news

last semester, tsinghua university president li luming communicated with students in class and found that most undergraduates have used artificial intelligence to complete their homework, and the results are better. he asked the students: in the era of artificial intelligence, where is the value of students' learning reflected? li luming also asked himself the same question: where do teachers who impart knowledge find their core position?

nowadays, the rapid iteration of artificial intelligence is enabling the high-quality development of higher education. so what is the boundary of integration? how to achieve the unity of "integration" and "combination"? at the end of august, a forum hosted by tsinghua university and co-organized by oxford university's global higher education center, peking university's faculty of education, and hong kong university's faculty of education attracted educators from many universities around the world.

re-examining the value of higher education in the era of artificial intelligence, exploring the reform of university course teaching and evaluation methods under artificial intelligence, and thinking about the transformation of talent training ideas under artificial intelligence... a series of core issues were exchanged, debated, and collided in tsinghua university. as stefania giannini, assistant director-general for education of unesco, emphasized in her video speech: while artificial intelligence expands the possibilities of higher education, it is also necessary to clarify the boundaries of artificial intelligence and see its far-reaching impact on education.

a gray area of ​​student collaboration with artificial intelligence

although artificial intelligence has long been integrated into students' classrooms, homework, and life, is it allowed to use artificial intelligence to complete academic work? will artificial intelligence homework be evaluated? this series of questions still have not been accurately answered. in the observation of luo jiahui, assistant professor at the education university of hong kong, "there is a sense of guilt behind the use of artificial intelligence."

specifically, luo jiahui saw clear annotations in student assignments, that is, students would note that this part used artificial intelligence tools, while another part emphasized the content of original work. some students also asked luo jiahui: "if i use too many artificial intelligence tools, will i be at a disadvantage compared to people who do original work?" "if i use artificial intelligence tools, will the professor doubt that my thinking lacks independence?"

luo jiahui keenly captured the common concerns of students. therefore, luo jiahui believes that the higher education system needs to build a future vision of students living and learning together with artificial intelligence. on the one hand, students should see the potential of human-machine collaboration, and on the other hand, the evaluation method of original work should be adjusted.

jean-gabriel ganassia, a computer scientist and philosopher at the sorbonne university in france, gave this thought: "as educators, our goal is to enable students to play a role in the future world." he believes that students should not only become consumers and manufacturers of artificial intelligence products, but also become citizens who can adapt to the future world.

jean-gabriel ganassia, who has been working in artificial intelligence for 45 years, said technology is changing too fast and educators need to help people play their roles in future careers.

"speaking of the biggest change, i think we need to cultivate two types of talents in the future, one is the talent who can raise questions, and the other is the talent who can solve problems." yang zonkai, president of wuhan university of technology, sees that the education track is changing, from imparting knowledge first to an education that connects openness, sharing, individuality, and value.

having served as president of three universities, yang zonkai has been constantly trying to deeply integrate information technology and education. in his view, "innovation in the way of educating people, the mode of running schools, the management system and the guarantee system are the difficulties that need to be overcome in the new track of education." however, to complete the transformation of the model, yang zonkai admitted: "this is a very difficult thing."

on campus, yang zonkai and his colleagues created new learning scenarios, allowing students to conduct in-depth learning in unmanned factories, unmanned docks, smart transportation, etc., introduced artificial intelligence supervision into the classroom, assisted teachers and students in completing more than 2,000 courses, and cultivated students' independent learning, management, and service capabilities.

professor mutlu kukurova, director of the learning analytics and artificial intelligence team at the university college london knowledge lab, has made a new change in assigning homework. he no longer focuses on the results of student papers, but on the production process of student papers. "artificial intelligence has significant potential in analyzing materials." mutlu kukurova will focus on students' potential to ask questions and solve problems.

"artificial intelligence tools themselves cannot change education. the transformation of the education system is a social-technological ecosystem issue," said mutlu kukurova.

what are the bottlenecks in integrating artificial intelligence into campus?

from the printing revolution to the internet technology revolution, and then to the advent of the artificial intelligence wave, technological changes have brought new forms, new methods, new styles to education, as well as new thinking and new challenges.

xue lan, a senior professor of liberal arts at tsinghua university and dean of tsinghua university schwarzman college, shared a story: a few years ago, a professor of chemical engineering asked students to send feedback to the "teaching assistant" via email about their learning problems, and the "teaching assistant" was very responsible and answered their questions in a timely manner. after the semester, the teaching assistant, who was rated "excellent" by the students, unveiled the mystery - it was actually an artificial intelligence teaching assistant.

xue lan said that after artificial intelligence enters campus, it is possible to promote the formation of large-scale personalized teaching models and make up for the "shortcomings" of the existing education model, such as building personalized learning centers and forming personal learning hubs to provide applications for students to understand knowledge.

this year, tsinghua university announced that it will build 100 pilot courses for ai-enabled teaching, and equip the freshmen of the class of 2024 with an "ai growth assistant" to provide 24-hour learning assistance, feedback, and evaluation; "brainiac buddy", a product of gpt-4-based ai interaction engineering, will become a teaching assistant for peking university students, helping them build a knowledge base and prepare for courses; southeast university's "university physics course intelligent ai teaching assistant system" has been put into use to help teachers adjust course strategies and help students build a knowledge system.

the changes don’t stop there. xue lan believes that the integration of artificial intelligence into campuses will bring about major changes in the education model and the way schools are organized. “artificial intelligence, as a personalized learning center, provides unlimited possibilities.”

wayne holmes, a professor at the institute of education at the university of london and a researcher at the editorial board of the scientific journal of the unesco international research center for artificial intelligence, has a different view on this: "artificial intelligence may design personalized paths, but that is not truly humane. real personalization is to help every student become better, but artificial intelligence cannot do that yet."

wayne holmes believes that artificial intelligence has made teachers and students lose control of things, and they can only complete tasks according to the design path of the artificial intelligence system. however, there are complex operations of the business sector behind the artificial intelligence system, involving complex factors of commercial interests. not only that, "we must consider students' social ability."

wayne holmes described a scenario where more than 30 students sat in a classroom with more than 30 computers and they did not communicate with each other. "students are indeed learning in an individualized way, but this should not be the direction we are pursuing."

many experts have also raised questions about the concerns brought about by artificial intelligence. yan guangcai, director and professor of the institute of higher education at east china normal university, raised a question: "when artificial intelligence is integrated into education, does it enable surface learning or deep learning for humans?"

from a historical perspective, human learning outcomes are based on the accumulation of experience and knowledge over a long period of time, while large models come from deep learning of databases. in the view of this educator, it is necessary to study whether artificial intelligence can help humans complete in-depth learning.

artificial intelligence tools have entered classrooms, libraries and other places, changing the learning model of education. however, embedding technology into education is only the first step. in this regard, many experts present said that the new track has just begun.

yang zongkai believes that campus financial management, network security, and data security will all face new reforms in the future. he said that the integration of technology will encounter bottlenecks to a certain extent and that systemic changes need to be completed. "where the boundaries are is still a process of continuous exploration."

a global education exploration with no standard answers

chang tongshan, senior researcher at the center for higher education research at the university of california, berkeley, and director of institutional research and academic planning at the office of the president of the university of california, has worked at the university of california for many years. faced with the sweeping wave of artificial intelligence technology, the university of california has established an artificial intelligence working group to discuss issues such as learning, campus governance, and risk control.

"they see the birth of generative ai tools as a once-in-a-century opportunity, specifically, an opportunity to improve the effectiveness of university governance." chang tongshan observed that american universities have held ai seminars on different topics, gathering experts to discuss and explore different paradigms for the application of ai in the field of education. in addition, most universities offer ai general courses to improve students' ability to use ai.

how to solve the problem of "integration" of teaching and artificial intelligence? the answer given by some american universities is: formulate clear rules.

taking the university of minnesota as an example, chang tongshan mentioned that the university divides the degree of embedding artificial intelligence into teaching courses into three levels, namely, allowed use, thoughtful use, and prohibited use. teachers need to clearly state in the syllabus whether students are allowed to use artificial intelligence tools, as well as the scope and requirements for the use of artificial intelligence, and guide students on how to use artificial intelligence.

there are clear rules not only in teaching, but also in scientific research activities. chang tongshan mentioned that some universities have clear boundaries for drafting papers, writing applications, and reviewing scientific research projects on artificial intelligence to help teachers and students regulate their behavior.

"top universities maintain a high level of sensitivity and exploratory spirit towards new technologies." chang tongshan believes that american universities are seizing new technologies and exploring new paths for university governance.

however, there is no "standard answer" for the integration of ai and education reform in the world. lei jing, vice dean and professor of the school of education at syracuse university, believes that american universities are not ready to deal with the new track of integrating ai into teaching.

according to a questionnaire survey of many universities in the united states, the proportion of universities that have formulated policies and plans for the use of artificial intelligence is still relatively low, and a considerable number of universities are still exploring planning ideas. in addition, nearly half of the university staff are worried that artificial intelligence will affect academic fairness. "but the consensus is that everyone feels that artificial intelligence is crucial to the teaching and management of universities, and action must be taken to cope with changes," said lei jing.

in china, the ministry of education has launched four actions to promote ai-enabled education, including the launch of the "ai learning" column on the national smart education public service platform, the intelligent upgrade of the national smart education public service platform to support personalized lifelong learning for all people, the implementation of the ai ​​large-scale model application demonstration action in the education system, and the establishment of an international exchange platform for digital education. the 20th national congress of the communist party of china included "education digitalization" in its report for the first time, proposing: "promote education digitalization and build a learning society and a learning power for lifelong learning for all people."

the future has arrived. from the debate on personalized education to the design of teaching tools and the transformation of teaching thinking, the boundary problem of "integration and integration" of artificial intelligence has been placed before the global education community. today, the consensus has been reached, but the problem has not been solved. the exploration of new ways to integrate artificial intelligence into campus is still continuing...

china youth daily and china youth network reporter yang jie source: china youth daily

source: china youth daily