1. The Hand That Returns the Badge
Asan, autumn 2024.
The HR office of Hyundai Motor's Asan plant. 2 p.m. Lee Jung-hoon, 52, sets a plastic card on the desk.
An employee badge. Name, employee number, rank, printed on its face. "Lee Jung-hoon, Production Engineering Division, Department Head." Four words summarizing 28 years. When it was first issued in 1996, the card was new. Every time he passed through the factory gate with it around his neck, the barrier rose. The barrier rising was a signal of authorized entry — and a confirmation: You matter inside these walls.
The HR clerk takes the card and slides it into a filing binder. Other cards are already slotted there. Lee Jung-hoon's is no different from the rest. Same dimensions, same material, inserted the same way. Whether 28 years or three, the card is the same size.
In Chapter 2 we wrote: "Someone is returning their employee badge." That someone is Lee Jung-hoon.
The badge was not simply an access pass. To be precise, it was also an access pass. Holding the card to the gate made the barrier rise, and only inside that gate did Lee Jung-hoon's 28 years hold meaning. If the gate won't open, he cannot hear the sounds; if he cannot hear the sounds, he cannot read the alignment of the dies; if he cannot read that, there is no reason for Lee Jung-hoon to be Lee Jung-hoon.
The employee badge was the material anchor of identity.
Whether Lee Jung-hoon's hand trembled as he handed the card over, we cannot know. In the prologue, he sat in a plastic chair at Incheon Airport, composed. The badge return had happened months earlier. In that moment of return, he says there was neither anger nor tears.
In the prologue, Lee Jung-hoon boarded a flight from Incheon to Hai Phong. To understand the weight of that choice, we must first see what he laid down. A single card he wore around his neck every day. What that card contained was not a name and an employee number — it was the entirety of one person's social existence.
The list of what a large-company employee badge means in Korea is long. Proof of employment for apartment lease contracts. A credential in the marriage market. A proxy for creditworthiness in loan reviews. Evidence of success reportable to parents and relatives. An inner confirmation that one is on "the normal track."
To return the badge is to return all of this at once. It is not an access pass being surrendered. It is an identity.
In Chapter 1, the Roman soldier lost his land and became a proletarius. In Chapter 2, the age when the Florentine artisan carved his name gave way to the factory's number. From land to hand, from hand to business card — the ground of identity has been migrating.
Lee Jung-hoon's return of his badge is the most recent scene in that migration.
And yet — what comes after returning the badge? When the soldier lost his land, Rome gave him bread. When the artisan lost the guild, the factory gave him wages. When Lee Jung-hoon loses his business card, who gives him what?
2. The Birth of the Career — A Construct Invented by the Twentieth Century
The word "career" originally meant a racetrack on which carriages ran. Derived from the Latin carrus, it became the French carrière: a straight-line course, a path with a fixed destination.
The word came to mean "a person's occupational path" only from the mid-nineteenth century onward. But the meaning we use today — a long-term occupational trajectory climbing a hierarchy within a single organization — is a concept that formed in the mid-twentieth century. It is not something natural. It was made.
The compound phrases "career path," "career development," and "career ladder" proliferated rapidly in English during the 1950s and 1960s — precisely the period when large-corporation bureaucracy reached its peak. This is not coincidental. The concept of the career was a product of the institution called the large corporation.
The material foundation for that institution was laid by Fordism.
Henry Ford's introduction of the moving assembly line at the Highland Park plant in 1913 was a revolution in production. But the essence of Fordism was not efficient production. Ford's 1914 declaration of the "five-dollar day" — twice the industry average at the time — was a strategy for turning workers into consumers of their own products.
The longer you work, the more you earn; the higher you climb. Fordism was the first to institutionalize this formula. The material foundation of the career construct was built here.
In his 1934 Prison Notebooks, Antonio Gramsci analyzed Fordism and observed that it was not merely a production method but a project for creating a new type of human being. The virtuous cycle of high wages, high productivity, and high consumption was a social control system that made workers into predictable beings. The metaphor of the career ladder arose naturally from within this system.
Frederick Winslow Taylor's Principles of Scientific Management (1911) completed this structure. Subdivide tasks; separate "those who plan" from "those who execute." This separation produced an explosive expansion of white-collar occupations. Those who plan, coordinate, supervise, report, and record — a new middle stratum. These are precisely "the people with careers."
An implicit premise formed: blue-collar workers earn wages but do not build careers.
The white-collar share of the American workforce rose from roughly 17 percent in 1900 to roughly 37 percent in 1950 and roughly 48 percent in 1970. The spread of the word "career" tracks this figure exactly.
In The Organization Man (1956), William H. Whyte dissected this new human type. His core thesis: "For the Organization Man, belonging is identity, belonging is security, belonging is meaning." Someone from IBM, someone from GE — a world was built in which affiliation mattered more than name. The equation business card = identity was formalized here.
Whyte wrote this as critique. It was a warning. But in Korea and Japan, this model was absorbed not as critique but as an object of aspiration.
The warning became the goal.
The American companies Whyte studied in the 1950s — GM, GE, IBM, DuPont — provided tens of thousands of employees with something close to lifetime employment stability. Suburban housing, company cars, health insurance. The corporation supplied the entire material foundation of life and received loyalty in return. The career model those companies created became the template for large corporations worldwide.
3. The Company Man and the Business Card — The Age When Loyalty Replaced Identity
In his 1958 study The Japanese Factory, James Abegglen defined the core of the Japanese employment relationship as "total commitment." In the West, employment is an exchange of labor for wages. In Japan, employment is the dedication of the whole person. The worker is not selling skill — he is affiliating his self to the organization.
"The worker does not see his relationship with the factory as a contract in which labor is provided in exchange for wages. Rather, the relationship is one of total commitment, a structure analogous to a family relationship." That is Abegglen's formulation.
In a family relationship, members do not leave the family when economic benefit decreases. If the company is a family, leaving the company is a betrayal of the family. To have one's identity annexed by the organization means that leaving the organization means losing the identity.
In his 1973 comparative study British Factory, Japanese Factory, Ronald Dore captured this difference more sharply. "The 'uchi' in the Japanese phrase uchi no kaisha means 'my household.'" In English, "my company" is a possessive. In Japanese, it is the language of family.
The three pillars of the lifetime employment system — lifetime employment, seniority wages, and enterprise unions — reinforced one another. Lifetime employment gave seniority its meaning; enterprise unions meant that strikes risked the company's destruction, so they were suppressed. The whole system rewarded the behavior of not leaving.
The Toyota Production System (TPS) added a new layer to this promise. Fordism's lifetime employment was a promise of "you keep this position." Toyotism's lifetime employment was a mutual promise: "the company invests in your knowledge, so you invest in the company." Through long-term employment, the worker's tacit knowledge accumulated, and that tacit knowledge became the source of kaizen — continuous improvement.
Lee Jung-hoon is legible within this model. He was not simply someone who stayed a long time. He was a repository of tacit knowledge built through sound and touch.
The term kaisha ningen — company man — was used in Japan during the 1970s and 1980s, sometimes self-deprecatingly, sometimes with pride. A person who commits his whole self to the company. The identity of the company man is structured in three layers.
Layer one, professional role: "I am the section chief of Sales Team 3 at XX Corporation."
Layer two, collective affiliation: "I am an XX person."
Layer three, personal identity: "What kind of person am I?"
Layers one and two overwhelm layer three. Mandatory retirement is an event that removes layers one and two simultaneously. It is not the husband who comes home — it is a being whose identity has been dismantled who comes home.
Korea absorbed this model in compressed form during the 1970s and 1980s. The expression uri hoesa — "our company" — carries in Korean the same emotional weight as uchi no kaisha. But there is a decisive difference. The Japanese company man was a cultural product built over decades. Korea's corporate-identity culture was compressed into a single generation. As Koo Hae-geun analyzed, the transition from agrarian to industrial society within a single generation dissolved the traditional resources of identity — land, clan, local community — and corporate affiliation filled the void.
What was assembled more abruptly is more fragile. The disruption that AI will deal will also be more abrupt.
Ezra Vogel, in Japan's New Middle Class (1963), analyzed how the identity of the Japanese salaryman derived not from individual interiority but from the group to which he belonged. The answer to "Who am I?" is replaced by "Where do I belong?" This structure operates identically in Korea. A society that opens every first meeting with "What do you do?" In Chapter 2 we examined Weber's Beruf — the equation of occupation with calling. Korea's business-card culture is the East Asian version of that equation.
In East Asian Confucian societies, the exchange of business cards is a highly ritualized act. The Japanese practice of meishi koukan (名刺交換) is its prototype. Cards are offered with both hands; upon receiving one, you bow; to pocket it immediately is a breach of etiquette. During meetings, the cards are arranged on the table in order of the counterparty's rank. A single sheet of paper carries the entirety of a person's social coordinates.
Korean business-card culture shares these East Asian roots but has its own particular intensity. The business card communicates three things simultaneously. First, company and rank. Second, position within the social hierarchy. Third, a proxy signal for trustworthiness. The credit of a Samsung section chief is borrowed from the credit of the organization called Samsung. The business card is the document that formalizes this borrowing arrangement. If in Japan a business card is a medium of deference, in Korea it is an instrument for the immediate determination of rank — without a card, it is not even settled who should speak first.
Sawon — juim — daeri — gwajang — chajang — bujang — isa. What this hierarchy determines is not just the scope of duties. Who greets first, who pays for the meal, who pours the drinks first. A society in which the formalities of social interaction cannot be settled without a title.
In Chapter 2 we watched the Florentine artisan carve his name into the back of the cabinet. That act of carving was a declaration: I am the one who made this. The two characters gwajang — "section chief" — on a business card are the bureaucratic version of that declaration. The difference is that the artisan's name was carved into his own work; the section chief's title was conferred by the organization.
Not something owned but something borrowed. What is borrowed must be returned.
4. Jeong Min-ho's Day — A Section Chief With a Card but an Empty Role
Nowon-gu, Seoul, spring 2025.
Jeong Min-ho, 45, arrives at work. He is a section chief on the planning team of a mid-sized auto-parts company. A company with revenues in the ₩300 billion range, 500 employees. His business card reads "Planning Team, Section Chief." He graduated from Chung-Ang University with a degree in business administration and joined this company in 2004. Twenty-one years now.
Trace Jeong Min-ho's day.
9 a.m. He sits down. Opens his computer. MS Copilot has sorted his inbox. The morning meeting agenda has been auto-generated by AI. A draft of the weekly report, completed in seven minutes, is waiting on his screen.
Three years ago, drafting the weekly report took a full day. Gathering data, building Excel pivot tables, constructing PowerPoint slides, polishing sentences. He believed that was his core competency. He had believed it for 20 years.
Now AI does it in seven minutes.
Jeong Min-ho's share is 20 minutes of review. Twenty minutes to confirm that there is no meaningful difference between what AI produced and what he would have produced. Sometimes he deletes AI's completed report and types it out by hand. The result is substantively identical. But the act of typing it himself maintains the sensation that "I am doing this work." Whether this is resistance, ritual, or self-deception, Jeong Min-ho himself does not know.
In meetings, his contributions overlap largely with the analysis AI has already offered. Three years ago, the industry context that only Jeong Min-ho knew gave his words weight. Now AI presents much of that context as data.
3 p.m. He does not know what to do with the time that remains. The work has not disappeared. His card says "section chief," he comes to the office, attends meetings, signs documents. But the substance of what the title "section chief" once performed is draining away. The title remains; the weight of the title has changed.
A 2024 survey by Microsoft Korea found that in companies that had adopted Copilot, report-drafting time fell by an average of 64 percent. Meeting preparation time fell 51 percent. Email processing time fell 45 percent. Yet 47 percent of employees could not say how they were using the time saved.
Jeong Min-ho is within that 47 percent.
He has not been dismissed. He is not a restructuring target. He tells himself: I haven't been cut yet, so I'm all right. But it is not all right — it is emptying.
If Lee Jung-hoon was pushed out, Jeong Min-ho is being hollowed out. The job exists, but the core tasks that constituted the job have been absorbed by AI. McKinsey called this "hollowing" — distinct from replacement. Replacement is when the job itself disappears; hollowing is when the job persists but its content is emptied.
Jeong Min-ho's annual salary is ₩68 million (roughly $49,000). His wife is an elementary school teacher. A leasehold apartment in Nowon-gu, Seoul, at ₩400 million (roughly $290,000), with a ₩200 million leasehold loan. Private tutoring costs for a son (14) and daughter (11): ₩1.3 million per month. Economically precarious he is not, but there is no slack. Within this economic structure, "transitional investment" — developing new competencies, preparing for a career shift — lacks both the financial and psychological resources to begin.
According to the Korea Labor Institute (2024), 61.3 percent of employed workers in their 40s and 50s said they felt their current job was insecure. Among those same respondents, only 14.2 percent said they were "actively preparing for a job change or career shift."
Feeling the anxiety but not acting on it. Jeong Min-ho is inside that condition.
In Book 5, Chapter 14, he recalls the day Lee Jung-hoon returned his employee badge. What was hardest for Lee was not the feeling of incompetence but the feeling of uselessness. Jeong Min-ho has not returned his badge. But the feeling of uselessness is growing at the same pace.
5. The Illusion of Prompt Engineering — Between 20 Years of Experience and Three Hours of Training
Geumcheon-gu, Seoul. A vocational training center. A Saturday morning in April 2025.
In a classroom offering a course titled "Digital Transformation in the AI Era," 24 participants in their 40s and 50s are seated. Jeong Min-ho is one of them. It is an external training program the company recommended. A weekend course, eight weeks. The instructor teaches prompt-writing technique. "The skill of giving AI clear instructions is the essential competency of the future."
Jeong Min-ho spent 20 years building Excel pivot tables. The person beside him is 52, a former regional branch manager at an insurance company. He whispers: "I heard that in six months AI will write its own prompts."
Targeting prompt engineering as a destination for career transition is structurally flawed.
First, the technology has a short lifespan. Based on LinkedIn data, job postings for "prompt engineer" surged in early 2023 and then fell by more than 60 percent from their peak after mid-2024. In their place, job titles demanding broader competencies — "AI engineer," "ML engineer" — increased. The prompt engineer role is converging into the ML engineer role — not a transition path for non-technical workers but a specialized role within the technical workforce.
Second, the floor is rising. In announcing the release of GPT-4o, OpenAI stated explicitly: "Natural-language instructions have become more natural, so users no longer need to know special prompting techniques." Anthropic's documentation for the Claude model states: "As models advance, the need for complex prompting techniques decreases." The output quality achievable through prompt engineering today will be achievable in six months without any special technique. It is like running on a conveyor belt — you must keep running just to stay in place.
Third, it is extension, not transition. Prompt engineering is a method for doing existing work faster using AI tools, not a method for moving up to the design layer. Jeong Min-ho learning prompt engineering does not restore the report-drafting capacity that AI Copilot has replaced.
Ethan Mollick addressed this directly: "Learning prompt engineering is fine. But making 'I am a prompt engineer' your identity is dangerous — not because AI will replace it in six months, but because that identity conceals the deeper expertise you already possess."
The pattern analyzed in Book 1 repeats. When the power loom displaced the handloom weaver's skill, some weavers tried to become "weavers who were better at operating the power loom." But operating a power loom was not a skill requiring seven years of training — it was semi-skilled work that could be learned in a few weeks of instruction. The structure of prompt engineering is the same.
In a survey by the Korea Employment Information Service (2023), the top criterion for 40–50-year-old retraining participants in choosing a course was "I heard there is high employment demand" — 41.2 percent. Second was "government funding is available" — 28.7 percent. "Because I wanted to" was only 17.3 percent. Training is determined by market signals or funding availability, not by analysis of individual strengths or capabilities.
Ministry of Employment and Labor (2024) statistics confirm this mismatch. The occupational mismatch rate for 40–50-year-old vocational training participants — the proportion who find work in a field different from what they trained in — is 57.3 percent. More than half find work in a field other than the one they trained for.
Prescription without diagnosis. Sweden's TRR (Trygghetsrådet, the Job Security Council) provides white-collar workers with competency analysis, individualized transition plans, and multi-year support. Korea's Tomorrow Learning Card (내일배움카드, Korea's public reskilling subsidy) has individuals choose their own course and the state covers the cost. The former is prescription after diagnosis; the latter is prescription without diagnosis.
That Jeong Min-ho is sitting in a prompt engineering course is a rational response. The system offers no better path. The problem is not with Jeong Min-ho. The real question is not "What skills does he lack?" The real question is: "In what manner, at what layer, can the 20 years of industry-context knowledge he possesses be deployed?" Korea has no system that answers the second question.
6. Two Chairs — The Watch and the Empty Desk
Tokyo, 1993.
Section Chief Tanaka retires after 32 years of service. At the ceremony, he is presented with a silver watch. Colleagues applaud. The department head says: "Thank you for your long dedication." The company logo is engraved on the silver watch.
In Japan, the ceremony of presenting a silver watch was not a simple expression of gratitude. The watch was a message — "your time is no longer the company's" — and simultaneously a pronouncement: "the time structure that defined you has ended."
A rite of separation.
The next day, Tanaka had his suit on and nowhere to go. The business cards had vanished from his wallet. He had a name but no title. For the following two years, he changed the hour of his morning walk to avoid encountering neighbors. Layers one and two of the company man's identity — professional role and collective affiliation — had been removed simultaneously; layer three, personal identity, could not hold.
Seoul, 2025.
Jeong Min-ho has not retired. His card reads "section chief." He arrives at 9 a.m. every morning and leaves at 6 p.m. His salary is deposited into his account; health insurance premiums are deducted. The form is intact.
But after AI Copilot completes the report in seven minutes, sorts the data, summarizes the emails — what remains is review and signature. The card exists, but what the "section chief" on the card once performed is empty.
Tanaka emptied out after leaving the company. Jeong Min-ho is emptying out while still seated at the company.
There is a difference. Tanaka had a rite of separation. A silver watch, applause, words of thanks. It was half-measure, but it was a rite. In Arnold van Gennep's three stages of the rite of passage — separation, transition, incorporation — Tanaka had the separation. What he lacked was incorporation into a new identity.
Jeong Min-ho has not even the rite of separation. No notice of dismissal, no retirement ceremony, no silver watch. He is being emptied without a sound.
Hollowing without ritual. This is the new shape of displacement in the age of AI.
Not a dramatic moment of surrendering a badge — but the process of sitting down at a desk at 9 a.m. every day while it grows less and less clear why he is sitting there. A dismissal notice has a date. Hollowing has no date. That one cannot specify when the emptying began makes it more cruel.
In Chapter 1, the Roman proletarius lost his land and lost his identity. Fordism supplied the business card as the foundation of identity in place of that land. AI is emptying out what lies beneath that business card.
The proletarius was displaced in a single blow. Jeong Min-ho is being emptied without being displaced.
7. What It Means That 70 Percent Are Exposed
The McKinsey Global Institute (2023) analyzed that 60–70 percent of automatable work hours with generative AI are concentrated in knowledge work — white-collar tasks. Specific figures make this concrete. Potential for automation of document drafting and editing: over 80 percent. Data collection and analysis: 75–85 percent. Customer correspondence (standard tasks): over 70 percent. Decision-support reports: over 75 percent.
These numbers translate Jeong Min-ho's day. And it is not only Jeong Min-ho's day.
Microsoft's 2024 Work Trend Index surveyed 31,000 knowledge workers in 31 countries. 75 percent of respondents said they use AI at work; 46 percent said they "cannot do important work without AI." AI is shifting from an auxiliary tool to an essential environment.
In the GitHub Copilot experiment, the group of developers using Copilot completed coding tasks 55.8 percent faster. In a controlled experiment with 758 BCG consultants, the group using GPT-4 showed 25.1 percent improvement in output quality and completed tasks 25.1 percent faster.
A particularly important finding emerged from that experiment: "Lower performers benefited most." AI elevated the work of average-performing workers to a top-tier level.
Paradoxically, this explains Jeong Min-ho's crisis. AI-driven upward leveling of mid-tier work eliminates the basis for differentiation of the "mid-tier professional." The gap narrows between a report produced by a section chief with 20 years' experience and a report produced with AI by a new hire who had three hours of AI training. The experience premium disappears.
A follow-up analysis by McKinsey (2024) generalized this phenomenon. The more "AI proficiency" becomes universal, the more rapidly the wage premium for possessing it contracts. Like "computer literacy" in the 1990s, beyond a certain point it becomes a baseline expectation rather than a differentiating factor. As Jeong Min-ho's 20 years of Excel proficiency already showed.
According to a 2023 survey by the National Information Society Agency (NIA), the adoption rate of AI/automation tools among companies with 100 or more employees was 48.3 percent; among adopting companies, 62.1 percent applied AI first to "document drafting and reporting tasks." 41.7 percent of adopters said the relevant employees had experienced a "change in role" following adoption — but only 19.3 percent said they had "specifically defined what the role had changed to."
Roles are changing, but organizations do not know what they should be changing into.
By Statistics Korea (2024) figures, clerical workers in Korea number roughly 4.97 million and managers roughly 470,000 — together approximately 5.44 million, or 19.6 percent of all employed persons.
The estimate that 60–75 percent of their core routine tasks — drafting reports, data collection and analysis, email and post-meeting processing — can be handled by AI tools at the current level of technology is a conservative figure.
Across the desks where 5.44 million people are seated, the weight of the business card is changing.
Between Lee Jung-hoon's employee badge and Jeong Min-ho's business card, there is a surface-level difference. Lee Jung-hoon had the tacit knowledge of hand and sense taken by AI machine vision; Jeong Min-ho is having the cognitive labor of analysis, synthesis, and reporting taken by AI Copilot. The line between blue-collar and white-collar disappears before AI. In the OECD's (2023) analysis of the distribution of automation risk among Korean jobs, high-risk exposure to AI encompasses both clerical (managerial-administrative) and production-technical workers together. The factory floor and the office are in the same structural position.
8. The Threshold — Career as Hypothesis
In Chapter 1 we saw the world in which land was identity. In Chapter 2 we saw the world in which the hand was identity. In Chapter 3 we are watching the world in which the business card is identity.
The three constructs share a common feature: all were invented. The equation land = citizen was made by Rome. The equation hand = artisan was made by the guild. The equation business card = career was made by Fordism and the Organization Man. Not laws of nature but constructs of particular eras.
Constructs are assembled, they operate, they break.
In the broken place, a new construct is assembled. That new construct will break too. The Roman construct took centuries to break. The guild construct took decades to break. The career construct is breaking now.
This is the pattern that the three chapters of Part I reveal.
Lee Jung-hoon returned his badge. He is someone who has passed through the moment when a construct breaks. He listened for sounds, read temperatures, detected the minute vibrations of dies — and remained faithful to the system. That faithfulness could not protect him.
Jeong Min-ho has not returned his badge. But what the badge guaranteed — the promise that loyalty will be rewarded with identity — is being rendered undeliverable by AI. A card that exists, whose content is draining away. A new kind of displacement.
Toyota president Akio Toyoda and Keidanren chairman Hiroaki Nakanishi publicly stated in 2019 that "maintaining lifetime employment is becoming difficult" — a formal announcement of the dissolution of a social contract that had held for 70 years. The promise that loyalty earns identity had been withdrawn by the party that made the promise.
In Korea, this dissolution had already begun with the foreign exchange crisis of 1997. The concept of "the job for life" collapsed. The non-regular employment rate rose from roughly 15 percent before 1997 to 38.2 percent in 2024. But even after the institution collapsed, the sentiment remained. The survival logic of "I must hold on at this company" dominated. The institutional foundation of the career construct had crumbled, but the psychological structure of career as identity remained intact.
By OECD standards, Korea's social protection expenditure as a share of GDP stands at 14.8 percent — well below the OECD average of 21.1 percent. In a country where "a good job" is a substitute for the social safety net, losing a job is losing the safety net itself. This is why the weight of the business card is heavier here than elsewhere. In Korea, a business card is not a contact detail. It is a document of survival.
The concentration of educational investment adds to this weight. Korean parents invest nine to 11 years and ₩100 million to ₩300 million (roughly $72,000 to $216,000) in their children's university entrance preparation. The destination of that investment is "a good job." Because the investment is large, the job that results is not merely a source of income — it is the justification for the investment. The business card is the material evidence of that justification.
Part I ends here.
The ground of identity that has migrated from land to hand, from hand to business card, is now shaking again. When land broke, the hand replaced it; when the hand broke, the business card replaced it. The business card is breaking.
What comes next? This is the starting point for Part II.
In Part II we move to the sites of that breaking — from historical contemplation to present urgency, to the place that asks what AI is actually replacing. In a law office in Gangnam, Seoul, a paralegal sits before an approval button. The same position as Lee Jung-hoon's confirmation button. A different industry, the same structure.
Threshold Question: After the business card is gone, can you still say what kind of person you are?