led astray by AI hallucination Xia cautions, however, that AI can quickly produce seemingly perfect explanations while presenting false information with convincing confidence, the phenomenon known as AI hallucination. That means students need less to receive answers than to develop the ability to question and verify, judging whether what a machine offers is actually reliable. ## Four recommendations He offers four suggestions for how universities might adapt. First, reorganize curricula around real-world problems rather than simply attaching an AI-plus label, so that students in law, journalism, engineering and literature learn by solving concrete problems. Second, reform assessment, moving the focus from cramming-based exams toward the learning process, for instance by documenting how AI is used, keeping learning logs, and weighing participation in group discussion. Third, redefine the teacher's role, so that teachers become editors, coaches and ethical advisors who provide the differentiated guidance AI cannot replicate. Fourth, keep universities as open systems, connecting the classroom to real society through projects, community service, internships and international collaboration. ## The abilities machines cannot replace What universities should really cultivate, in his view, are the abilities machines struggle to replace: independent thinking and problem-solving, critical analysis of AI-generated content, cross-cultural communication, ethical judgment and a sense of responsibility, and an understanding of the nuances AI cannot capture. These are precisely what give people their core value in an age of intelligent machines. The discussion has a policy and global backdrop. In 2025 China explicitly called for integrating AI into teaching methods and curricula while emphasizing students' independent thinking and cooperation. The World Economic Forum's Future of Jobs Report 2025 also noted that technological change will reshape employment and skill structures by 2030. As AI grows better at teaching, the answer for universities may lie not in competing with machines over who remembers more, but in cultivating what machines cannot learn. Leave the carrying of knowledge to the tools, and keep judgment, creativity and responsibility with people. That may be exactly the part higher education needs to hold on to in the age of AI. --- Originally published by Times of China. Free to cite with attribution and a link to https://timesofchina.cn.com/en/article/what-universities-teach-when-ai-teaches.