We used to picture competence as a stack of credentials — how many certifications you held, your language-test scores, your specialized knowledge. Things you could measure and list on a résumé. But as AI takes over that territory, the value of the things that are hard to measure is climbing.

The Bank of Korea ran an analysis of how exposed different occupations are to AI replacement. Doctors and practitioners of Korean traditional medicine topped the list at 99%. Accountants came in at 81%; judges, prosecutors, and lawyers at 79%. Bodyguards and singers, on the other hand, scored 0%, and childcare workers 25%.

Doesn't that seem backwards? The jobs that take the longest to train for and pay the most are the first to go.

It's called Moravec's Paradox. The idea is that what's hard for humans — math, logic, legal interpretation — is easy for AI, while what's easy for humans — walking, manual dexterity, reading a facial expression — is hard for AI. The pecking order of professions that people believed in for decades is being flipped wholesale.

A ₩30M Desk Job vs. a ₩70M Trade

The job platform Catch surveyed Gen Z job-seekers. Asked to choose between a white-collar job paying 30 million won a year (about $22,000) and a blue-collar job paying 70 million (about $51,000), more of them picked blue-collar. The reasons were clear-cut: the higher pay (67%), the lower risk of being laid off (13%), and less of the overtime-and-promotion stress.

And sure enough, you can now spot people in their twenties and thirties on wallpapering, tiling, and plumbing sites without much trouble. Trades that were once avoided are being re-rated as "jobs that are safe from AI."

But that leaves one uncomfortable question. So what are the people working in offices supposed to do?

Skill and Knowledge Already Belong to AI

The Korea Development Institute (KDI) put out a report warning that if the current labor-market structure holds, up to 90% of human labor could be replaceable by 2030. In the US, the payments company Block cut a workforce of more than 10,000 employees to under 6,000 — better than 40% of its staff, gone.

There's a common thread. The competence of the people being replaced was concentrated in "skill" and "knowledge." Being good with Excel, memorizing legal statutes, knowing how to clean up a dataset. AI does all of it faster, more accurately, and more cheaply.

So what's left?

A survey of 1,532 people by Dooit Survey and Research&Lab turned up something interesting. The shared keywords for the "competencies AI can't easily replace" were on-the-ground judgment, real-time problem-solving, and emotional connection. None of that is skill, and none of it is knowledge.

It's attitude.

When Attitude Becomes the Competence

Making the call when the situation is uncertain. Finding the next move when a project has just collapsed. Building consensus in a meeting room where opinions are colliding. You can't learn this from a manual, and there's no data for AI to train on.

The OECD is saying the same thing: AI-driven change in the workplace isn't simple replacement but a reconfiguration of roles, and alongside technical ability, distinctly human capacities — creativity, critical thinking, emotional intelligence — will only grow more important.

In the end, the line a career has to defend is shifting from "what do I know" to "what kind of person am I."

There's no need to fear AI. But if you've assumed that the skills and knowledge you've banked so far are the sum of who you are, that assumption is worth a second look. Because what AI can't take, in the end, is the part that only comes from people.