I collected these notes during the waiting time between rounds of vibe coding today. They are thoughts that have crossed my mind about how the world might change in the age of agents, written down as they came, in no particular order of sequence or importance. This is a casual piece that might look very different if I wrote it again a week from now. After finishing the list, I asked four models—Claude Opus 4.6, Grok 4.20, Gemini 3.1 Pro, and GPT-5.4—to review it. The percentage at the end of each item is the average of their independent estimates of its "probability of happening within three years." Some of the numbers were quite different from what I had in mind.
There is no need to see these changes only as crises. Wherever a crack appears, a window of opportunity always opens on the other side.
1. Middle management roles disappear
As agents handle task allocation, progress tracking, and performance reports, senior managers can oversee more than five times as much. Beyond top management, many middle managers who used to move information up and down the hierarchy become unnecessary. 65%
2. Office buildings empty out
Vacancy rates rise sharply, and property values fall by half. Offices are redefined from places to work into places to meet and find inspiration, but almost no buildings are designed that way. 55%
3. The content industry expands
Production costs fall to a tenth, and supply grows dozens of times over. The cost of making things approaches zero, while the cost of being discovered approaches infinity. 90%
4. The billable-hour economy collapses
The hourly billing model breaks down. The entire B2B knowledge industry, including consulting, law, and accounting, has to rewrite its rates. People whose profession was selling their time now have to prove their judgment, not their hours. 80%
5. Call centers disappear
AI handles 90% of calls within three years. In Korea, 400,000 people need to be reassigned. Soon, speaking to a human at all comes at a premium. 80%
6. Traditional résumés lose their meaning
Practical assignments replace academic credentials, while culture interviews become even more important. The central hiring question flips from "What can you do?" to "What do you know how to delegate?" 55%
7. The neighborhood clinic becomes an app
Seventy percent of primary care consists of pattern-based diagnoses and prescriptions. The technology is already ready; only medical laws and licensing systems determine the timeline. In a field where regulation moves more slowly than technology, patients wait not for technology but for permission. 35%
8. Newsroom staffing is cut in half
Traditional news media become less profitable and cut more than half their reporters. Only small, elite teams in areas such as investigative reporting survive. Once the value of breaking news evaporates, all that remains is "the ability to judge what truly matters." 70%
9. Taste divides social classes
As execution becomes a level playing field, the ability to define problems and a sense of aesthetics determine income gaps. The democratization of execution creates an aristocracy of discernment. 85%
10. Human labor becomes a luxury
The Handmade by Human label commands a large premium. Handmade furniture, human chefs, and in-person counseling become high-end services. Inefficiency is human labor's last competitive advantage. 75%
11. One-person SaaS companies proliferate
Agents handle development, design, customer support, and marketing. One person runs a software business bringing in hundreds of millions of won a month. The ability to get by without a team becomes a more essential skill for founders than the ability to build one. 70%
12. Translation and interpreting jobs disappear
Multilingual agents reach the level of professional interpreters. Only literary translation and court and diplomatic interpreting remain. As language barriers disappear, the cultural barriers hidden behind them come into sharper view. 85%
13. The employment premium of a college degree is halved
As the ability to use agents overtakes degrees, more people without degrees are hired. What universities sell shifts from knowledge to networks and signaling, but whether four years of tuition is a fair price for that signal is another question. 45%
14. Personalized education shakes up private tutoring
Agent tutors analyze each student's weaknesses in real time. But what private tutoring centers really sell is childcare and the management of parental anxiety. An industry that sells reassurance more than grades is hard for agents to touch. 45%
15. Corporate legal teams shrink by half
Agents handle much of contract review, regulatory compliance, and risk analysis. Demand for junior lawyers falls. The ladder from junior to senior becomes a cliff. 60%
16. Insurance claims review is automated
Human involvement all but disappears from filing a claim to receiving payment. Insurance shifts at its core from an industry that examines the past to one that predicts the future. 75%
17. Agents handle half of public service requests
More than half of routine administrative work is automated. But cuts to the civil service face strong political resistance, making this the area with the widest gap between what is technically possible and what is politically allowed. 45%
18. The accounting and tax software market is reshaped
Agents handle everything from reading receipts to filing taxes. Seven out of ten small and medium-sized businesses complete their taxes without an accountant. The reason for an accountant's existence narrows from balancing the books to making judgments beyond the numbers. 85%
19. Real estate brokerage is restructured
Agents search listings, compare market prices, and review contracts. Beyond real estate, all the intermediary professions that built their business models on information asymmetry fall one after another. 55%
20. Personal agents become essential infrastructure
Most urban workers use two to five agents in their daily lives. Subscription fees become a new fixed expense. As with smartphones, not using them is no longer a choice but a way of falling behind. 80%
21. Portfolio agents replace wealth managers
Services that automatically rebalance assets by having agents analyze market data alongside spending patterns become commonplace. Human advisers serve the wealthy as concierges, but agents are enough for everyone else. Class divisions in finance become more blatant. 50%
22. AI mental health counseling takes over initial consultations
Agents handle initial consultations for mild depression and anxiety. Waiting times fall from days to seconds. As the boundary between opening up and receiving treatment blurs, it is the algorithm, not the patient, that makes the judgment. 35%
23. Last-mile logistics costs fall
Autonomous delivery combines with AI route optimization, and the human role in delivery work shrinks sharply. Logistics costs fall substantially, and predictive logistics redefine the very idea of waiting. 55%
24. Power grid AI cuts peak demand
AI predicts weather, industrial operating rates, and household patterns in real time to adjust power second by second. Optimizing energy storage systems alone has already been shown to reduce peaks. The point where the energy AI saves crosses the energy it consumes is coming soon. 70%
25. Retail inventories shrink dramatically
AI demand forecasting sharply reduces waste in the fashion industry. As technology introduced to reduce inventory eliminates overproduction itself, discounting no longer works as a language of marketing. 60%
26. Cyberattack detection moves to milliseconds
AI security agents detect and isolate anomalies within milliseconds. False-positive rates fall sharply. Attacks grow stronger alongside defenses, leaving security in an endless arms race of agent against agent. 75%
27. VC screening is automated
AI analyzes pitch decks, financials, and founder backgrounds to filter deals. The odds of finding deals outside Silicon Valley rise. The more screening is automated, the more contrarian investors emerge who believe there are gems among what AI has filtered out. 60%
28. Predictive maintenance sharply reduces unplanned downtime in manufacturing
AI predicts equipment failures in advance. Manufacturing utilization rates rise sharply. When equipment is replaced before it breaks rather than fixed afterward, the very concept of repair disappears. 65%
29. HR performance reviews become continuous feedback
AI continuously analyzes work output and collaboration patterns. Time spent on year-end reviews falls by more than 90%. The moment continuous evaluation improves fairness, it becomes indistinguishable from continuous surveillance. 45%
30. Regulatory forecasting agents reshape the lobbying market
Tools emerge that predict regulatory changes by synthesizing bills, public statements, and public opinion. Political variables and limits to prediction accuracy remain substantial, but as the distance between predicting regulation and making it narrows, democracy faces new questions. 40%
