ChatGPT has been released for two years now, and the media continues to pay attention to the increasing "AI" content in academic research. The latest report shows that many schools in China have begun to explore the application boundaries of artificial intelligence technology in student theses. Some universities require students to fill out an explanation form for the use of artificial intelligence, some universities use existing AIGC detection tools for testing, and some universities have issued special regulations for the use of AI in undergraduate graduation theses. At present, there is a lack of effective tools for AI rate detection in academic papers. Some tools that claim to be able to detect AI usage can be applied in reverse, just like traditional "plagiarism detection" tools can also teach authors to "remove duplicates". Therefore, universities have limited methods in this issue, and in fact, they only bundle the governance levers that can be used, such as warnings, self inspections, teacher identification, and rule guidance. The lack of restraint measures seems to exacerbate the anxiety of universities, especially liberal arts educators, towards originality and criticality. The upgrading of technology can disrupt the balance of power in the system, often catching people off guard and seemingly making problems more cutting-edge than ever before. In fact, sometimes it just puts old problems in a polarized environment. The issue of originality, especially in humanities papers, is a chronic problem that may not be objected to. The papers that scholars are concerned about today, which are highly dependent on AI generation, may have been highly dependent on "Control C+Control V" 20 years ago, and highly dependent on "weight reducing tools" 10 years ago, with no difference in nature. Originality consciousness is a precious self-awareness, even a kind of confidence, which needs to be built on a deep trust in the sanctity of the profession. The more trust is generated by the profession itself and the professional narrators, the more students will have original intentions. Of course, even so, not everyone has it. Another anxiety of educators is that AI generated papers lack humanistic spirit and critical thinking, but their speed and simplicity are strongly attracting students to adopt them. Undoubtedly so, but we need to further consider where this deadly attraction comes from? Today's academic evaluation standards are generally oriented towards high efficiency and high achievement standards, and all professionals are constantly searching for ways to highly compress time and bring the highest cost-effectiveness of input-output. Otherwise, what can be used to pass the countdown of "either promotion or departure", what can be used to reach the order volume of papers and projects, and what can be used to firmly grasp the age and apply for various "funds" and "talents"? Can we find the most efficient and productive method? The result is more punitive than rewarding. That is to say, finding it only means passing the level, while not finding it means being eliminated. The actual scope of college students' learning in universities far exceeds professional knowledge, which inevitably includes understanding and recognition of this set of "academic algorithms". If this learned 'academic algorithm' itself is anti humanistic, then this paradox cannot be solved in the thinking of younger generations. From this perspective, preventing AI from generating papers may just be a sub proposition of a big problem. From past experience, it can be seen that technological means and technological governance means will inevitably reach a balance within a certain period of time, so more precise AI detection tools will emerge sooner or later. What needs to be considered is whether the aforementioned issues, such as lack of originality, criticality, and humanism, will disappear with the upgrading of tools. (New Society)