Why Are Chinese Business Owners Becoming So Fond of Robots?

Originally published on Substack.

It’s not simply because robots are cheaper than people. It’s because managing people is becoming increasingly expensive.

Chinese business owners have actually liked robots for a long time.

They just didn’t always have a strong reason to say so.

For the past two or three decades, one of China’s biggest manufacturing advantages was remarkably simple:

Labor was cheap, and there was plenty of it.

A factory needed more workers? Hire them.

Orders increased? Add overtime.

A production process needed ten people? Find ten people.

If someone didn’t work out, replace them.

This was one of the fundamental economic mechanics behind Chinese manufacturing for decades.

It allowed countless companies to industrialize at extraordinarily low cost, and it gave many business owners their first taste of the enormous leverage that comes with manufacturing at scale.

But that logic is changing.

Not because China suddenly has no people.

As of the end of 2025, China still had 851 million people aged 16–59, representing 60.6% of the population, and more than 725 million people employed. China still has an enormous labor force.

What is changing is something more subtle:

The cost of labor is no longer just the wage.

It is management.

1. A company is never really buying just eight hours of someone’s time

Suppose a factory hires one employee.

The owner sees:

Salary, social insurance, accommodation, recruiting costs.

But the real operating cost also includes:

Training.

Communication.

Scheduling.

Leave.

Morale.

Turnover.

Labor relations.

Safety responsibilities.

Quality fluctuations.

Managerial overhead.

Coordination between teams.

And sometimes the cost of an entire production line being disrupted because one person suddenly quits.

So an employee’s salary is never their true cost to the company.

Salary is the visible cost.

Management is the hidden cost.

And the larger the organization becomes, the more significant that hidden cost becomes.

That is why, when companies seriously evaluate automation, many owners are no longer asking only:

“How much does the robot cost compared with a worker?”

They are asking:

“If I buy this machine, can I eliminate the need to manage ten people?”

Those are two completely different calculations.

2. Why are Chinese manufacturers particularly receptive to robots?

Because Chinese companies have never had much of a psychological barrier to replacing manual work with machines.

Over the past several decades, Chinese manufacturing has gone through wave after wave of automation.

Manual material handling became conveyor systems.

Manual inspection became machine vision.

Manual welding became robotic welding.

Manual loading and unloading became automated production lines.

One worker operating one machine became one person supervising an entire production system.

Robots did not suddenly arrive in Chinese factories.

They have been arriving for decades.

According to the International Federation of Robotics, China installed 295,000 industrial robots in 2024, accounting for more than half of all new industrial robot installations worldwide. By the end of 2024, the installed base in China had exceeded two million industrial robots.

There is a very simple reason for this:

Chinese companies have already demonstrated that they know how to turn robots into productivity.

More importantly, China’s supply chain has continued to drive down the cost of the robots themselves.

Motors, reducers, servos, controllers, machine vision, sensors, batteries, structural components, precision machining, electronics manufacturing, and countless other capabilities have all benefited from enormous manufacturing scale.

This creates an interesting feedback loop:

Companies are willing to buy robots.

Because companies buy robots, robot manufacturers get more orders.

More orders create economies of scale.

Economies of scale make robots cheaper.

Cheaper robots make more companies willing to buy them.

That is one reason China has become the world’s largest robotics market.

Increasingly, the question for many manufacturers is no longer:

“Should we automate?”

It is:

“At what point does this particular process become worth automating?”

3. Government and business are increasingly aligned on this

This is not simply a slogan.

In 2023, China’s Ministry of Industry and Information Technology and 16 other government departments released the “Robot+” Application Action Plan. One of its targets was to double manufacturing robot density by 2025 compared with 2020, while accelerating robot adoption across key application scenarios.

For humanoid robots, manufacturing has also been explicitly identified as an important application area.

China’s Ministry of Industry and Information Technology’s guidance on humanoid robot innovation and development calls for accelerating large-scale commercialization by 2027 and integrating humanoid robots more deeply into the real economy, including applications in manufacturing sectors such as consumer electronics and automobiles.

The logic is straightforward.

For the government:

Robots mean higher manufacturing productivity.

They mean industrial upgrading.

They mean supply-chain upgrading.

And they provide one possible way to maintain manufacturing competitiveness as the demographic structure changes.

For companies, robots offer something different:

They can turn part of an unpredictable labor cost into a relatively measurable equipment cost.

4. This is why I think the real reason Chinese companies like robots has been under-discussed

It is not necessarily because robots are smarter.

And it is not simply because robots do not get tired.

It is because:

Robots are easier to manage.

A robot does not have a bad day and suddenly become 10% less productive.

It does not work more slowly because its boss criticized it yesterday.

It does not tell you at the end of the month:

“I’m leaving.”

It does not require an hour-long meeting to resolve a conflict with a colleague.

Of course, robots break down.

But when a machine breaks, an engineer can usually ask:

What is the motor temperature?

What is the current?

What is the vibration level?

What is the fault code?

Has the component reached its expected service life?

Are spare parts available?

At least these problems can be engineered, measured, and systematized.

Problems involving people are often not engineering problems.

5. This is also why “higher education means easier to manage” does not necessarily hold

A recent case involving Chinese automotive lighting manufacturer Xingyu deserves attention from this perspective.

According to a public statement, Xingyu had recruited 440 university graduates for its 2026 intake, of whom 107 had their employment contracts terminated. Local authorities said the company’s approach to communication and negotiations had been overly simplistic and lacked sufficient effective communication. Xingyu subsequently apologized publicly and announced compensation measures.

The incident should first be understood as a specific labor-relations and management failure. It would be an overreach to conclude:

“College graduates are difficult to manage.”

But the incident does raise a more important question:

What kind of people does a company actually need?

Higher education generally brings stronger knowledge and analytical capabilities.

But as people become more capable, their expectations around career development, autonomy, professional identity, and management may also change.

If a company hires university graduates but treats them essentially as interchangeable production-line labor, they may naturally ask:

Why?

Where is the growth opportunity?

Does this process actually make sense?

What exactly is my responsibility?

Why do we have to do it this way?

Is there a better way?

A machine will not ask those questions.

It simply executes the task it has been programmed or instructed to perform.

So the problem is not:

“University graduates are bad.”

It is:

The more capable people are, the more carefully an organization needs to create a meaningful place for that capability.

If an organization can offer only standardized, repetitive work while hiring large numbers of highly educated people, friction is almost inevitable.

That is not necessarily a problem with the people.

And it is not necessarily a problem with education.

It is often a problem of organizational design and talent structure being mismatched.

6. The people robots replace may not simply be “low-skilled workers”

This is what I think will become increasingly important over the next several years.

When we talk about robotics, we often say:

Robots will replace factory workers.

That is certainly part of the story.

Robots will continue to replace repetitive, dangerous, dirty, physically exhausting, and highly standardized work.

But as AI moves into robotics, something more interesting starts to happen.

Work that previously required substantial human supervision can increasingly be turned into processes that software can define, monitor, and control.

Today:

One supervisor manages 20 people.

Tomorrow:

One supervisor may manage 10 people plus 30 robots.

Eventually:

One engineer may manage an entire automated production system.

Humans handle exceptions.

Machines handle the normal flow.

That is a profound organizational change.

Because the deepest transformation brought by industrialization has never simply been about machines replacing people.

It has been about:

Turning human experience into processes; turning processes into software; and turning software into actions machines can execute.

7. Why humanoid robots make this even more interesting

Traditional industrial robots have one major limitation:

They need a world designed for robots.

Fixed positions.

Fixed tooling.

Fixed materials.

Fixed trajectories.

Fixed environments.

For decades, automation essentially followed one principle:

Change the factory to accommodate the machine.

Humanoid robotics is attempting something different:

Put machines into environments that were already designed for humans.

If that ultimately works, the potential market becomes much larger.

Most production and service environments in the world were designed around the human body.

Door handles.

Shelves.

Tools.

Workbenches.

Cars.

Warehouses.

Kitchens.

Stores.

If a robot can use these existing systems directly, companies no longer need to redesign the entire environment around the robot.

That is one reason manufacturing, logistics, and service environments are repeatedly discussed as potential humanoid-robot applications.

But this is where some restraint is necessary.

A company being willing to buy a robot does not mean it is willing to pay for a demo.

What ultimately determines whether a robot enters a real business is much more mundane:

What does it cost?

How many hours can it work each day?

What is the failure rate?

How much does maintenance cost?

Can it operate reliably?

Can it integrate with the existing production line?

Who is responsible if there is a safety incident?

How long until the investment pays back?

Until these questions are answered, even the most impressive humanoid robot remains a demonstration rather than a product.

8. What Chinese companies really want is “computable labor”

This is why I increasingly think:

The biggest competition in robotics may not be about who looks most human.

It may be about:

Who can behave most like a good employee while allowing the company to calculate the economics clearly?

A business owner ultimately wants to know:

Can it work 20 hours a day, or only eight?

What does each completed task cost?

How many failures occur per 10,000 operations?

How long does it take to repair?

What will maintenance cost over a year?

If I add one robot, can I eliminate the need to hire three people?

Can I reduce the number of shift supervisors?

Can I eliminate night shifts?

Can I keep the hardest-to-recruit workers focused on higher-value tasks?

If these questions can be answered, the robot stops being merely a technology product.

It becomes a new factor of production.

9. This is also where the real opportunity for Chinese robotics lies

One of the easiest mistakes for Chinese robotics companies to make is to think of themselves primarily as technology companies.

They spend enormous amounts of time discussing:

Models.

Parameters.

Degrees of freedom.

Motors.

Dexterous hands.

Benchmarks.

All of these things matter.

But for an industrial customer, the final question is remarkably simple:

“Can you actually do this job for me?”

And not just once during a demonstration.

Every day.

Continuously.

With someone available to repair it when it fails.

With different materials.

At different workstations.

And, after everything is included, at a lower cost than the human alternative.

That is what productization in robotics really means.

10. So I am not pessimistic about “machines replacing people”

Because machines should not replace human value.

They should replace:

the things humans do not need to personally do.

If someone spends eight hours a day carrying materials, inspecting repetitive parts, performing the same motion, recording data, waiting, and correcting predictable errors, replacing those tasks with machines is not necessarily a bad thing.

The more important question is something else:

As machines become cheaper and the organizational cost of human labor rises, companies will increasingly ask: Which jobs actually need to be done by humans?

The answer could reshape many industries over the next decade.

It will also reshape organizations.

In the past, one of a factory’s most important assets was its workforce.

In the future, it may be:

People + software + robots + data + supply chain.

Human value will not disappear.

But it will increasingly concentrate in areas that are difficult to standardize:

Judgment.

Creation.

Relationships.

Responsibility.

Decision-making.

Complex problem solving.

And, ultimately, taking responsibility for the entire system.

That may be the genuinely difficult, and genuinely hopeful, part of the robotics era.

It is not simply that machines will replace people.

It is that:

Work that cannot create higher value will become increasingly difficult to sustain at its old price.

Meanwhile, people who can define problems, manage machines, build products, solve exceptions, and take responsibility for outcomes may become more valuable than ever.

For decades, one of China’s greatest competitive advantages was its ability to make human labor extremely inexpensive and then scale production enormously.

The next stage may be different.

The goal will no longer be to find cheaper people.

It will be to find a form of production that is more measurable, more stable, and easier to scale than human labor.

Robots are simply the most visible part of that transition.

What is really changing is the way companies define labor itself.