As generative AI moves deeper into workplaces and classrooms, SK Group Chairman Choi Tae-won said the skills that matter most are shifting away from memorized knowledge and academic credentials. He argued that in the AI era, the core abilities people need are thinking, adaptability, empathy, and physical skills.
Choi said education systems can no longer stay centered on knowledge transfer, test scores, and degree-based certification. Instead, they need to train people to identify problems, solve them in real working and living environments, and create value. He also said companies should not adopt AI simply to reduce headcount. If AI raises productivity, businesses should redeploy the labor that is freed up into new tasks they previously could not handle, including product development, service upgrades, and new market creation.
Four capabilities Choi says matter in the AI era
The first is thinking. As AI becomes faster at searching, organizing, and producing answers, human value no longer rests on recalling information from memory. Choi said it rests on asking the right questions, judging the quality of answers, and understanding the context and limits behind the information. AI can produce large amounts of plausible-looking content, but it does not always know whether the target itself is correct, and it can rely on wrong information. Humans still have to set direction, verify output, and bear the consequences of decisions.
The second is adaptability. AI tools, workflows, and industry demand continue to change, and career paths built around a single body of expertise over several decades may gradually lose relevance. Workers need to keep learning new tools and find new ways to work when existing roles are redesigned. Choi framed adaptability as more than learning software. It also includes recovering from failure and adjusting to role changes and industry shifts.
The third is empathy. AI can simulate conversation, analyze emotion, and generate personalized responses, but understanding another person’s circumstances, needs, and value conflicts remains a human task. Choi said management, healthcare, education, services, negotiation, and product design all require an understanding of why different people make specific choices. As technical skills become easier to access, he said the gap between companies may depend more on who can genuinely understand customers and employees.
The fourth is physical skill. Choi used the term broadly, covering not only manual labor but also work that depends on sensory judgment, operational experience, hand-eye coordination, and decisions made on site. Even if AI can handle large volumes of digital tasks, many jobs in manufacturing, repair, care work, food service, construction, and field services still need to be performed by people or highly mature robots. Until robots are widely adopted, he said people who can turn abstract judgment into practical execution will retain clear value.
Degrees are no longer enough on their own
Choi cited SK hynix’s removal of university graduation requirements for some positions as an example of a broader change. He said the period in which a diploma could directly guarantee job capability is ending. In the past, companies often used educational background as a fast screening tool. Schools, majors, and grades were treated as signals for knowledge level, learning ability, and work discipline. But AI is changing how knowledge is accessed. A large share of information that once had to be acquired through classrooms, books, and formal training can now be organized and explained by AI in real time.
That does not mean higher education has lost its value, he said. It means academic credentials alone are no longer enough to show whether someone can solve problems in a modern environment. According to the article, companies may place more weight on what applicants have actually built or completed, how they dealt with failure, whether they can work with people from different backgrounds, and whether they can make reasonable judgments when information is incomplete. Portfolios, project experience, on-site capability, and a record of solving problems may become more direct proof of ability than a single degree.
Education faces pressure to move beyond standard answers
Choi’s view also challenges current education systems. Traditional education often measures learning through standardized materials, model answers, and test rankings. Those are the kinds of tasks AI handles best when rules are clear. If students can use AI to obtain solution steps, article summaries, and draft reports, the focus of education cannot stop at whether an answer was produced. It has to look at whether students understand the problem, can verify information, and can form judgments of their own.
Under that approach, future coursework may need more project-based practice, cross-disciplinary collaboration, public expression, and real-world problem handling. Teachers may shift from one-way delivery of knowledge to helping students build frameworks for thinking, assess information quality, and reflect on decisions. Choi said the harder task is not teaching students how to use one specific AI tool, but helping them preserve independent judgment and learning ability even as tools change quickly.
AI adoption should not be reduced to layoffs
On concerns that AI could replace jobs, Choi said the end goal of corporate AI adoption should not be limited to shrinking labor costs. AI can automate some work, including document classification, meeting summaries, data entry, initial customer service responses, and standardized analysis.
But if companies convert all the time and labor saved by AI directly into layoffs, short-term costs may fall while long-term growth does not necessarily follow. Choi said companies should redirect labor freed up by AI-driven efficiency toward work they previously lacked the time, budget, or capacity to do. That could include developing new products, researching new markets, improving customer experience, redesigning internal processes, and addressing problems that had been left unattended for long periods.