[Im Ki-beom’s AI innovation story] AI education cannot be achieved through teacher passion alone

This summer, 10,268 elementary and middle school teachers across the country received training as ‘Leading Teachers in the Use of Artificial Intelligence (AI)’. It is a 35-hour course operated by the Ministry of Education in six regions from June 23 to August 7. It is structured as a 6-hour basic course, a 24-hour intensive course focusing on practice for each subject group, and a 5-hour sharing course where class cases are shared.

One current teacher wrote in a media contribution in August that three things were missing from this training. It is time to apply the proven learning effects, curriculum-aligned design, and what you have learned to actual classes.

In my previous column, concluding the university section, I stated that task redesign and course evaluation cannot be left solely to the dedication of individual professors, and this time I would like to address the changing roles of teachers and professors.

The author's perception of reality was that while the government promoted large-scale investments in semiconductors, physical AI, and AI data centers, it failed to establish a system to train teachers to teach these in vocational high schools. Now I am faced with the same question again.

“No matter how well you decide what to teach, education will not change if the people who teach it are not prepared.”

The author argued that elementary schools should first establish the basics that humans must have, middle schools should learn how to doubt and verify AI's answers, and high schools should expand collaboration with AI according to career and educational purposes. The university proposed allowing the use of AI in principle, but limiting it based on learning goals, and redesigning assignments and evaluations.

The people who will actually implement all of this in the classroom are teachers and professors. But can we ask them to suddenly become AI experts one day, create new evaluation methods, reorganize classes, and even check each student's AI use process? It's not realistic.

It is easy to conclude the article by saying, “Teachers and professors must also change to fit the AI ​​era.” Nothing changes with just those words.

◇ ‘AI’ added to a teacher’s already busy day

The existing role of teachers does not diminish with the arrival of the AI era. The work of teaching, observing, evaluating, and guiding life remains the same. In addition, learning new technologies, applying them to classes, and verifying the results produced by AI are being added.

It’s not that teachers are ignoring change. According to the results of the 2024 Teachers and Teaching and Learning International Survey (TALIS) conducted by the Organization for Economic Co-operation and Development (OECD) targeting 55 education systems, 42.7% of Korean middle school teachers had experience using AI in their classes. It is higher than the OECD average (36.3%).

In the same survey, Korean teachers' weekly working hours were 43.1 hours, which was longer than the OECD average (41.0 hours). 46.9% of teachers cited excessive administrative work as a cause of stress, and 56.9% of teachers cited responding to parents' complaints. The structure makes it difficult to set aside time to learn new teaching methods.

A similar voice was raised in the policy brief published by the Korea Educational Development Institute (KEDI) in June this year. As a result of interviews with superintendents and teachers in charge of AI at metropolitan and provincial offices of education, many pointed out that the level of AI use varies widely among teachers and that the training content is far from actual classes.

The situation is no different for university professors. If you have to create new assignments while doing research, teaching, student guidance, and administrative work, look into the students' AI use process, and even conduct oral evaluations if necessary, the burden will inevitably increase.

It is difficult to expect proper change if educational innovation in the AI ​​era is left to the passion and sacrifice of individual teachers and professors. In my view, the first thing you need is time and support. You need to give yourself time to learn new teaching methods and change your classes. Applying new evaluation methods requires having the tools and people to help. It is difficult to assume that a few hours of training is enough to prepare teachers for new tasks.

◇ Can I get a ‘Ready’ stamp with 35 hours of training?

The 'AI Competency System for Teachers' that UNESCO released along with the AI Competency System for students in September 2024 does not only require teachers to have AI knowledge or how to use tools. It presented 15 competencies across 5 areas, including human-centered perspective, AI ethics, AI basics and use, AI teaching methods, and professional development using AI, and developed them in three stages of 'acquisition - deepening - creation'. Teachers' AI capabilities are not viewed as skills that can be learned once, but as expertise that can be built throughout one's career.

The Ministry of Education also took a similar direction. In March of this year, we developed the first AI education teacher training course with the Korea Education and Research Information Service (KERIS) and divided it into three stages: basic, advanced, and professional. We announced that we will train about 3,000 people this year and more than 10,000 by 2029. The competency system for leading teacher training was also refined to reflect international trends.

The framework is being prepared. The remaining task is the conditions under which the teacher can actually take the steps. AI technology changes every few months. If a structure involves receiving training once during vacation and then being buried in classes and administrative work once the semester begins, the step-by-step competency system remains only in documents. It is difficult to last long by separating learning time from teachers' vacations and after work.

Devices that create room for learning within working hours, such as adjusting the number of class hours or reducing administrative tasks, must be implemented.

I do not have the expertise to answer how to fix the teacher qualification system. However, one principle seems clear.

“If the industry and knowledge continue to change, teachers’ expertise must also continue to expand and change.”

When a new field is created, there must be a path for teachers in existing majors to expand their expertise, and teaching, learning and evaluation using AI must be included as a basic part of the prospective teacher training process. These changes cannot be left to the personal choices of a few enthusiastic teachers.

◇ Universities should support education just as they support research.

University professors are experts in their fields. However, not many professors have been systematically trained on how to teach and evaluate students. In the AI ​​era, it is increasingly difficult to answer how to change assignments and evaluations with only major knowledge.

This does not mean that there is no support at all. The Higher Education Training Institute of the Korean Council for University Education opened 21 new training courses, including basic and advanced courses in generative AI teaching methods, in the second half of this year. However, there is a distance between attending a training once and redesigning each of my lecture assignments and observing the students' AI utilization process.

Universities support professors' research with research funds and research support personnel. Education also needs the same level of support. We need to create common AI usage principles and evaluation guidelines at the university level, and strengthen the teaching and learning support organization, educational design experts, teaching assistants, and evaluation support system so that professors can receive help when designing classes and evaluations tailored to the characteristics of their majors.

Before demanding change from teachers and professors, we must first create conditions for change.

The author is not an education expert, and is not in a position to provide specific answers on how to design teacher training, qualification systems, training, and evaluation systems. However, the reason I bring up this issue is because, as someone who studies AI, as a scholar who is still studying it, and as a parent raising children, I feel that the questions that the current changes pose to our education are too heavy.

Perhaps it was another heavy task handed down to experts in the education field. So I am sorry and cautious.

What is clear is that the homework should not be passed on to individual teachers and professors. Educational authorities, schools, universities, researchers, and field experts must work together to continuously develop the expertise of teachers and create conditions and systems to respond to change.

Concrete answers will come from field experience and intense discussions among education experts. There is only one thing I wish for.

It is a system in which the curriculum, evaluation, and the expertise of the teacher move together. We cannot handle the next change by relying on the individual efforts of a few teachers and professors every time a new technology comes out.

Even in the AI ​​era, the weight of the term “100-year-old education” cannot become lighter. However, shouldn't the next 100-year plan be to establish a system that is prepared so that education can change in time as the world changes? This is a time when we need a system that can adapt to changes on its own rather than a solid system that will last for a hundred years.

Lim Lim, artificial intelligence expert

▲ Chief AI Officer (CAIO) of Korea Special Information Agency ▲ Visiting Professor at Seoul Graduate School of Science (aSSIST) ▲ Executive Director and Academic Division Chairperson of the AI Management Association ▲ Former CIO of Compaq Co., Ltd. ▲ Former Director of Shinhan DS Digital Strategy Research Institute