In the age of AI, teach how to assign work

Kim Byoung-pil

The author is a professor of technology management at KAIST. 

As artificial intelligence makes it possible to generate assignments, reports and presentations almost instantly, educators are facing a pressing question: What should schools teach now? Many teachers and professors share this concern. Students use AI to complete their work, while instructors rely on AI to evaluate it. Learners, teachers and evaluators alike are leaning on the same tools. As this cycle repeats, unease is growing. The process of learning continues, but little learning seems to remain. Assignments, presentations and evaluations risk becoming empty forms, with substance fading away.

Some argue that AI use should be restricted, especially for younger students who may become overly dependent and fail to learn how to think for themselves. But banning AI is unlikely to offer a real solution. Pushing AI out of the classroom does not mean students will live in a future without it. Instead, a more fundamental question must be asked: What should be the purpose of education in the age of AI?

Until now, schools have developed in a direction that emphasizes training people to perform tasks well. Modern public education expanded access to universal schooling and fostered habits such as arriving on time and completing assigned work. It also contributed to cultivating active participants in democracy while equipping individuals with the skills needed in an industrial society. But AI is rapidly encroaching on this territory. Following instructions and executing tasks are becoming areas where machines can outperform humans.

As a result, the purpose of education must also change. It is no longer enough to teach how to work. Education must now also teach how to assign work. One way to understand this shift is to recall the education of a crown prince during the Joseon Dynasty (1392-1910). A crown prince was surrounded by numerous officials who could carry out any task upon command. Yet education was still essential. Even if the prince did not need to perform tasks personally, he had to learn how to give instructions properly. He needed to judge what was important, decide who should handle each task and determine whether the reports he received were accurate.

For this reason, Joseon had institutions dedicated to educating crown princes. They spent their days studying Confucian classics and undergoing examinations. One key text was “Seonghakjipyo” (1575), written by the scholar Yi I for King Seonjo. In the book, Yi emphasized the importance of appointing capable people. He wrote that a ruler’s foremost duty is to recognize talent and assign responsibilities effectively. Managing AI today is not so different. Not all systems are trustworthy. Some produce plausible but incorrect answers, fabricate information or rush to conclusions. It is therefore essential to distinguish reliable tools from unreliable ones and assign tasks accordingly.

The next step is learning how to give instructions to AI effectively. Students must learn how to define goals clearly, write precise prompts, verify outputs and revise results on their own. Above all, they need to develop the ability to draw boundaries, deciding which tasks to delegate to AI and which to handle themselves.

If the content of learning changes, evaluation must also change. If grading is based only on the final product, it ends up evaluating AI rather than the student. Educators should instead look at what questions students asked, what judgments they made along the way and how they edited or discarded AI-generated responses. This approach can help cultivate students who are capable of designing tasks and critically assessing answers.

This issue is not limited to higher education. It applies equally to primary and secondary education. In the future, children will turn to AI whenever they encounter something they do not know. If assignments simply demand correct answers, little learning will remain. The process of reaching an answer is more important than the answer itself.

Being a good ruler is harder than being a good subordinate. The same is true in education today. Schools must go beyond teaching how to work independently. They must teach how to use AI without being controlled by it. Students need more than the ability to produce quick answers. They must learn to think with AI while making final judgments on their own.

That is why education continues to matter, even in the age of AI.

This article was originally written in Korean and translated by a bilingual reporter with the help of generative AI tools. It was then edited by a native English-speaking editor. All AI-assisted translations are reviewed and refined by our newsroom.