Tech giants start by disassembling systems

Chapter 77 Words from the Heart

Su Chen did not write an open letter. Nor did he use the names Tianying and Haotai.

He wrote a long article titled "Why can't we buy flight controllers?"

The article begins with a reenactment of a scene. In the winter of 2005, on his first day at Hongyuan, he went to the workshop to see the first prototype. Zhou Ming brought over a machine with an exposed, dark brown circuit board, on which sat an 18-pin chip. Zhou Ming said, "This is a ready-made flight control board we bought. As a result, three out of ten of these failed as soon as they were powered on. We complained to the solution provider, but they said it wasn't a quality issue, but rather that our operating environment was incompatible. How was the operating environment incompatible? Nobody could explain it clearly."

Then he wrote down his first decision—to make his own flight controller.

Nobody was optimistic. At the time, the company only had 1.2 million yuan and could only hire one programmer. Writing a flight control system from scratch is a task that would take a senior engineer three years in the coding world. They worked straight through, spending three and a half months creating the first version of the F1 flight control system. It was terrible—it went out of control at 20 meters, and would collapse in a little wind. But it was something they wrote themselves. They could modify it countless times.

Change each parameter one by one. Adjust each scene one by one.

When the F2 was launched, its flight control system had already undergone 44 iterations. By the time the F3 was launched, it was already past its 5th generation. The amount of flight control code in the F4 was more than forty times that of the first generation.

"I'll remember these numbers for the rest of my life. Because with each iteration, a small gap related to the user, the scenario, and the environment is filled. With each gap filled, the product becomes more persistent. This kind of thing can't be bought with solutions. Purchased solutions are someone else's parameters, someone else's iterations, someone else's patching of gaps. How does it relate to your own product's users, scenarios, and usage environment? It's irrelevant. What you buy is always someone else's bones."

In the middle of the article, Su Chen wrote three short stories based on real-life scenarios.

The first issue was rice paddies in Chiang Mai, Thailand. A plant protection customer using an early version of Hongyuan's flight controller discovered that the flight altitude would drift under certain twilight angles of sunlight. It was later found that the barometer was affected by the thermal expansion of moisture from the rice paddies. Correcting this issue took two version updates. The second issue was a misjudgment of sudden changes in sunlight in palm plantations in Indonesia. The third issue was electromagnetic filtering under high-voltage power lines in Shaoguan.

"Can you find these three questions in the documentation of the flight control solution you purchased? No. Because these questions simply don't appear on the solution provider's test list. Solution providers serve most scenarios. They can't handle users in those somewhat unusual, different, and more complex usage environments. Leading companies with deep expertise in a particular industry will find that some customers are precisely in these complex corners."

"Let's put it another way—if all the clients in this industry were in standard scenarios, then indeed buying a solution would suffice. But our clients include workers in palm plantations, inspectors under high-voltage lines, pesticide spraying teams in rice paddies at midday, and maintenance workers on photovoltaic power stations. Each of them operates in a unique environment. Technicians, the door to this industry has no standard scenarios. Finding clients in this industry is like trying to open a bunch of different locks with a universal key—most locks won't open."

Su Chen wrote a paragraph at the end of the article.

"I've seen investors say they don't care about the long term. I understand. Capital has its own on/off switch and exit point; that's its right. But an industry as a whole can't be led by short-term capital sentiment. An industry needs people willing to work on things for three, five, or ten years. Platform companies are built this way, not bought."

"Flight control systems can't be bought. What you buy is a solution, not a capability. The difference isn't huge, but it's enough to determine how far a company can go."

The article was published on Hongyuan's official WeChat account at 5:18 a.m.

At 7:00 AM, the number of views exceeded 10,000. By 9:00 AM, it had surpassed 100,000. At noon, it exceeded 500,000. At 2:00 PM, 36Kr reprinted the entire article. At 5:00 PM, the electronic version of Caijing magazine excerpted key paragraphs into a separate article. In the evening, a link to the reprinted article appeared on the homepage of China's largest business forum.

The comments section was like boiling water.

"Su Chen's article didn't mention Tianying, Haotai, or Xia Kanghao even once, but every single sentence was like a blood-soaked Longquan sword."

"The analogy of the key and lock is brilliant. Anyone who's been in the industry for ten years will feel a pang of sympathy after reading it."

"This is a true tech entrepreneur. He doesn't take attacks on anyone; he simply speaks the truth about the industry."

Two days later, something even more unexpected happened. One of the most influential financial newspapers in China adapted Su Chen's article into a commentary titled "Capital and Technology: Two Narratives of Industry Maturity," removing all the sentences that specifically named Haotai, but retaining the core key-and-lock metaphor and the discussion of platform companies. It was featured on the front page of the newspaper's business section.

Su Chen's name has gone beyond the industry.

That same afternoon, he received a WeChat message from Professor Chen Hongyuan.

"You've come to Guangzhou for a while. Someone wants to introduce you to someone. This person is a professor here."

Su Chen replied with a simple "Okay." He didn't ask who it was. Anyone Chen Hongyuan specifically summoned must be worth meeting.

He arrived in Guangzhou on Saturday morning. Chen Hongyuan had made a reservation at a Chaozhou restaurant in Tianhe City. When he pushed open the door, there were already two people sitting inside. Chen Hongyuan. And someone Su Chen knew but had never met.

Wu Zheng.

The former CTO of Skyhawk. He left the company a week before the acquisition was completed.

Su Chen's expression remained unchanged, but he was slightly startled.

"President Su," Wu Zheng stood up to shake hands, his voice steady, "I read your article. I read it three times."

The three sat down. Chen Hongyuan spoke first.

"Let me make the connection. Professor Wu is my undergraduate classmate from South China University of Technology. We met again last year at an industry forum. I won't interrupt over this tea today; Professor Wu can speak for himself."

He poured tea for the two of them and then sat down to one side.

Wu Zheng held the teacup in his hand and remained silent for a while.

"I founded Tianying 19 years ago with Lu Weimin. For the first two years, we wrote our own flight control system. At that time, Tianying only had three engineers, and I and two assistants were coding in a residential house. But after the price war started, we switched to external solutions. Lu Weimin's reason was—we couldn't afford the time cost of developing our own system."

He glanced up and stared at Su Chen.

"I agreed back then because I considered drones as a product. It wasn't until a year and a half ago, when I brought up the issue of developing our own flight control system again, that I realized—a drone isn't a product. It's a platform. And a platform must be self-developed. You already know what I'm saying. I won't repeat myself."

He fed him a sip of tea.

"When I left, I had a twelve-person self-developed flight control team. I recruited them one by one over the past twenty-one years. Most of them were young people I recruited from three universities in China. After Haotai took over—Haotai gave me three options. One was to continue as CTO but listen to them. Two was to become a consultant and leave with a five million yuan advance payment. Three was to leave automatically, using the stock options earned from the patent application development to be converted into cash."

"I chose the third option. I thought my life would just stay like this forever—going home to send my kids to school, and when I felt like it, I'd try my hand at something like the 'Three-Year Plan.' My daughter would definitely say I wouldn't be able to use such a huge amount of money to learn from others. The day before yesterday afternoon, I saw your article while searching for knowledge on this tech forum."

Wu Zheng's voice lowered.

"I couldn't help but feel a chill when I read the part about the keys and locks. Skyhawk has made too much money over the years, and in the end, they didn't use it on the keys."

He looked up at Su Chen.

"I've been thinking about something lately. Of the twelve disciples of Tianying, eight are people I personally trained from the beginning. I left a letter for Haotai, recommending they stay on. But Haotai won't keep them. This is a difficult situation. I want to ask you—do you need them?"

Su Chen remained silent for a while.

He wasn't considering whether he needed it or not. He was considering what kind of acceptance would be most appropriate.

"need."

"But I'm sorry—I can't accept them all at once. Everyone has to go through our regular interview process. I can have HR contact them tonight. Those who pass our interview will stay, and those who don't will not."

Wu Zheng nodded.

"That's how it should be."

He paused for a moment, then asked, "And what about myself?"

Su Chen smiled.

"Mr. Wu, I won't be interviewing you. If you're willing to come, we'll reserve a position for you as the chief flight control architect. Mr. Zhou will send you the details of the basic framework and stock options tomorrow. But there's one condition."

"you say."

"You don't need to bring Skyhawks. If they want to come, follow the proper procedures. You're here to create, not to seek revenge."

Wu Zheng looked at him for a few seconds. Then he reached out his hand.

Chen Hongyuan nodded and picked up the teapot to add tea to the three cups.

That evening, Su Chen was driving back to Shenzhen. As darkness fell, he called Zhou Ming and told him about the newly created position of Chief Flight Control Architect. Zhou Ming didn't ask many questions, only saying, "It's a good thing we were able to get him here."

Then Su Chen hung up the phone.

He drove on the highway for a long time. He wasn't thinking about Wu Zheng, Haotai, or that article. He was thinking about Zhou Ming.

Zhou Ming, who showed him a piece of crystalline silicon board in the leaky workshop ten years ago. Zhou Ming was the only one in that small factory that was on the verge of bankruptcy who was still willing to believe in him.

That initial trust is the key. The red line isn't just a red line. The halo isn't just a halo. The real foundation is a group of people willing to write a single line of code for years on one thing.

He stepped on the gas.