Education in the AI era: Teaching how to define problems

Home > Opinion > Columns

Education in the AI era: Teaching how to define problems

Published: 23 Dec. 2025, 00:01

Audio report: written by reporters, read by AI

Lee Kwang-hyung

The author is the president of KAIST.

“Knowledge is power.”

The phrase, attributed to English empiricist Francis Bacon (1561-1626), has long guided and shaped modern civilization. Yet in the age of artificial intelligence, that maxim is steadily losing its force. AI is diminishing the value of knowledge and memorization skills that humans have traditionally prized. Within the next decade, we are likely to enter an era of technological singularity, in which AI broadly surpasses human intelligence. For now, knowledge and memory still matter, but their importance will decline sharply over time.

Nobel laureate in Chemistry Demis Hassabis receives his award from Sweden's King Carl Gustaf at the Nobel Prize ceremony at the Konserthuset in Stockholm on Dec. 10, 2024. The Royal Swedish Academy of Sciences decided to award the 2024 Nobel Prize in Chemistry to David Baker for computational protein design and to Demis Hassabis and John M. Jumper for protein structure prediction. [REUTERS/YONHAP]

Professions that heavily rely on specialized expertise have been the first to feel the pressure. Patent attorneys, lawyers, doctors, tax accountants and translators are often cited as examples. The decrease in entry-level positions across these fields clearly signals that a shift is already underway.

AI is now moving beyond the realm of knowledge into areas once considered uniquely human. It writes literary works, composes music and creates visual art, mimicking human creativity through increasingly novel attempts.

An introduction webpage for "Panshi," meaning “bedrock,” a science-focused AI research platform announced last August by the Chinese Academy of Sciences and promoted in English as "ScienceOne" [PANSHI WEBSITE CAPTURE]

Scientific research is no exception. Last year’s Nobel Prize in Chemistry was awarded to researchers who developed AI capable of predicting protein structures. Tasks that once took humans months or years can now be completed in minutes. In August, the Chinese Academy of Sciences unveiled an AI research platform — known in Chinese as “Panshi,” meaning “bedrock,” and in English as “ScienceOne” — designed to support research across mathematics, physics, chemistry, biology, geoscience and astronomy. Once humans define a research question, the model can review literature and even advise on research design and hypothesis testing.

AI-driven robots that conduct biological and chemical experiments and process data are also being developed, greatly reducing the time needed for manual laboratory work. Labs that still rely entirely on human hands may eventually be seen as outdated and avoided.

Even U.S. President Donald Trump has made significant decisions regarding AI. Late last month, he launched a major initiative dubbed the “Genesis Mission,” likening it to a second Manhattan Project. The executive order directed the government to build the infrastructure necessary to integrate AI into scientific research.

In computer programming, AI’s progress has been equally striking. Although AI itself is a program, it now writes programs on its own. As a result, global tech companies such as Google have significantly reduced the number of coding positions. In September, Google’s AI model Gemini achieved gold medal-level results at an international coding competition. It solved 10 of 12 problems at the International Collegiate Programming Contest, ranking second overall among all human teams.

As AI's role expands, the role of humans must inevitably change. For the children being educated today, who will live through the 2040s and 2050s, human work will look very different. Broadly speaking, three core roles are likely to define humanity's future.

The first is defining problems. AI will increasingly excel at solving problems. Humans, by contrast, will identify inefficiencies and areas for improvement in everyday life, define them as problems and then have AI address them.

Some may argue that AI will eventually outperform humans even at identifying and defining problems. But humans ultimately use technology, and judgment remains a human prerogative. Humans best understand the shortcomings and inconveniences of current technologies. Dissatisfied with the status quo, people constantly seek something better. Identifying what needs improvement and defining the problem will become humanity’s most important task that shapes the trajectory of civilization itself.

The second role is building and using AI effectively. AI is ultimately a human creation. Decisions about how it is designed and deployed remain human choices, making this role all the more significant. Planning new services and generating social value through AI will remain human responsibilities.

AI may become more intelligent than its creators. To use and control such systems properly, humans must understand them deeply. Studying AI is essential to avoid being dominated by it. Those who understand AI best will shape the world, and as AI’s influence grows, so will the power of those who build it.

The third role is advancing a humanistic society. Modern civilization has been led by humans, but we must learn how to recognize and coexist with AI. Ignoring or rejecting it would undermine our own efficiency and risk future conflict.

Humanity’s task is to reframe today’s human-centered social order into one in which humans and AI coexist. Reflecting on human values, cooperating with others and building relationships between humans and AI are responsibilities that no one else can assume.

U.S. President Donald Trump , center, delivers remarks with U.S. Senator Ted Cruz, left, and U.S. Secretary of Commerce Howard Lutnick during a signing ceremony on artificial intelligence in the Oval Office at the White House in Washington on Dec. 11. [EPA/YONHAP]

How, then, should education change in the AI era? Three possible approaches stand out.

Firstly, all education should be AI-based. Classroom instruction and homework should use AI tools. Lectures focused on transmitting knowledge should be cut in half, with the remainder devoted to discovering, defining and solving problems. Education must instill confidence in students to pursue creativity, challenges and even failures.

Secondly, the ability to develop and apply AI must be cultivated systematically. Students should understand AI principles, algorithms and basic coding. A society that relies only on AI built by others will ultimately be governed by the values embedded within it.

Thirdly, education in the humanities, social sciences, culture and the arts must be strengthened. These fields nurture the foundations needed to sustain a human-centered society. Overreliance on AI risks shrinking critical thinking and reflection, and combined with reduced physical activity, it could even lead to human decline.

As AI takes over more tasks, humans will gain more leisure time. How we use that time will shape our quality of life. The ability to engage meaningfully in culture and the arts will become essential. In the future, the phrase “knowledge is power” may give way to a new one: “Using AI well is power.”

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.