Hello, everyone. Welcome to tonight's speech.
This is my fourth annual speech. Many people are here, and many more are watching online. Thank you all.
Two months ago, I attended Wuhan University's graduation ceremony.
This year marks its 130th anniversary, and more than 17,000 people attended. Being invited back as an alumnus was the highest honor. Facing the faculty and students, I was overwhelmed as old memories came flooding back.
01
Memories of Wuhan University
1.1 Where the Dream Began
In 1987, thirty-six years ago, I entered Wuhan University's computer science department.
In the library, I read Fire in the Valley and found my lifelong dream.
The book filled me with such excitement that I could not sleep. Beneath the bright stars, I walked around the university track all night. A fire burned inside me: I would build a great company. My dream had been ignited.
But how could a first-year student from a small county town, with no skills or resources, possibly build a great company?
I had no idea how to realize such an outrageous dream, so I stopped thinking about it and focused on studying first.
1.2 First Goal: Complete Every Credit in Two Years
How could I make my studies extraordinary? I set an audacious goal: complete all four years of credits in two years. The credit system had only just begun, and few people imagined such a thing.
How did I do it? With limited time today, I will explain how I overcame three difficulties. If you are at university, this may help you.
First, how should you choose courses?
Taking twice the classes, doing twice the homework, and sitting twice the exams was university on hell mode. Hard work was manageable; choosing courses was the real problem. After puzzling over it, I decided to ask someone.
I was rather shy, but I forced myself to knock on the doors of third- and fourth-year students and look for people from my hometown.
It worked wonderfully. Several senior students eagerly explained the entire curriculum for hours: which courses were easy or hard, which had prerequisites, and even stories about the teachers. They also gave me their textbooks, study guides, and notes.
After that conversation, everything suddenly became clear. I basically understood how to navigate university.
It taught me something important: when you encounter a problem, immediately ask someone who understands it. People often think they must solve problems alone. In reality, whatever your problem is, someone else has probably faced and solved it, perhaps with a standard answer. You only need to ask.
Second, how do you teach yourself?
Twice the classes meant learning many subjects independently. Our computer department grew out of mathematics and was effectively half a mathematics department, with extremely difficult courses. Students often call advanced mathematics hard, but we studied specialized mathematics courses that were much harder.
When studying alone, I often became stuck on things I could not understand. What could I do? I simply forced myself to keep reading until understanding finally came.
If something was incomprehensible, I skipped it and read onward like a novel, from beginning to end. After repeating the book several times, I could generally understand it.
I later understood why: knowledge is not entirely linear but mostly a network. Ideas do not always have an absolute sequence. Skipping something earlier does not prevent learning what follows, and later knowledge can make earlier material easier.
Let me add this: society changes constantly, and what school teaches is never enough. Only lifelong learning can prepare us for future challenges, and lifelong learning depends on self-education.
Third, how did I handle attendance checks?
Taking twice the courses inevitably created timetable conflicts. Missing class meant risking an attendance check, a serious matter at university. What could I do?
Teachers checked attendance because they feared students were not studying seriously. I reasoned that if I made an impression and showed I was not lazy, it should be fine.
So I sat in the front row where teachers could see me, raised my hand first when they asked questions, and approached them during breaks whenever I needed help.
Through these efforts, most teachers formed a good impression of me, and occasional absences ceased to matter.
That was how I overcame every obstacle and completed all four years of credits.
Wuhan University's teachers were exceptionally kind and gave me tremendous guidance. Remembering them still fills me with gratitude.
1.3 Second Goal: Become an Excellent Programmer
I studied computing. Earning credits was not enough; writing excellent programs was what mattered.
So my second goal was to become an excellent programmer.
Computers were scarce more than thirty years ago. In my first year, our entire class shared one Motorola 68000 minicomputer. It was far weaker than a phone from ten years ago, served over twenty terminals, and was painfully slow.
Even then, access was rationed. Each person received only two tickets a week, each valid for two hours.
With only a few hours weekly, becoming good at programming was extremely difficult.
Whenever possible, I queued outside the computer room hoping someone would not appear. Otherwise, I watched classmates work. If someone needed help, I advised them and felt a great sense of achievement.
To use computer time efficiently, I wrote my code on paper beforehand and entered it immediately. To type faster, I traced a keyboard onto printing paper and taped it to my desk. During every class, I secretly practiced by tapping the desk. Soon classmates copied me, and teachers were baffled by the constant tapping.
When no computer was available, I wrote programs on paper. Textbook examples were usually excellent, so I closed the book and rewrote each one my own way until I matched its quality.
My first programming course was Pascal. In my second year, a classmate told me the teacher had included my assignment in a new edition of the textbook. I was thrilled that my work had been recognized.
I later helped with projects in a professor's laboratory. After countless nights at its computers, I gradually became the technical expert among my classmates.
A few years ago, I saw assembly code on Zhihu that I had written in my second year. It was a DOS memory-management tool called RI. It became popular in China, and I released all its source code.
People still discuss it more than thirty years later. As a programmer, that makes me proud.
1.4 Third Goal: Publish an Academic Paper
My third challenge was publishing in a leading academic journal, a difficult goal for an undergraduate.
Professor Zhang Dexiang suggested organizing my research on computer viruses into a paper.
China had recently experienced its first major computer-virus outbreak. I had studied the subject extensively and even written one of China's early antivirus programs.
I was unsure whether it counted as academic work, but followed Professor Zhang's advice and boldly wrote the paper.
Papers were handwritten then. I filled dozens of manuscript pages and asked friends to review them. One said my handwriting was poor and that attractive handwriting improved the odds. He volunteered to spend three days copying the entire paper neatly.
I mailed it to the authoritative journal Computer Research and Development. About six months later, I learned it had been accepted.
I was ecstatic. My first submission had succeeded. More surprisingly, formal publication did not come until August 1992, two years later. Apparently, that was normal.
This always reminds me that daring to imagine and act matters. Young people need boldness. Many things are less complicated than they seem; you discover that by trying.
1.5 Dreams and Growth
That was my student life at Wuhan University.
Many people have similar youthful experiences: a book, film, or person ignites a dream.
I was fortunate to take my dream seriously, break it into achievable goals, and give everything to each one.
Dreams have immense power. I found ways to turn many impossibilities into possibilities.
That experience built a solid foundation for my life.
02
The Entrepreneurial Journey
After completing my credits, I had time for interesting projects.
2.1 My First Commercial Software
During the May Day holiday of my second year, I met Wang Quanguo on Wuhan's electronics street. A highly skilled alumnus four years older than me, he had remained at the university as a teacher and worked for a university-run networking company.
I often went there to use their newer computers. I programmed on demonstration machines and helped customers with questions. The teachers thought I worked there, but I received no pay, only lunch.
Wang and I became friends. He respected my ability, and that summer we decided to write encryption software together.
Why encryption? It was technically difficult, and we wanted to tackle something hard.
After work, we lowered the shutters and had the whole office to ourselves.
I used two computers, one for coding and one for testing. It felt like the joke: when we get rich, we will buy two bowls of soy milk, drink one, and throw away the other. It was wonderful.
We were very happy, often working until three or four in the morning and sleeping on two long sofas. However late we slept, we rose on time because the office opened at nine.
After fifteen days and nights of work, both of us had lost considerable weight, but we finished on schedule.
My first commercial program was BITLOK. It attracted attention among encryption tools and became popular with programmers and developers. Priced above RMB 2,000, it sold well. We upgraded it over several generations and sold it for six or seven years, earning substantial money.
Success on my first attempt was exciting, but I knew it depended on Wang. I was a second-year student under twenty; he was already a teacher with far greater maturity, experience, and resources. However capable one person is, that ability is limited. Success is easier with complementary partners.
Wang and I became lifelong friends. Thirty years later, we took a back-to-back Kappa-style photograph to commemorate our first collaboration.
2.2 My First Startup
By August 1990, after more than a year on Wuhan's electronics street, I had gained a modest reputation.
An older student invited me to start a business. I eagerly agreed, thrilled to have a chance to realize my ambitions.
We rented room 103 in a hotel and named the company Sansè. The four founders were friends and felt awkward discussing equity, so we divided it equally: 25 percent each.
Everyone wanted to accomplish something great. We worked furiously, often all night. But a fatal question soon appeared: who was in charge?
Discussing everything among four people was inefficient and produced constant arguments. We elected a chairman to decide, but held several elections within months. Everyone took a turn except me; I was too young to command authority. A classmate recently showed me my old card: technical director.
We sold computers, wrote software, offered typing and copying, and did any business available in pursuit of our dream.
Reality was harsh. We made a little money initially, but as the business expanded, profits became harder to earn.
During evening study, the other directors often pulled me out for meetings. None of us had experience, so naturally we found no good solutions.
Eventually, we could not pay salaries or even afford food. It was agonizing.
One director boasted that although he was bad at business, he was good at mahjong. We sent him to play with the cafeteria cooks, and he won many meal tickets, sustaining us for a while. Eventually he won so many that they stopped playing.
The company finally collapsed. Without even a farewell meal, we liquidated it. I received a 286 computer and carried it back to campus.
Walking across campus the next morning, the sunshine seemed brilliant. At last, the chaos was over.
That ended my first startup. Repeated reviews showed how commercially naive and inexperienced we were. We never established a leader or decided what business to pursue; we simply did whatever appeared and began in confusion.
Sansè's failure was a bucket of cold water that awakened me. Building a company was extraordinarily hard. Technology alone was insufficient; I had much to learn, especially business fundamentals.
Years later, I met a student startup with four equal shareholders. Every business card said chairman. Four chairmen! I asked who was actually responsible, and they said they were classmates who discussed everything together.
I suddenly saw my younger self and Sansè. I doubted their journey would be smooth and could only wish them luck.
2.3 Entrepreneurship and Growth
After graduation, Qiu Bojun invited me to Kingsoft. This year marks Kingsoft's thirty-fifth anniversary. It was where I truly grew, alongside my elder colleagues Qiu Bojun and Zhang Xuanlong and brothers including Zou Tao, Zhang Qingyuan, and Guo Weiwei.
At Kingsoft, we often say that one person may go faster, but only a group can go farther. That is the most important lesson from its thirty-five years.
In April 2010, I set out again. Carrying my youthful dream, I founded Xiaomi with a group of partners. I will not repeat Xiaomi's story here.
Looking back on more than thirty years of entrepreneurship, my journey has been one of continually pursuing dreams and growing.
03
The Pursuit of the Premium Market
Speaking of growth, our pursuit of the premium market over the past three years has been my most painful yet rewarding experience of the last decade.
Xiaomi's premium journey began with Mi 10.
It started smoothly, but Mi 11 suffered a major setback. We worked extremely hard on Mi 12, yet it still fell short. No one expected premium products to be so difficult or require such investment. Discouragement spread through the company.
3.1 The Fiercest Debate in Xiaomi's History
I spoke with everyone at length. Many views were so blunt that they surprised me.
Some asked whether, since premium products were so difficult, we could stop making them.
Others said Xiaomi's brand image was fixed. Could it become premium, or should we create another brand?
This was before last year's Spring Festival. I had assumed everyone supported the premium strategy and never imagined such disagreement. I felt heavy and anxious throughout the holiday.
On the first working day afterward, I urgently convened the largest special strategy meeting in Xiaomi's history, attended by every department's general manager.
I encouraged open discussion. The mood was friendly after the holiday, but debate raged from noon until evening. I repeatedly argued that our founding dream was to make Xiaomi a globally renowned technology brand. However difficult it was, we had to use the Xiaomi name for premium products.
Consumer electronics is fiercely competitive. Only premium products can force technological breakthroughs and secure room for future survival and development.
We finally agreed that moving upmarket was essential to development and a battle for survival. We could not waver and had to persevere. Premiumization formally became a group strategy.
3.2 Exploring Imaging
Let me use imaging to explain our exploration.
We have always valued imaging and established a separate camera department in 2017.
Competitors used DXO rankings to demonstrate imaging quality and technical strength. We invested heavily and set a simple, aggressive goal: surpass them and rank first worldwide on DXO.
We sounded the charge.
We entered the top three, then ranked second, then tied for first. To take sole possession of first place, the team spent over two more years preparing Mi 11 Ultra, with a custom sensor, maximum specifications, and intensive tuning.
The day before launch, we received the news: number one worldwide on DXO.
The camera department erupted. After years of effort, the long-awaited success had arrived. Unable to wait for the formal launch, they held a grand celebration that evening.
Xiaomi Science and Technology Park lit up with the words "DXO No. 1." We were immersed in joy.
Yet users fiercely criticized the product on its first day. With such powerful imaging and a global number-one DXO rating, why were they dissatisfied? Painful reflection taught us that maximum specifications and rankings did not necessarily mean good photographs or satisfied users.
How could we make photographs users loved? My university experience suggested asking someone who understood.
3.3 Xiaomi's Partnership with Leica
I proposed speaking with Leica.
Colleagues told me Leica had approached us twice and been rejected by different teams.
What? We had rejected them twice?
After investigating, I learned it was not arrogance. Colleagues were conflicted: Leica had worked successfully with another brand for five years. Could we surpass that? If not, would we become an industry joke?
I disagreed. If Leica was the world's best teacher and we studied earnestly, our ability would certainly grow. Every photographer I asked called Leica a legend with unmatched standing and expertise.
We resolved to sign formally with Leica.
3.4 Photography Classes
Our first major step after signing was unexpected: we opened photography classes.
I knew that only by understanding and loving photography could we truly master imaging. We invited Leica to teach at Xiaomi Science and Technology Park and required every core team member to attend.
We have now held six Leica master classes. I studied more seriously than at university and missed none, partly because I was class monitor and embarrassed to request leave. It was demanding: everyone was busy, yet weekends brought classes and, worse, homework.
Every assignment drove us crazy. Once, rushing to finish, I went to the campus Starbucks early to photograph colleagues. Many had no idea what I was doing and fled before finishing their coffee.
Once planted, photography culture quickly took root. The company organized training and employee exhibitions, and participation was enthusiastic.
We also befriended many professional photographers and held a major exhibition on the Great Wall. Take a look.
3.5 Two Leica Looks and the Photography Kit
As our understanding grew through Leica, we firmly chose humanistic photography and resolved to create an authentic Leica look on phones.
This was difficult. Mainstream phone images use vivid colors that look attractive at first glance despite being somewhat artificial. Leica's humanistic tones are less immediately flattering. What if ordinary users rejected them?
After repeated discussion, we proposed preserving the classic Leica look while adding "Leica Vibrant" for mainstream users. Leica was surprised that Xiaomi would double its development investment for one phone's imaging and gladly agreed.
Thus the two Leica photographic styles were born.
The first phone with both, Xiaomi 12S Ultra, performed extremely well after launch and was loved by ordinary users and photographers alike.
One photographer immediately raised his thumb after seeing samples. He never imagined a phone could create such texture and said, "This phone is incredible. This is real Leica." We made the line an advertisement.
There is another story about the photography kit.
It became so popular that scalpers raised its RMB 799 launch price to RMB 1,400. Yet I had killed the project twice. Only after studying photography did I understand enthusiasts and support its third proposal immediately. The same product produced a completely different decision after my understanding changed.
I especially thank Xiaomi's product team for persisting. You were right.
Two generations of Ultra products earned Xiaomi imaging an excellent reputation among both consumers and professional photographers.
3.6 Xiaomi 13
This exploration and change is happening throughout Xiaomi.
Our understanding of premium products steadily deepened, and their performance kept improving.
Xiaomi 13 won excellent reviews and sales, but nearly died twice during development. Let me tell you the story.
The first time was at the project approval meeting.
After Xiaomi 12 and 12S both missed expectations and suffered serious losses, debate raged over whether Xiaomi 13 should remain compact. More than half opposed it.
A compact flagship is extremely difficult. Users want not merely comfortable size but a complete flagship with no compromises. Fitting every capability and experience into a small body is an enormous engineering challenge. Even if achieved, costs are high and users may not pay, because small phones appear less valuable than large ones.
I understood these concerns but believed feel-in-the-hand was the most important user experience. It cannot appear on a specification sheet and is hard to explain, but users understand instantly when holding the product.
The project narrowly passed.
Our development team fought with its back to the wall. During last year's pandemic, members lived and ate in the office to maintain progress, like a collective military camp.
They delivered. Xiaomi 13 was exceptional, with many experiences far surpassing the iPhone.
It nearly died a second time at the ordering meeting before launch.
Despite its excellence, colleagues lacked confidence. The previous two generations caused heavy losses. Could we withstand losing hundreds of millions or even more than a billion yuan again? Competitors at this price used large screens and impressive specifications. Could we really win?
The meeting was even more oppressive. I could not decide recklessly, so I repeatedly persuaded everyone: succeeding at the high end requires patience, and user acceptance takes time, indeed longer than expected. The market was also highly homogeneous, and users needed something distinctive. Seeing my persistence, everyone finally agreed.
After its troubled journey, the compact flagship Xiaomi 13 finally arrived.
Mi Fans joked, "Buy even-numbered Xiaomi flagships, never odd ones." They would wait for Xiaomi 14 and refuse Xiaomi 13. The joke startled our already fragile nerves.
But the result surprised everyone: both reputation and sales flourished. Mi Fans called it Xiaomi's "fragrant thirteen."
Only after three generations of compact flagships and immense determination, investment, and patience did Xiaomi 13 succeed. It became our signature move from leading specifications to prioritizing experience.
This time, everyone sincerely embraced "experience first."
We summarized three principles for premium product design: no weaknesses; move from leading specifications to experience first; and deeply integrate hardware and software.
With this understanding and consensus, I believe future products will be even better.
3.7 Understanding and Growth
Xiaomi's premium journey took over three years, tens of billions of yuan, and the combined effort of tens of thousands of people. After many twists and turns, we achieved a phased breakthrough. Every small success in this battle was hard won and rested on a change and breakthrough in understanding.
Only breakthroughs in understanding bring genuine growth. That is my greatest lesson of the past decade.
04
People Become Great Through Dreams
And Grow by Persisting in Them
Today I have shared several pivotal moments of growth:
At Wuhan University, I formed a dream and learned to divide it into goals. At Kingsoft, I transformed from programmer to manager. By founding Xiaomi around a vision of changing Chinese manufacturing and an innovative business model, I began turning my dream into reality.
Every experience and transformation was a breakthrough in understanding and a pivotal step in growth.
In recent years, many people have felt anxious and lost.
Amid the torrent of changing times, how can we find inner peace?
My answer is growth.
Only steady, grounded growth gives us the confidence, courage, and resolve to meet every unknown challenge.
Only steady, grounded growth can fill our hearts, put light in our eyes, and keep us strong at every moment.
Life is a marathon. Successes and failures at any single moment matter less than we think.
You will find answers to all of life's difficult questions through growth.
People become great through dreams and grow by persisting in them.
My friends, may your dreams still shine brightly after the passage of time.
Whatever your dream is today, I sincerely wish that it comes true and that you enjoy a happy and fulfilling life.