Twenty years ago, when Chinese entrepreneurs went to the US, the most important stop was Silicon Valley, where they would visit Google to learn about the internet, Apple to learn about products, Stanford to learn about innovation, and Sand Hill Road to learn about venture capital.
At that time, the whole world had a very clear division of labor in the tech industry: the United States was responsible for invention, and China was responsible for manufacturing.
But 20 years later, an interesting phenomenon has emerged, with the direction reversing. Reuters reported on September 3 that an increasing number of overseas investors, entrepreneurs, and corporate executives are making special trips to Shenzhen, Hangzhou, Shanghai, Beijing, and Hefei to visit Chinese artificial intelligence, robotics, electric vehicle, and advanced manufacturing companies, and this is no longer just individual travel, but has even started to become a business.
A five-day tech tour organized by Shanghai-based research institution Baiguan costs up to $15,000. Another Shanghai tech tour company, GloPen, said its consulting volume increased by 50% in 2026, with clients mainly from Europe and Singapore, and it now organizes over 100 one-day corporate visits every month. Even factory tours have become a scarce resource. Xiaomi's Beijing car factory has received over 250,000 visitors since March 2024, with some tour spots won through lottery being resold online for as much as 2,000 yuan.
These individuals spend thousands or even tens of thousands of dollars to fly to China, not to visit an ordinary factory, but to truly understand why Chinese tech companies are moving so fast.
This may be one of the biggest changes happening in global tech competition today. In the past, when people studied China, they focused on why China could manufacture things so cheaply; nowadays, more and more people are studying why China can turn an idea into a product so quickly. These two questions may seem similar, but they represent two completely different eras.
The World is Reassessing
Made in China
For decades, the most memorable label associated with "Made in China" has been cost. Cheap labor, massive factories, and a complete export system have made China the world's factory. However, if one still understands Chinese industry through the lens of "low-cost manufacturing" today, they are likely severely outdated.
One of the most intuitive sets of data is that in the mid-1990s, China accounted for around 5% of global manufacturing output, and today this number is close to 30%. In other words, nearly one-third of global manufacturing output comes from China, meaning that for every three dollars of manufacturing output worldwide, approximately one dollar comes from China.
More importantly, what has changed is not just the quantity, but also the structure. Previously, China's exports were clothing, shoes, toys, and home appliances, but today's increasingly important export products have become new energy vehicles, power batteries, photovoltaics, drones, industrial robots, and smart hardware.
This means that Chinese manufacturing is undergoing a very significant upgrade: from a "cost center" to an "innovation infrastructure".
New energy vehicles are a typical example. In 2025, the world produced nearly 22 million electric vehicles, with China producing around 16 million, accounting for nearly three-quarters of global electric vehicle production. China also has more than 80% of the world's battery cell production capacity, about 85% of the positive electrode material production capacity, and more than 90% of the negative electrode material production capacity. This means that a large part of the core supply chain behind a new energy vehicle company can be found in China. As a result, Chinese new energy vehicle companies have acquired a very important capability: rapid iteration.
It used to be normal for a car to take five or six years to develop, but today the competitive pace of China's new energy vehicle market is so fast that many traditional car companies are struggling to keep up. What do consumers need? Smart cockpits, assisted driving, charging speed, range, refrigerators, screens, software ecosystems... The market provides feedback, companies modify their products, suppliers make adjustments in tandem, and new models re-enter the market.
So, what's really worth studying is not "why Chinese cars are so cheap," but why Chinese automakers can iterate so quickly.
What's Truly Frightening Isn't 'Made in China'
Instead of Chinese Iteration
Looking at robots again, in 2024, 54% of newly installed industrial robots in factories worldwide were in China, with approximately 295,000 industrial robots installed in the country in one year. This means that for every two new industrial robots added globally, more than one is going into Chinese factories. The total number of industrial robots operating in Chinese factories has exceeded 2 million units.
The significance of this data set goes beyond the number of robots we see doing backflips in short videos, as AI and robots are truly entering industries, not just performing on stage, but rather welding, transporting, sorting, assembling, inspecting, and handling logistics in factories.
The difference between these three issues is not just a matter of technology, but an entire industrial system.
What Shenzhen is Really Selling to the World
So why is Shenzhen one of the core cities for this round of overseas technology inspections? It's because the most worthy of study in Shenzhen has never been a particular company, but rather the industrial density behind it.
To build a robot, you need chips, cameras, lidar, sensors, motors, gearboxes, lead screws, batteries, structural components, molds, control systems, software, contract manufacturers, testing equipment, and logistics. If these enterprises are scattered across five countries, a design modification could mean weeks of communication. However, if these suppliers are concentrated in a highly dense industrial network, things are completely different. Engineers can identify a problem in the morning and find a supplier in the afternoon; re-sample a few days later; discover the cost is too high and switch materials; find the structure is unreasonable and modify the mold. After the first generation enters the market, the second generation can be developed immediately based on customer feedback. As a result, what is truly compressed is time.
What has truly taken shape in Shenzhen is not just a simple supply chain, but a massive "technological time compressor".
Drones are an extreme case. Shenzhen is home to over 1,500 drone-related companies, with Shenzhen-based enterprises' consumer drones once occupying around 74% of the global market. Meanwhile, Shenzhen-based DJI holds an estimated 70%-80% share of the global non-military and non-governmental drone market, according to industry research institutions.
Why is this industry concentrated in Shenzhen? The answer is not just because DJI was born here, but because the motors, batteries, cameras, transmitters, chips, structural components, gimbals, software, and precision manufacturing capabilities required for drones already exist within this industrial network.
It's not that a great company creates an entire industry chain, but to a certain extent, a sufficiently dense industry network is more likely to continuously foster great companies.
What's Really Emerging in China
An 'Industrial Compression Zone'
So I am increasingly inclined to use a term to describe the capability that is taking shape in China's tech industry today: industrial compression field.
What is an industrial compression field? It refers to compressing technology, engineers, supply chains, manufacturing, capital, markets, and application scenarios into a highly dense industrial network.
Looking at any one aspect individually, China may not necessarily have an absolute advantage. The US still has the world's top universities, still has a strong foundation for basic scientific research, still has the world's leading AI chip companies, and still has Silicon Valley's vast network of venture capital. However, what makes China truly unique is that a large number of tech elements are starting to cluster together in physical space.
This proximity will produce a huge multiplier effect. Automotive companies need batteries, battery companies need materials, and material companies need energy; robots need motors, sensors, and batteries, drones need chips, vision, and communication, and AI needs servers, which in turn require chips, power, and data centers. As a result, different industries begin to feed into each other, and technology is no longer advancing along a single industry chain, but instead is constantly intersecting in a vast industrial network.
This is also why the most worthy of study in China today is no longer a particular company, but rather why so many industries are able to evolve rapidly here at the same time.
A More Important Case
Why Tesla Built Its Gigafactory
Put it in Shanghai?
To understand China's supply chain, there is another classic case: Tesla's Shanghai Gigafactory.
Many people used to understand this factory with a single logic: the Chinese market is large. But what's truly noteworthy is the industrial significance it has formed later on. The Shanghai factory not only serves the Chinese market, but has also become an important base for Tesla's global export system. By the second quarter of 2026, more than half of the cars produced by Tesla's Shanghai factory were for export. A quintessential US tech automotive company has turned its Chinese factory into a key production hub serving European, Asia-Pacific, and other markets, and the underlying reason ultimately boils down to a few key words: supply chain, efficiency, scale, and industrial cluster.
This is also why the so-called "decoupling" today is far more complicated than a political slogan. Because a factory can be relocated, but the hundreds of suppliers surrounding it, the tens of thousands of engineers, the logistics system, the mold system, the material system, and the decades of accumulated engineering experience cannot be replicated with a single click. The truly valuable assets have never been the factories themselves, but the invisible, yet continuously improving, industrial collaboration network behind them.
The AI Era
Why Does Manufacturing Matter More Than Ever?
Many people may ask: in the AI era, shouldn't algorithms be the most important thing? Why are we still discussing manufacturing? On the contrary, the more AI develops, the more important manufacturing may become.
Due to the internet revolution of the past twenty years, which mainly took place in the digital world, companies like Google, Facebook, TikTok, and WeChat are essentially software-based, and the biggest characteristic of software is that its replication cost is close to zero - an app can be developed and quickly cover hundreds of millions of users. However, the next stage of AI is entering another world: Physical AI. With robots, autonomous driving, drones, AI glasses, smart cars, smart factories, energy storage systems, and AI terminals, when artificial intelligence begins to have a "physical body", the problems become completely different.
AI is no longer limited to just GPUs, but also requires motors, gearboxes, sensors, batteries, cameras, materials, factories, and supply chains. As of 2024, there are already over 4.7 million industrial robots in operation worldwide, with more than 500,000 added annually, and China accounts for more than half of the world's new industrial robot installations. This means that while the first half of AI may have taken place in servers, the second half will likely take place in factories.
At this point, China's manufacturing capabilities accumulated over the past few decades have suddenly taken on a whole new strategic significance.
What Capital Truly Needs to Reprice
The Key Is 'Industrialization Capability'
From an investor's perspective, I think there's a bigger change underlying this issue. In the past, tech investors often asked: how far ahead is your technology? This question will still be important in the future, but perhaps another question will also need to be added: how fast can you achieve industrialization?
As AI is reducing the cost of knowledge dissemination, a model that is half a year ahead may soon be caught up by competitors; an algorithm that is a year ahead may not necessarily create a ten-year moat; and many technological innovations are quickly understood and replicated by engineers worldwide after they emerge. Therefore, the things that will be truly difficult to replicate in the future may become increasingly "heavy": supply chains, engineering experience, manufacturing capabilities, industrial clusters, customer networks, application scenarios, and scale advantages. These things may not seem as appealing as a new AI model, but they may form a more solid moat.
Taking BYD as an example, it is very clear. In August 2026, BYD's global sales reached approximately 440,000 vehicles, a year-on-year increase of 17.8%; but what is truly noteworthy is that its overseas sales for the month reached approximately 189,000 vehicles, a year-on-year increase of over 134%. Meanwhile, in the first half of 2026, BYD underwent a significant change for the first time: its overseas revenue surpassed its domestic revenue in China.
This means Chinese tech manufacturers are entering the next stage. In the past, it was Chinese production and global brands selling; later it became Chinese brands, Chinese production, and selling in the Chinese market; now it's evolving into Chinese brands, Chinese technology, Chinese supply chains, and selling in the global market. This is where "Made in China" truly deserves a revaluation by capital.
But We Must Also See:
China Didn't Actually Win
All Tech Competition
It's essential to remain clear-headed. "Foreigners coming to China to learn about technology" does not equate to "China has completely surpassed the US." If that were the case, this article would be pointless. The US still holds a large number of the most advanced core intellectual property rights, boasts the world's most powerful group of tech companies, a top-notch university system, a risk capital system, and extremely strong basic research capabilities. In the fields of high-end AI chips, basic software, some scientific instruments, original algorithms, and cutting-edge scientific research, China still has obvious weaknesses. Even the people involved in the China tech tour, as reported by Reuters, warned that non-Chinese tech companies still own a large share of the global market, the most advanced intellectual property, and enormous profits.
What's truly worth discussing isn't who has already "won," but rather that the evaluation system for global tech competition is changing. In the past, a country's tech prowess was assessed by looking at its research papers, patents, laboratories, and top companies; in the future, a new metric must be added: the speed of industrialization. Whether a technology is leading-edge remains important, but more and more, its business value ultimately depends on whether it can be turned into a stable, replicable, and scalable real product within a sufficiently short period of time.
The Real Battle Ahead
Could Be 'Tech Time'
If we compress the past century's tech competition into a single view, a very interesting pattern emerges. The first industrial revolution was about machines, the second was about electricity and mass production, the information revolution was about chips and software, the internet revolution was about traffic and network effects, and the AI era may ultimately be about competing for a more abstract resource: time.
Who can train models faster, who can manufacture chips faster, who can build data centers faster, who can produce robots faster, who can reduce costs faster, and who can make a technology cover 100 million users faster - ultimately, the deciding factor for competitive advantage may be the same variable: how long it takes for a technology to go from emergence to large-scale commercial application.
In the past, China's greatest advantage was cost, but in the future, the advantage that truly deserves attention may gradually shift to speed. The speed in question is not just about working hard, but rather the result of an entire industrial system working together in tandem.
A country that can continuously shorten the cycle from identifying problems to developing products and then achieving scale is essentially competing for a new factor of production - technological time. Whoever can compress technological time more effectively will be more likely to gain a scale advantage in the next round of industrial competition.
Why the World Is Turning to China
So when re-examining the foreign investors and entrepreneurs flocking to Shenzhen, Hangzhou, and Shanghai, it becomes apparent that what they truly want to see may not be a particular robot or new energy vehicle, or even a star company. What they really want to understand is: why do these things happen so densely and simultaneously in China? Why can robots be mass-produced quickly, why can new energy vehicles iterate rapidly, why can drones form a complete industrial chain, why can a hardware entrepreneur find suppliers so quickly, and why can new technology enter the real market for testing so swiftly?
Behind these issues lies a single answer: China is evolving from the "world's factory" into a massive industrial laboratory.
In the past, the rest of the world would send design specifications to China, and China would be responsible for manufacturing them. However, a new possibility is emerging: research and development, design, engineering, manufacturing, marketing, and iteration are starting to form a closed loop in the same place. Once this closed loop is truly formed, the significance will be completely different, because manufacturing is no longer just the last step of innovation, but manufacturing itself begins to participate in innovation.
The world is re-examining China today, and what it is truly studying is not the success of a particular product, but rather the efficient mechanism that integrates technology, industry, and market.
Twenty years ago, Chinese entrepreneurs went to Silicon Valley to learn how to create the future. Today, more and more people are coming to China to study how to turn the future into reality faster. This may be the real era change worth noting behind the trend of "foreigners coming to China to learn about technology".
In the future, a truly powerful tech nation will not only have laboratories, but must also possess two capabilities: the ability to achieve original innovation from 0 to 1, and the ability to achieve industrialization from 1 to 1 million. The former determines the height of technology, while the latter determines the speed at which technology changes the world. Ultimately, capital seeks the intersection of these two capabilities.
So perhaps in the future, when we assess the competitiveness of a nation, a city, or even a tech company, we should add a new metric: how much time does it take from the birth of an idea, to the first product rolling off the line, to the one-millionth unit being sold? Whoever can keep compressing that timeline may hold the scarcest resource in the next round of tech competition — the ability to turn the future into reality ahead of schedule.