"Is 'Jobs Disappearing' an Exaggeration? Bill Gates Warns of a Major Employment Shift in the AI Era"

Will our jobs become more convenient with the evolution of artificial intelligence, or will jobs themselves disappear?

In response to this question, a person who has built one of the world's leading technology companies is showing more caution than ever before.

In August 2026, Bill Gates, co-founder of Microsoft, discussed the changes AI will bring to society and the economy in a lengthy essay. He particularly emphasized the impact on employment.

What Gates is concerned about is not simply the automation of some jobs.

He believes that AI will take on roles in a wide range of fields that have traditionally required human knowledge and judgment, such as law, customer service, healthcare, software development, and manufacturing, and that these changes could occur much more rapidly than past industrial revolutions.

Looking back at history, it is not uncommon for new technologies to render old professions obsolete.

Mechanization in agriculture reduced the number of agricultural workers, automation in factories changed manufacturing processes, and the advent of the internet transformed many businesses.

On the other hand, past technological innovations have also given birth to new industries and jobs.

However, Gates points out that the current AI revolution has characteristics that differ from previous ones.

AI is beginning to replace not only "human physical abilities" but also "human cognitive abilities" themselves.

The era when "jobs that require thinking are safe" might end

Until now, jobs most susceptible to automation were typically simple factory tasks or routine clerical work.

However, the emergence of generative AI has changed the situation.

Writing articles.

Reviewing contracts.

Writing programs.

Analyzing large amounts of data.

Answering customer inquiries.

Creating documents.

Producing images.

These tasks were once thought to be difficult for computers.

Nowadays, it is becoming increasingly common for AI to handle the majority of these tasks, with humans only verifying the final results.

Gates envisions a major turning point where AI's accuracy improves to the extent that even human final verification is no longer necessary.

From a business perspective, if a system exists that can continuously produce results equivalent to or better than humans at a lower cost, the economic temptation to adopt it is extremely high.

AI does not take vacations and can work overnight.

There is no need to spend time on hiring or training.

It can replicate the same tasks in large quantities.

If this happens, a gap in competitiveness will emerge between "companies that adopt AI" and "companies that continue human-centered operations," forcing the latter to automate as well.

Thus, the issue is not whether managers like AI, but the possibility that market competition itself will accelerate automation.

The first to be affected might be "newcomers"

A particularly important aspect of employment changes due to AI is its impact on young people.

In many professions, newcomers have traditionally gained experience through relatively simple tasks.

Young programmers would write simple code.

New analysts would organize documents and data.

Junior staff at law firms would research literature and precedents.

Marketing personnel would gather market information and create draft materials.

These tasks were not mere chores.

They were also "training grounds" for learning the job, developing judgment, and advancing to more advanced work in the future.

However, this is precisely the area where generative AI excels.

A study published by Stanford University's Digital Economy Lab in August 2026 reported that while large-scale job losses due to AI were not confirmed across the U.S., employment growth among young people aged 22-25 in occupations susceptible to AI was significantly lagging compared to occupations with low AI exposure.

According to the research team, this gap expanded to 19% by mid-2026.

However, the researchers themselves caution that this does not directly prove AI as the cause, as multiple factors such as interest rates, economic conditions, and company hiring policies exist in the labor market.

Nevertheless, the possibility that adjustments are progressing not only in the form of "mass layoffs" but also as "not hiring young people as much as before" is significant.

The employment issue in the AI era may not manifest dramatically as millions of employees being laid off overnight but rather as a gradual reduction in job openings for newcomers, narrowing the very entrance for humans to start their careers.

This could lead to even more serious problems in the future.

If newcomers are not hired, there will be no mid-level employees in ten years.

The next wave may hit blue-collar jobs

It might be premature to think that only "jobs that can be done on a computer" are at risk.

Gates cites the fusion of AI and robots as the next major change.

While AI has rapidly become smarter, robots that handle and move objects in the real world have progressed relatively slowly.

However, in recent years, improvements in sensors, AI, and control technology for robots have advanced research and development of machines capable of flexible tasks based on human instructions.

Gates points out the possibility that by the 2030s, "smart robots" will compete with humans even in some physical tasks such as construction and hospitality.

If this happens, the employment issue due to AI will not just be a story of white-collar workers versus generative AI.

The wave of automation could spread to jobs dealing with the real world, such as in factories, warehouses, hotels, food service, construction, and logistics.

Price is particularly important.

If a job requiring $20 per hour for a human can be done by a robot with equivalent capabilities for $10, there will be a strong incentive for companies to adopt it.

If one company adopts it and lowers the price of its products or services, competitors will follow suit.

This creates a cycle where automation begets more automation.

The emergence of the "AI and robot tax"

As a way to prepare for such a future, Gates has long supported the idea of taxing robots.

Under the current social system, hiring humans incurs taxes and social insurance premiums related to wages.

On the other hand, investments in machines or software are treated as capital investments by companies, which can sometimes be tax-advantageous.

This structure could create a distortion where replacing humans with machines is more beneficial for companies.

Thus, the idea arises to impose a certain tax burden on AI usage and robots, directing that revenue toward supporting unemployed workers, vocational training, and social security.

The goal is not to ban AI.

It is to slightly slow down the extreme pace of replacement of humans with AI driven solely by economic rationality, allowing society time to adapt.

However, if it were to be institutionalized, there would be many challenges.

How to define "one job" done by AI.

How to tax AI used internally by companies.

Can AI that enhances human productivity be distinguished from AI that replaces humans entirely?

Will it put companies at a disadvantage in competition with overseas firms?

If AI usage in socially beneficial areas like healthcare and education becomes too costly, it could be counterproductive.

The robot tax should be seen not as a simple solution but as a starting point for discussions on how to redesign the tax system in the AI era.

The idea of preserving "jobs only humans can do"

Another intriguing concept from Gates is "Human Reserved."

It is the idea of deliberately leaving certain areas to humans, even if AI and robots can technically do them.

In nature reserves, even land where roads and buildings could technically be constructed is left undeveloped to protect nature.

The idea is to apply the same concept to jobs.

A prime example could be caregiving.

In the future, robots might be able to deliver meals to the elderly and manage their medications.

However, should actions such as reading a person's condition from their expressions, conversing, alleviating anxiety, and accompanying them in the final stages of life be evaluated solely on efficiency?

In jobs where the relationship between people itself has value, such as education, childcare, caregiving, healthcare, and counseling, the idea emerges that they should not be entirely replaced by AI just because "AI is cheaper."

Meanwhile, Japan has different circumstances.

With a declining workforce due to a low birthrate and aging population, and a severe shortage of caregiving personnel, there might be a decision to actively utilize robots rather than restrict them to protect human jobs.

Therefore, "protecting human jobs from AI" is not the correct answer for every country.

The optimal solution varies depending on population structure, social security systems, unemployment rates, and the degree of labor shortages.

On social media, there are harsh reactions like "Is he finally saying it?"

Gates' warnings have sparked discussions on social media and online communities.

Particularly in Reddit's technology-related communities, reactions focused more on "it's late to issue a warning" than on Gates' actual concerns.

AI researchers and critics have long pointed out the impact on employment, and there is irony in a person who built one of the world's largest technology companies now expressing heightened concern.

Additionally, criticisms have been posted questioning whether the tech industry has been expanding AI investments while simultaneously reducing staff.

This reflects a distrust not just towards Gates personally, but towards the entire AI industry.

Some feel there is a contradiction in tech companies explaining to investors that "AI will improve productivity" while telling society that "we need to prepare for the impact on employment."

On the other hand, there are reactions showing understanding of the idea of a robot tax, with sentiments like "there's a need to review the current system where human labor incurs taxes and social security burdens, while only automation is favored."

In the AI community, opinions are divided on how highly AI's capabilities should be evaluated.

While some believe the "AI boom is overrated," others think "current generative AI is already practical enough, and if its capabilities improve further, the impact on employment is unavoidable."

Thus, the focus of the debate is shifting from "whether AI will affect jobs" to "how quickly and to what extent it will impact them."

However, it is important to note that posts on Reddit and similar platforms represent user opinions and are not representative of broader public opinion polls.

The anxiety on social media is also reflected in public opinion polls

Nonetheless, the anxiety about AI is not something that can be dismissed as a phenomenon limited to some internet users.

A survey conducted by the Pew Research Center in June 2026 and published in August found that 52% of U.S. adults feel "more anxious than hopeful" about the increasing use of AI in daily life.

In 2021, the same response was given by 37%, indicating a significant long-term increase in caution.

Particularly among those aged 18-29, 55% in 2026 responded that they feel "more anxious than hopeful."

It is symbolic that the younger generation, often thought to be the most proactive users of AI, is wary of its impact.

For young people, AI is not a distant future issue.

Skills acquired over several years in university or vocational training might be replaceable by AI by the time they graduate.

The very choice of "what to learn to be safe" in career planning is becoming more difficult.

However, it's premature to conclude that "mass unemployment has already begun"

It is important not to take Gates' warnings as a predetermined future.

Current data does not allow for a definitive conclusion that large-scale unemployment due to AI is occurring across the U.S.

Stanford's research also did not confirm large-scale job losses spreading throughout the economy.

Moreover, AI has the effect of enhancing human capabilities, not just replacing jobs.

A single doctor can see more patients.

Small businesses can perform data analysis on par with large corporations.

Programmers can develop services in a shorter time.

Language barriers are lowered.

Teachers can create materials tailored to each student.

Such productivity improvements could lead to new services and demand, potentially increasing employment as a result.

Therefore, the real issue is not a binary choice of "whether AI will take jobs or not."

It is a distribution issue of who will receive the benefits generated by AI and who will bear the costs of