A client sits across a Tokyo meeting table and says, "The numbers look fine. I still don't want to sell." The minutes of that meeting can now be drafted by AI before anyone is back at their desk. The hesitation in the room cannot. The definition of human capital — what we at INA write as 人財 (jinzai), people treated as assets, not as "human resources" (人材) — is starting to change in a serious way. Headlines still say AI will take jobs. The real movement is narrower. AI first replaces routine admin and processing, not the person sitting in the chair. That is why the question companies now have to answer is not how many people they can remove. It is what they leave with people, and what they hand to a machine.
Is AI Taking "Jobs" — or First Taking Routine Admin Work?
A junior staff member used to spend Tuesday morning typing up yesterday's meeting. Now a recording and a prompt produce a usable draft before coffee. The job title did not vanish. The Tuesday morning did. As generative AI has spread, tasks such as writing minutes, summarizing documents, handling the first response to an inquiry, and sorting data have become much easier to automate. An ILO paper published in 2025 makes the same point: most jobs will not vanish. Parts of the work inside those jobs will be rearranged. The question that matters is not "will the job title disappear?" It is "will the content of the job change?"
Microsoft's Work Trend Index, published on 23 April 2025, pointed in the same direction. Redesign the role on the assumption that people and AI agents will work together. In other words, the point of adopting AI is not to cut people. It is to move processing work onto machines so that people can return to the work they should have been concentrating on.
This is a uniquely Japanese way of framing the problem, and it is easy to miss if you read the topic through a US or UK head office. In much of the English-speaking corporate world, an AI rollout is still sold as a headcount-efficiency story: fewer full-time roles after the software is live. Japan's owner-managed firms — the overwhelming majority of companies here — usually ask a different first question. Who still sits with the client when the spreadsheet is clean and the person across the table is not ready? If you are hiring in Japan, or buying or partnering with a Japanese operator, importing the Western "efficiency after automation" slide as your operating plan will select for the wrong people.
If management gets this wrong, the whole project collapses into a short efficiency story. What will you do with the hours that come back? Without an answer to that, AI adoption ends as a one-off lift in output. That is why I believe the AI era makes the idea that "people are the greatest asset" more important, not less.
Why Do People Still Remain? Five Forms of Value AI Cannot Fully Replace
A client repeats the same sentence twice, a little slower the second time. The transcript captures the words. It does not capture that they are about to walk away from the deal. AI is good at going fast, missing little, and finishing at a steady quality. If people try to compete with AI on that same ground, their worth tends to fall. So what remains as value only a person can provide? I believe five strengths now sit at the center of that worth.
The Ability to Read a Counterpart's Context in Conversation
The same words change meaning with the other person's position, feelings, and situation. What a customer is actually afraid of, the unease a junior staff member cannot yet put into words, the circumstance a counterparty is carrying behind the numbers — none of that appears unless you read the conversation as it unfolds. Dialogue is not an exchange of information. It is the act of building a working relationship. AI still cannot fill that depth on its own.
The Ability to Decide When There Is No Right Answer
AI is good at lining up options. The last step — "we go this way" — is still a person's. Information is incomplete. The interests of the people involved are still moving. You still have to go forward. That is when a decision that someone will own becomes necessary. The worth of management sits exactly here.
The Resolve to Take Responsibility, and the Trust That Follows
Work does not move on correctness alone. Who owns the outcome, and who explains it at the end, changes how safe an organization and a client feel. A company is stronger when someone will bring the inconvenient facts as well as the good news, and then take the judgment. AI can produce an answer. A human being still answers for the result.
The Flexibility to Relearn After Failure
In the AI era, the person who can relearn as conditions change is worth more than the person who already knows the "right" method. Materials from Japan's Ministry of Health, Labour and Welfare (厚生労働省) on reskilling make the same point: moving people toward DX work requires continuous learning. Without a culture that does not fear failure, that relearning does not continue. Updating how you work, again and again, is now a core quality of 人財.
Empathy and Meaning-Making That Move People
An organization does not move on orders alone. The person who can put into words why the work is being done, and whose life it is for, is the person who moves the people around them. Empathy is not softness. It is the ability to imagine every party involved, share the purpose, and create a sense that the work is worth doing. People who can do that become more valuable, not less, as AI spreads.
How Should We Evaluate 人財 in the AI Era? Moving Off Processing-Speed Metrics
Look at last quarter's personnel review in a typical Japanese office and you can still see the old scorecard. Speed. Accuracy. How many items someone cleared. Many firms have long rewarded exactly that, and those qualities still matter. But we have entered an era in which AI can cover much of that ground. If the evaluation axis stays the same, the organization stops growing.
What is needed now is a shift from scoring processing capacity to scoring the ability to design value. Can this person put the real problem into words? Can they bring other people in and still move the work forward? Can they judge, with a clear will, even when the situation is vague? Can they start relearning on their own? These strengths are hard to read from a short-term number. Over a longer horizon, they move the worth of the firm itself.
An OECD report published in April 2024 showed the same direction. As AI spreads, the skills the labour market wants will change, and simple clerical processing will become harder to differentiate on. That is why hiring, development, and evaluation all have to look not only at "how much can this person get through?" but at "how much meaning can this person make?"
This is the second place where a Western operator in Japan can misread the room. US and UK performance systems often run on quarterly OKRs, stack-ranked "impact," and a clean story about individual output. A Japanese owner-manager still weights how a person handles a long client relationship, a messy handover, or a junior who is stuck. If you hire here with a Silicon Valley scorecard, you will keep the people who look busy in a dashboard and lose the people who actually keep landlords, tenants, and counterparties in the deal. The practical risk is not "Japan is slow." It is that you will staff a Japan business with the wrong strengths and call it merit.
How Does INA Think About Investing in People? Management That Does Not Pit Technology Against Human Skill
We do not introduce a new tool and then ask, a month later, who we no longer need. I do not believe AI and human skill stand opposed to each other. The further technology advances, the sharper the worth of human skill becomes. If admin work can be left to AI, people should spend their time on understanding the client, raising the quality of a proposal, building a team, and designing the next challenge — the higher-value ground.
What that requires is not tool adoption alone. It means treating the work as an investment in 人財: redesigning roles, offering chances to take on a challenge, building an environment with psychological safety, and putting in place the support that makes relearning possible. Not AI for cutting people. AI so that people can put more of their worth to work. That difference in starting point will, I believe, open a large gap in the worth of a company a few years from now.
Looked at over the long term, what makes a firm strong is not that it owns convenient tools. It is that it can use technology and still widen what its people are able to do. That is why I believe the AI era makes the idea of being a "human-capital investment company" a more realistic management strategy, not a softer one. For an English-speaking investor or operator hiring or partnering in Japan, this is the practical opportunity. Use AI to give hours back to the people who can run the quiet alignment around a deal, stay with a wobbling client, and say the downside out loud. Keep those people. That is the scarce asset in this market. Cutting them to show a cleaner org chart is the risk.
Who Still Holds Value in the AI Era? Not the Person Who Finishes Tasks, but the Person Who Designs Value
- What AI replaces first is not the person. It is routine admin work.
- The worth of 人財 is moving toward conversation and decision-making. It is also moving toward the capacity to relearn, to be trusted, and to move people.
- Companies need to shift their evaluation from processing capacity to the ability to design value.
- The purpose of adopting AI is not headcount reduction. It is a redesign so that people can move into higher-value work.
- Sustainable growth is built by an organization that can fuse technology with human skill.
Frequently Asked Questions (FAQ)
Q1. Once AI spreads, will clerical jobs really become unnecessary?
A. It is more accurate to say that the content of the role will change, rather than that the role will become unnecessary. The share of routine processing will fall. Judgment, coordination, conversation, and the design of improvements will matter more, not less.
Q2. What kind of person should companies prioritize in hiring in the AI era?
A. Not only the person who waits for instructions and then executes them accurately. You want 人財 who can think in a vague situation, talk with the people around them, and keep making value while they relearn. Adaptability, and a posture that builds trust, both matter.
Q3. Is investing in people the same as adding more training programs?
A. It is not the same. Training is only one part. It becomes an investment in 人財 only when it also includes role design, chances to take on a challenge, the evaluation system, psychological safety, and an environment in which people can keep learning.
Q4. What is the failure companies most easily fall into when they adopt AI?
A. Treating efficiency itself as the purpose. Unless you decide which higher-value work the freed-up hours will be reinvested in, AI adoption ends as a one-off improvement.
What Are the Sources and References?
- Ministry of Health, Labour and Welfare (厚生労働省) Career Development and Reskilling Promotion Program, "Career Design Toward DX Talent"
- OECD, "Artificial Intelligence and the Changing Demand for Skills in the Labour Market"
- OECD, "Building an AI-Ready Public Workforce"
- Microsoft, "The 2025 Annual Work Trend Index: The Frontier Firm is born"
- ILO, "Generative AI and jobs: A 2025 update"
