Preface:

A century ago, red-headed ships carried porcelain, tea, and sundries.

Today, the ships are invisible, and the cargo is almost weightless. They are algorithms, models, robots, and intellectual property.

Author | Fang Wensan

Image Source | Internet

Reaching the Global Summit: Chaoshan Businesspeople Take Their Goods to the Upstream of AI

Looking at Yang Zhilin's background, you'll notice a distinct Chaoshan imprint, though the cargo has changed entirely.

Beyond technology, what's even more intriguing is the company's business approach, which is steeped in traditional Chaoshan business practices.

In March 2026, Moonshot AI's annualized revenue surpassed $100 million for the first time, $200 million in May, and stabilized at $300 million by mid-June. API revenue accounted for over 70% of this three-fold growth in three months.

K3's API pricing is about four times higher than the previous generation, charging 20 yuan per million input tokens and 100 yuan per million output tokens, directly entering the pricing range of mainstream U.S. models.

The capital market has voted with its wallet. During the Series C funding in December 2025, the company was valued at $4.3 billion.

On July 29, an over $3.5 billion Series F round was announced, with subscriptions exceeding the target by more than threefold, leading to an early closure of the funding round. The post-money valuation reached $35 billion. The Pre-IPO round, originally scheduled for August, was moved up overnight, with a pre-money valuation set at $50 billion. The Hong Kong Stock Exchange listing is expected to happen within six months.

Placing this trajectory within the history of Chaoshan business adds profound meaning.

The first generation of Chaoshan businesspeople sold rice, sugar, and sundries. Xie Yichu took vegetable seeds to Southeast Asia and founded Chia Tai.

The second generation sold real estate and electronics, with Li Ka-shing and Huang Guangyu leading the way.

The third generation, represented by Pony Ma, sold connections, putting 1.4 billion people into a single chat box.

Now, with Yang Zhilin, Chaoshan people are selling intelligence itself.

The goods are getting lighter, but the pricing power is getting stronger—exactly what this business group has been obsessed with for centuries.

The Hard Tech Cycle Perfectly Matches the Core Traits of Chaoshan Businesspeople

In the first half of the internet era, Chaoshan businesspeople seemed somewhat left out.

It was an era of subsidies for traffic and burning money for scale, emphasizing blitzkrieg and winner-takes-all strategies. The traditional Chaoshan approach, however, focused on cash is king, sticking to what you know, and making money quietly.

Except for Tencent, Chaoshan people's presence in that frenzy was far from commensurate with their wealth.

The hard tech era has flipped the rules. Chips, electronic materials, and large models are all long-term, high-stakes businesses with decade-long investments and returns measured in patience, earning money from technological compounding. This rhythm seems tailor-made for Chaoshan businesspeople.

Patience is the first trait. Wang Laichun could stay on an assembly line for 10 years, and Tri-ring could persevere with a ceramic material for 55 years, with R&D investment accounting for 7.25% of revenue. This ability to sit tight was a burden in the traffic era but a moat in the hard tech era.

Financial acumen is the second trait. Luxshare's gross margin is only around 10%, but Wang Laichun still managed to build a hundred-billion-dollar empire through Ultimate turnover (extreme turnover) and over 25.2 billion yuan in consecutive acquisitions in recent years.

When Moonshot AI had over 10 billion yuan in cash on hand by the end of 2025, Yang Zhilin's internal letter was calm, stating that the company was in no hurry to go public and could still raise substantial funds in the primary market. This obsession with cash flow would make Chaoshan people smile knowingly.

Networking is the third trait. The clan and chamber of commerce networks of Chaoshan businesspeople are world-renowned. Today, they have transformed into supply chain collaboration networks and alumni talent networks. The executive team at Lixiang Innovation all hail from the Luxshare system, with the sister's chairman position and the brother's founder status appearing in the same prospectus.

When the industrial cycle shifted from traffic monetization to technological compounding, the old traits of Chaoshan businesspeople became levers.

K3 Reveals a Decentralized Tech Innovation Network

The Chaoshan tech innovation cluster brought to light by K3 is not just a random collection of names.

Yang Zhilin is betting on foundational models, Huang Yuanhao founded Orbbec, entering 3D vision and robotics perception, and Wang Shiquan founded Flexiv, researching adaptive robots.

Qiu Chunxin turned his doctoral research on environmental perception into LiDAR technology, while Xiao Jianxiong moved from Princeton to L4 autonomous driving.

Together, these paths cover different joints of machine intelligence, from models, vision, and sensors to robotic motion and automotive decision-making.

The distance between these entrepreneurs and traditional Chaoshan businesspeople is first evident in the nature of their businesses.

In Moonshot AI's early days, there were no stores, factories, or mature cash flows. What was continuously consumed were computing power, data, and top talent.

Technological judgment often preceded revenue, and commercial returns came more slowly. Such entrepreneurship was hard to measure directly with familiar inventory, channel, and turnover experiences.

However, the instinct to turn opportunities into businesses did not disappear; it just changed tools.

When Huang Yuanhao was early in industrial-grade 3D sensing, he abandoned a product route with millions already invested to pivot to a larger consumer market while continuing to Attack chip (tackle chips), optics, and mass production.

By 2025, Orbbec's revenue reached 941 million yuan, up 66.66% year-on-year, with a net profit of 128 million yuan, marking its first annual profit since going public.

RoboSense's path was closer to manufacturing. Qiu Chunxin brought his doctoral research out of the lab, through chips, optics, algorithms, automotive certification, and mass production.

In 2025, the company's revenue was approximately 1.94 billion yuan, with LiDAR sales of about 912,000 units, up 67.6% year-on-year, including about 303,000 units in the robotics sector. The fourth quarter saw its first quarterly profit.

Behind these numbers, there are no easy genius curves. Papers must be dissected into the supply chain, precision must face cost scrutiny, and technology must be repeatedly calibrated between yield, delivery, and customer needs.

This explains the truly scarce ability of the new generation of Chaoshan tech entrepreneurs. They must understand both papers and customers, labs and production lines. Being good at business today is more like a bilingual ability.

The most durable assets of traditional Chaoshan businesspeople were never just goods but the ability to organize trust, raise funds, and find buyers.

In the past, the qiaopi network used family and local ties to reduce cross-border transaction costs. Today, cutting-edge tech startups face companies with no factory inventory to mortgage but continuously consume computing power and talent, changing the trust mechanism.

From Yang Zhilin to Huang Yuanhao and Qiu Chunxin, this generation of entrepreneurs has Continued (carried on) the Chaoshan sensitivity to market acumen and operational efficiency while having to learn slower-paced businesses.

Foundational models, sensors, and robots often burn money for years, with technological judgment preceding cash flow, and scale manufacturing grinding ideals into greasy hands.

The commercialization impulse that traditional business groups excel at is still present, but it is now recalibrated by scientific training, engineering discipline, and institutional capital.

Clan networks recede to the background, while education and capital take over the interface.

Gongfu tea emphasizes precision—a second's difference in heat control (heat) changes the tea's flavor.

This reverence for subtlety translates into Olympiad classes for children and Shantou University, founded with donations from Li Ka-shing, in cities.

Yang Zhilin's trajectory from Jinshan High School to Tsinghua and then CMU proves that this business group's daredevil spirit has upgraded—calculating before acting, then betting big.

Huang Yuanhao followed the same high school path through Peking University, the National University of Singapore, City University of Hong Kong, and the MIT SMART Research Center.

Xiao Jianxiong grew up in Chaozhou and later entered the frontier of autonomous driving via the Hong Kong University of Science and Technology, MIT, and Princeton.

Attributing the emergence of these founders solely to regional character is both facile and obscures generations of investment. Chaoshan society has long viewed education as a durable good that expands family options.

According to incomplete statistics cited by the Overseas Chinese Affairs Office of the State Council in 2008, from 1987 to 1996, over 10,000 overseas Chaoshan people donated 1.5 billion yuan to establish schools in the Chaoshan region. By 2008, overseas Chaoshan people had donated to build, expand, or assist in building over 2,000 schools.

Homeland didn't directly complete their entrepreneurship but gave them the ability to leave it.

The fellow townspeople network didn't disband but receded from the core transaction mechanism to become a connector. The Chaoshan Entrepreneurs Association's website shows it has over 1,000 members, with seven branches in Shantou, Shenzhen-Hong Kong, Guangzhou, Beijing, Shanghai, Singapore, and North America, and 12 bases.

These nodes provide talent matching, experience exchange, and capital access. Technological viability is still judged by peer reviews, customer orders, and market competition. Hometown connections create encounters; professional ability determines how far cooperation goes.

Chen Xiang, founder of Turing Future Capital, recalled Flexiv's early industrialization discussion, which happened during a weekend Chaoshan dinner in the U.S., shifting from rice noodles to robots, followed by capital connecting with the Stanford team.

This entrepreneurial anecdote still carries the warmth of a acquaintance society (acquaintance society), but the investment target was adaptive robots, with subsequent involvement from institutional capital like Meituan, Yunfeng, and Gaorong. The dinner table retained the old business group's entry point, but due diligence, patents, and products took over the rest.

From "Good at Studying" to "Daring to Entrepreneur": There's Still a Door in Between

Of course, focusing solely on education doesn't automatically create tech entrepreneurs. Many regions value education, and there are many top students.

What's more unique about Chaoshan is the seemingly smooth transition from education to business.

Some researchers choose to stay in labs, some engineers join large companies, but many tech talents with a Chaoshan background are willing to take their technology to market.

Huang Yuanhao of Orbbec is representative. Even in research, he thought about how technology could become a product. Zhang Junbin, founder of Narwal Robotics, directly linked his entrepreneurial impulse to his upbringing.

What truly carries on here is perhaps not some "business talent" but a lower psychological barrier to commerce.

For many families, entrepreneurship is not a distant career.

Even if children later go to Tsinghua, Stanford, or Carnegie Mellon to study AI, robotics, and computer vision, they may not be unfamiliar with "making something and selling it."

Thus, an interesting combination emerges: the previous generation provided business intuition, the education system added technical skills, and venture capital supplied the capital leverage.

When these three collide, entrepreneurship is no longer just opening a store or taking over a factory.

It can be a LiDAR system, a robotic control system, or a 2.8 trillion-parameter large model.

The business spirit hasn't faded; it's just that the average transaction value has become a bit intimidating.

Money doesn't directly become chips, robots, or large models.

It first becomes classrooms, labs, scholarships, and university admission letters, reappearing in industries as engineers and entrepreneurs twenty years later.

Industry research tends to focus on funding rounds, valuations, and IPO bells but rarely looks back twenty years.

In fact, the true starting point of a hard tech company is often when the founder first enters a better school, participates in a competition, or gets the chance to continue studying in their teens.

Behind the so-called new tech elite often stand generations of choices.

Ending:

Hometown still opens the door, but whether you get to the table increasingly depends on professional ability.

This is a crucial step in the modernization of traditional business groups.

Personal connections haven't been eliminated by modern commerce; modern commerce has just redefined their limits.

Partial Sources: Phoenix Weekly: "Half of China's Tech Tycoons Come from Chaoshan?", RoboSense: "Company Introduction 202607", Economic Observer: "Dialogue with Narwal Robotics Founder Zhang Junbin: Ten Years of Entrepreneurship, I Still Dare Not Call Myself an Entrepreneur"