生成式AI对高等教育的破坏性影响日益凸显，然而这一危机的根源远不止于技术本身。据研究数据，学生对AI工具的使用已相当广泛。2024年一项针对8000人的研究发现40%学生承认通过算法作弊1，2025年RMIT一项调查显示近80%学生使用AI技术1，美国UC伯克利2024年研究表明三分之二本科生使用AI1，英国2025年分析更是表明生成式AI在该国学生中"近乎普遍"1。
大学管理层对此现象的回应饱受争议。澳大利亚五所大学（阿德莱德、纽卡斯尔、迪肯、RMIT）鼓励员工使用AI工具加快批改1。西悉尼大学官员Cath Ellis撰写的文章因过度依赖AI被《时代周刊》和《悉尼晨锋报》删除1，美国达特茅斯学院教务长Santiago Schnell在《华盛顿邮报》发表的文章也被指依赖生成文本1。2024年学者Kylie Moore-Gilbert指控"澳大利亚所有大学"从事"大规模工业级欺诈"1。
然而，AI危机背后是数十年新自由主义改革造成的系统性崩坏。自1980年代约翰·道金斯改革以来，澳大利亚大学政府资金从80%下降到约40%1，相应地学费大幅上升——2027年澳大利亚本地学生在法律、会计等学科的学费将超过18000澳元/年1，三年学位超过54000澳元1。国际学生学费更为高昂，墨尔本大学国际学生文科学位可能花费18.7万澳元，理科约20万澳元1。
教职员工面临严峻困境。2024年澳大利亚大学劳动力中60%处于不稳定就业状态1，最近调查显示88%永久职位员工表示自2019年以来未经认可的工作量增加1。高等教育员工报告年均工作3256小时，一名教员称2025年比获得报酬的工作多做25%1。工作量激增体现在多个维度：53%员工报告课程总体规模增加，62%报告实验室扩展，47%报告工作坊人数增加1。员工信心崩塌，悉尼大学艺术和社会科学系内部调查中，零百分比员工同意"我信任悉尼大学为员工、学生和更广泛社区的最佳利益行动"的声明1。
与此同时，管理层薪酬与支出高企。2024年38所公立大学副校长平均薪酬为102万澳元1，最近调查发现300名大学管理人员收入超过州州长1。澳大利亚大学每年花费18亿澳元雇佣管理咨询公司1。根据NTEU估计，整个教育部门总欠薪达到4.5亿澳元1。
大学财务模式依赖国际学生收入，导致教育商业化。悉尼大学47%收入来自国际学生1，国际学生投资组织主任Ariya Masud称国际学生被视为"提取多十亿美元产业的ATM机"1。非退款签证费从2300澳元增至4600澳元1。2024年《卫报》报道称澳大利亚学者感到被"迫让数百名涉嫌抄袭和其他作弊的学生通过以维持大学收入流"1。
学生对教学质量的体验不佳。墨尔本大学本科生观看六年前录制的视频片段1，反映出教学资源的不足。早期9月悉尼大学员工因包括AI相关政策在内的诉求而罢工1。罢工后不到两周，管理层与OpenAI签署多年合作伙伴关系，旨在在"所有学习层面"推进AI代理使用1。
Generative artificial intelligence has exposed fundamental fractures within higher education systems that predate the technology itself. While AI misuse has become a visible crisis on campuses, the underlying dysfunction stems from decades of neoliberal restructuring that has systematically eroded the institutional foundations of universities.
The scale of AI adoption among students is substantial. A 2024 study of 8,000 respondents found that 40% of students admitted to cheating through algorithmic means 1, while a 2025 RMIT survey revealed nearly 80% of students use AI technology 1. In the United States, research from UC Berkeley in 2024 showed two-thirds of undergraduate students employ AI 1, and 2025 analysis in Britain indicated generative AI use is "nearly universal" among students 1. Simultaneously, five Australian universities—Adelaide, Newcastle, Deakin, RMIT, and one other—have encouraged staff to use AI tools to expedite marking 1. The paradox deepened when Dartmouth's provost Santiago Schnell published an article in the Washington Post warning against AI dependence, only to face accusations that the piece itself relied on generative text 1. Similarly, Western Sydney University official Cath Ellis saw her article removed by Time magazine and the Sydney Morning Herald for excessive AI reliance 1.
The institutional pressures driving this crisis trace directly to financial restructuring. Since the 1980s Dawkins reforms, Australian university government funding declined from 80% to approximately 40% of budgets 1. Currently, 60% of the university workforce operates under precarious employment arrangements 1, while 88% of permanent staff report unrecognized workload increases since 2019 1. Meanwhile, universities spend A$1.8 billion annually on management consulting firms 1, and vice-chancellors at 38 public universities earn an average of A$1.02 million 1. A recent investigation found 300 university administrators earn more than state governors 1. An internal survey at the University of Sydney's Faculty of Arts and Social Sciences found zero percent of employees agreed they trust the university to act in the best interests of staff, students, and the broader community 1.
The reliance on international student revenue intensifies these dynamics. Sydney University derives 47% of its income from international students 1, while Melbourne University charges international students up to A$187,000 for arts degrees and approximately A$200,000 for science degrees 1. The director of the International Student Investment Organization described international students as an "ATM machine to extract billions of dollars" 1, while visa fees increased from A$2,300 to A$4,600 1. Scholars report feeling compelled to pass hundreds of allegedly plagiarizing and cheating students to maintain university revenue flows 1. In 2024, American statistician Nicholas Polson published over 200 papers and 14 books within a year with AI assistance 1, while submissions to Organization Science increased 42% since ChatGPT's emergence, even as human submissions declined 1.
The timing of institutional response proves revealing. Sydney University staff struck in early September, with demands including AI-related policies 1. Less than two weeks later, management signed a multiyear partnership with OpenAI aimed at advancing AI agent use across "all learning levels" 1. This sequence suggests that universities, rather than confronting systemic pressures, are institutionalizing the very technology that both reflects and amplifies their existing crises.
评论
还没有评论，欢迎留下第一条。