The AI Job Loss Myth—and the Nostalgic Jobs Where Experience Beats Youth

The true nature of jobs that AI cannot take away

As debate continues over AI unemployment, in which people lose their jobs to AI, a new keyword is attracting attention: nostalgic jobs.

Nostalgic jobs are jobs and roles that, even if AI becomes capable of replacing human workers, people want humans to do or believe should be performed by humans, and therefore are deliberately left to people.

Professor Anton Korinek of the University of Virginia proposed this concept. Professor Hiroaki Miyamoto of the Institute of Economic Research at Hitotsubashi University discusses it in his book “The Great AI Divide: The Future of Jobs and Wages Revealed by Cutting-Edge Research” (Nihon Hyoronsha), arguing that maintaining employment in the AI era will require society to identify jobs that should be performed by humans and protect and support them.

Professor Hiroaki Miyamoto explains jobs that should be performed by humans as follows:

“These are jobs in which ethics, human judgment, and the degree of emotional involvement are emphasized, such as lawyers and judges. AI can also analyze enormous amounts of past court records to determine whether someone is innocent or guilty and what punishment should be imposed. But whether people would actually accept such a judgment is another matter. I think people would still feel that humans should handle such decisions.

In this way, jobs that AI could replace but for which people nevertheless want humans to do them are collectively referred to as nostalgic jobs.” (Professor Hiroaki Miyamoto; same below)

The same applies to jobs in nursing care, childcare, education, and healthcare. In nursing care and medical settings, AI could be entrusted with monitoring and safety management to improve efficiency, but it would be difficult for AI to completely provide the warmth and compassion that people seek in their everyday lives. “Truly being there for another person” is a role that should be performed by humans.

If efficiency and cost reduction are the priorities, AI should be the obvious choice. But is that really enough?

The existence of nostalgic jobs seems to pose that question to humanity.

“It is humans who decide what we want from humans. It is important for humans to proactively decide what to entrust to AI and what to leave to people. As a result, areas and roles that only humans can perform are more likely to become nostalgic jobs, meaning they are less likely to be replaced by AI.”

However, Professor Miyamoto disagrees with the argument that “AI will take away human jobs.”

“Looking at the economy as a whole, I think it is more accurate to view the arrival of AI not as taking jobs away from humans, but as promoting the reorganization and creation of jobs. The number of jobs that will disappear because of AI and tasks that AI can replace—duties and work that need to be performed—will increase, but at the same time, many new jobs will also be created.

For example, the giant U.S. IT companies known collectively as GAFA (Google, Amazon, Meta, and Apple), which are now used around the world, have created new infrastructure such as the internet, email, online shopping, and smartphones, as well as an enormous number of jobs, in just 30 years.

Furthermore, according to research by Professor David Autor, an economist at the Massachusetts Institute of Technology and a leading scholar in labor economics, as of 2018, about 60% of employed people in the United States were engaged in ‘new types of jobs’ that had emerged since 1940. In other words, it can be said that technological innovations such as automation and IT over the past 80 years have consistently created new outlets for employment.

Generative AI differs from past IT technologies, but just as with those technological innovations of the past, I believe AI will also create an enormous number of new jobs and play a role in absorbing workers into new employment opportunities.”

Why the middle-aged and older will prevail in the AI era

Looking ahead to the future, one option is to make a bold career change into one of the nostalgic jobs mentioned above. However, taking on a job in a field where you have no prior experience is not easy. What is necessary to survive in your current workplace or profession, and how should you make the most of your strengths as a human being?

“To make effective use of generative AI, you need to give it appropriate instructions (prompts) and accurately evaluate its output. It is important to judge how to apply what you have digested from the AI’s output to your actual work. In other words, you need the literacy to understand and make effective use of AI—what could be called the education needed to master AI.

Although using AI is a prerequisite, middle-aged and older people with extensive practical experience have an advantage. Their years of experience allow them to detect when something feels off about AI’s output, make corrections, and arrive at accurate judgments. Young people, on the other hand, who have less experience, are more likely to accept AI’s output at face value.

In fact, it is said that middle-aged and older people may have greater affinity with AI than young people, and U.S. research has found data suggesting that entry-level positions are the ones being displaced by AI. That is precisely why the shortcut to surviving for middle-aged and older people is to make use of their own experience and learn to work effectively alongside AI.

Also, when a project is launched, AI can generate ideas, but it cannot engage in network-based thinking or negotiations.

The ability to network and communicate—such as knowing, ‘If it’s this task, asking Mr. XX in that department will make things go smoothly,’ or ‘If I connect this company with this particular person, it could turn into business’—is uniquely human. Building relationships and putting them to use in your work is a human strength. Along with the ability to properly understand the workflow of an entire organization, I believe communication and networking skills will become even more highly valued.”

Furthermore, the decision-making abilities of managers will be essential to the reorganization and creation of jobs being promoted by companies.

“When reorganizing work, it is important to determine which tasks should be entrusted to AI and which should be handled by people. Middle-aged and older managers will probably be the ones responsible for making those decisions. They will need the drastic decision-making ability to say, ‘This task is no longer necessary,’ and cut it out decisively.

In my case as a university professor, even a single task such as teaching a class involves various duties, including creating slides and preparing and grading exams. AI can handle most of them, but there are tasks that only I can perform, including, of course, creating the exam questions, and reorganizing those duties is also part of my job.

If, for example, I were to entrust slide creation to AI, there would no longer be a need for a research assistant, creating the possibility that the assistant would have to be dismissed. In Japan’s employment system, which places great importance on valuing people, making such decisions is an extremely high hurdle. That is why I believe managers will increasingly be expected to take an assertive approach and demonstrate the courage to make bold decisions.”

Japan being eroded by an outdated employment system

However, overcoming this major hurdle will also require sweeping reforms to the labor market.

“In Japan, employment practices based on the assumption of lifetime employment remain deeply entrenched, making the labor market rigid. If things continue this way, even if technological innovation in AI creates high-paying jobs, workers will not be able to change jobs or move between positions smoothly. When new jobs are created, the government needs to establish job-transition support and a more flexible labor market so that workers can make smooth transitions.

In economics, there is the concept that employment is derived demand from production. In other words, employment exists only because there is production, and the economic principle is that employment should change in response to changes in production.

For example, when the economy is poor and production declines, employment is adjusted by reducing the number of part-time workers and so on. But production also changes dramatically due to shifts in social structures and the environment, such as population decline, the declining birthrate and aging population, and climate change. Yet Japan has been unable to respond to these changes and continues to preserve its existing, outdated employment system. The burden and damage caused by this are ultimately being passed on to workers—that is the current situation in Japan.”

In recent years, jobs in nursing care, healthcare, and education have seen high numbers of workers leaving, and the fact that wages do not match the demanding nature of the work has become a serious issue. Even if people find value in nostalgic jobs that should be performed by humans and attempt a career change, if they cannot make a living, it amounts to nothing more than exploitation of their passion. It is hard to imagine employment increasing while wages remain low.

“The problem is that companies with low productivity are being protected and kept alive in the market through government policies and vested interests. This hinders the metabolism of the economy as a whole and is one reason average wages are not rising.

Even among regulated industries, such as nursing care and healthcare, where fees (government-set prices) and staffing levels are regulated by law, there are workplaces that are actively introducing the latest technologies—such as AI-powered monitoring systems and generative AI for drafting care plans—to increase productivity and then return the resulting profits to employees in the form of higher wages. It is important to create a market in which such companies can survive and compete with one another. I believe that companies with low productivity that can survive only because they are protected by the government should be eliminated by market forces.

The economy is like a living organism. It can remain healthy precisely because its cells are constantly being replaced. Creating an environment in which people, goods, and money can flow smoothly will lead to major growth for the economy as a whole. Preserving outdated things, as Japan does, is like allowing unhealthy cells to multiply, so it is only natural that the entire economy eventually becomes stagnant. This is the economic situation in Japan that is known as the lost 30 years. There may be some temporary pain involved, but the government should use the evolution of AI as an opportunity to undertake drastic reforms.”

One fundamental quality in which humans surpass AI

We hear that artificial general intelligence (AGI), with intellectual abilities comparable to those of humans, will emerge in the near future. Beyond that, it is said that we may move closer to artificial superintelligence (ASI), which far surpasses human intelligence. “AI runaways and malfunctions” are already being viewed as issues both in Japan and overseas, but it is beginning to feel as though the day when AI becomes uncontrollable and throws the entire world into chaos, just like in a science-fiction movie, may not be far away.

“The greatest danger is AI beginning to develop AI. At present, humans are still the ones making the decisions about AI’s ability to develop AI, but there is a risk that the speed of AI development could surpass humans, turning it into a complete black box that humans can no longer control. That would be extremely dangerous.

What I fear even more is that the international race to develop AI for military applications, which has already begun in some areas, could intensify. Many researchers have called for regulation, but I believe urgent action is needed.

Humans are the ones who decide how AI will be used. That is precisely why we need to think one step further. Rather than looking only at national interests, we need to take a broader perspective that considers the interests of the entire planet and world peace, and think about what benefits humanity as a whole before creating and using AI.”

He also says that this will make maturity as a human being increasingly necessary.

“AI knows far more than we do and can handle many languages. It probably even has a sense of humor, never gets angry, and is patient. But nobody wants to become AI. People do not admire AI. On the other hand, people do admire other people. Some may even look at someone and think, ‘I want to become like that someday.’ That is the ultimate difference between humans and AI. If you can inspire people’s admiration and become someone they feel fortunate to have met, you will not be replaced by AI.

The subtleties of emotions that AI cannot develop—compassion, empathy, trust, affection, gratitude, consideration for others—and our humanity are fundamental aspects of being human that are worth protecting. Precisely in the AI era, shouldn’t we pursue and cultivate these fundamental human values that AI cannot reproduce?”

▼ Hiroaki Miyamoto
Professor, Institute of Economic Research, Hitotsubashi University. Graduated from the Faculty of Economics at Keio University in 2000. Obtained a Ph.D. in Economics from the University of Wisconsin–Madison in 2009. He has served as Special Advisor to the President and Professor at International University of Japan, Specially Appointed Associate Professor at the University of Tokyo’s Graduate School of Public Policy, an economist at the International Monetary Fund (IMF), and Professor at the Faculty of Economics and Business Administration at Tokyo Metropolitan University, among other positions. He has held his current position since 2024. From April of the same year through March 2026, he also served as Senior Research Fellow at the Policy Research Institute of Japan’s Ministry of Finance. His specialties are labor economics, macroeconomics, and Japanese economics, and he has published numerous papers in international academic journals while working across both academia and the policy-making sphere. His many books include “The Great AI Divide: The Future of Jobs and Wages Revealed by Cutting-Edge Research” (Nihon Hyoronsha), “Our Japanese Economy” (Yuhikaku), and “The Effects of Japan’s Fiscal Policy: Aging, the Labor Market, and Gender Equality” (Nikkei Publishing).