Kong Kat Wong Ideas Festival

Hi guys. Also to start, everyone who's got a cell phone, please put it in airplane mode. Thank you. Well, we should be able to be here, you know, sitting here with you guys. It's been too many years. It's been 22 years, more than 22 years. He already told you how I am, okay. But the last few days, thanks to the Chu and Cherien team, it was exciting to reset some place that is so familiar. I would say those were the three most memorable and happy years in my young life at that time. It also changed my way of thinking a lot. This afternoon was especially meaningful to me. I got to actually visit four, actually five of the professors I had in the past. Well, they all look the same, yeah. And that's not the scary part. The scary part is they're still working so diligently in the office. When I visited, I didn't expect to see that.

Campus is the same. I can still find my way. I really don't want something to look exactly the same. I arrived in Chicago two days ago, came into town right after dinner with our friends here. I actually walked around campus a little bit. The first place I went is Hicks, the library. It's like I was programmed to do that every night for three years, and now the first thing I do is just walk there. Then I went downstairs. There used to be a study lounge to the right across from the library. It used to be open 24 hours. The library would close at 11:30 or something, but the study lounge would remain open for the night. That's actually where I spent most of the night during those three years. Sorry, a little bit different. I don't see people delivering Papa John's after 11:00 anymore in the lounge. Anyway, it's good to be back. Today I'm very excited about other things too. OTO and Tim, let's have some conversation to help me. In the last two days, being in the environment has helped me recollect my memory slowly. In the old building, I almost felt like someone who had memory loss when I came back, but now I'm gradually building my memory. Lots of little funny stories, some of them I would never share with anyone, came back in my mind.

It was actually a simple, logical decision on my part. My dad was an engineer. He introduced me to computers early on because I saw him. We didn't have enough money to buy the full-blown latest state-of-the-art computer, so I would see him coming home every day bringing different parts, and it took him a couple months to actually assemble a computer that would work when I was much younger. So I was attached to the concept of computers. I wanted to go where the computer major was actually good. When I looked for schools, Purdue was really high up in the ranking for computer-related subjects, like EE and aerospace engineering, of course. That's why I came to Purdue, wanting to study computer. What I didn't plan for was that I was going to topology to learn the hard way, CS of computer. Later on in my life, in Xiaomi, we start doing lots of hardware, but at the time I discovered the school of computer science. I don't know if Dr. Dan SMO is here today. This afternoon I just shared with him that he literally changed my career by giving us such a great program, CS 180. For those of you that are in CS 180, after that class I realized that software actually was the career I wanted to go into. So I stuck with computer science for the next three years, and after that I just went directly to Microsoft because Microsoft was doing a lot of software.

But me, I shaved more with no offense. I went to his office this afternoon and I said, 'Professor, do you remember me?' And he said, 'Yes, KK, I remember you vividly.' I couldn't believe it. He has so many students, 30 years of teaching CS 180. How can he remember me? Then the memory recovery that I told you about came back. I remember maybe the reason why I left an impression on him was because in the very beginning of the class, I was not doing well. I actually flunked my first exam in his class. He was going through the results, and I was missing some simple answers, at least in his opinion. Then he was speaking to the class the next day and said, 'Well, computer science might not be for all of you. Some of you that have a hard time picking up the basic concept might want to reconsider.' But he never gave me the meaning look. Dr. Tensefor is the most patient man in the world. He really just said that to make sure we don't go into the wrong field. But for the whole semester, I was able to go from flunking to, in the end, I like to think I ended up the top student in his class during the final. He was avoiding eye contact with me, slowly giving me the expression of encouragement every time I got better and better. I told him, 'You helped me fix not just my programming, but what you truly gave me was the experience of overcoming my confidence problem.' I told him it was a really meaningful semester because I came in not knowing enough about programming than many other students, but in the proper learning environment, I was able to meet the expectation and excel. I told him the same thing happened to me again and again down the road. When I graduated from Purdue, I thought I was pretty good because I got all A's in CS and other classes. Then when I go to Microsoft, guess what? The same thing happened again. I saw all these talented people from everywhere, and the same process of care where I have to, in the beginning, feel a little bit inferior, but with the right way of approaching problems, I was able to catch up to how other people are doing and in the end excel. That happened again and again for all the years until this afternoon. I personally thank him for that. That's why he was the first professor I needed to sneak away and pay a visit to.

Well, I told you already. I literally came out in a study lounge. I don't know why I do that because we had a habit of not sleeping through the night. That's actually what many of the computer science students like. We tend to think clearer when we don't see the sun. So we basically just pulled all-nighters, me and my classmates, in a study lounge. We knew how to order the pizza with the most cheese. I don't know if you used to do that. You would pick the first table closest to the door, and around midnight you would be placing orders for Papa John's. But when the delivery guy came, they would ask for Peter, and Peter probably had already gone and didn't wait for the pizza. Then we would say, 'Okay, free pizza,' and we would be the first one to pick it up. That's how we got very affordable pizza. But what kind of student was I? I like to see myself as a diligent student. At the high point, I concurrently worked for three professors as an undergraduate. That's also why I had to be in the study lounge. There's a drawback though: it may be really hard to get up in the morning to attend classes. But it turned out to be okay because I was looking for the professors that were teaching me, so they knew I was actually working and I was up to speed on the material. They didn't give me a hard time for not going to their classes all the time. It was unusual in a sense that undergraduate students got to work for these great professors because they usually take graduate students. I still remember the tactics I took to get the job. I would go to a professor and say, 'Professor, do you need someone to help code your algorithm?' And they said, 'Sorry, I only take graduate students.' And I would say, 'Can I work for you for free? And if I fail, in two weeks fire me. Otherwise, you don't let me do that for you.' I don't think they actually took me on for the fact that they didn't have to pay me because all of them ended up paying me after one month. I think they actually took me in because they were feeling my passion to learn from them. My initiative actually moved them to some degree. But that's how I got my first jobs. Those three years were super fun years for me. When I was chatting with someone, he asked me if there's one thing that I would do differently. I said I would have partied more because after starting working, time just flies. Before you know it, you're back at your school after 22 years telling the students how diligent you were.

The guy next to me, the shoulder guy, was my roommate. We had a lot of fun together. He's the kind of guy that always goes to parties, and he was the president of the Hong Kong Student Association for three years. We still stay in close touch. He's in Hong Kong. Two of the other students that have not graduated yet, they're working in California. After this trip, I'm going to stop in California and I've taken a lot of pictures of the new campus. I'm going to share them with them.

That's in the W building hall next to the fountain, is that right? The W is along the engineering mall. I don't know, I can't tell. Anybody? The building probably changed, but you can still see it. The color tone remains the same for all buildings at Purdue. On the right-hand side, you can see that we were still doing those newsletters and making copies with the photocopier. Very old-school. Brings back a lot of memories.

I actually did an internship with Microsoft before my last semester. It was really impressive for a computer science student to go to the biggest software company in the world at the time. I learned a whole lot from Microsoft. The first five or six years, I had the most fun being a young programmer. Even if you didn't pay me, I would be home programming every night anyway. I was shocked that anyone would actually pay me to do the things that I would do for free. I still remember in the first five or six years, my wife, who was still my girlfriend at the time, would tell you that I never went home before 2 AM at Microsoft. In the morning every day, I would be in the office programming alongside not just me, but my manager and my manager's manager. It was a very hot core age for Microsoft. It was already very big, but the environment was very startup-like. This is the late nineties, 97, 98, 99, before 2000. I started working in the MSN group, and then I moved on to work on the data warehouse project. For those of you that don't know, the project unit manager at the time overseeing this project was Satya, the current CEO of Microsoft. It was a startup project within Microsoft. We had to do the e-commerce platform. This is when I actually worked on a data warehouse project. My first task was to analyze the web traffic log for MSN, which was the number three biggest site at the time. I refreshed this part of the memory with a professor this afternoon. The project was so fun. When I first joined the group, it took them more than 24 hours of processing to handle their daily log, which means they couldn't catch up with their daily log. 24 hours to catch up with 24 hours of new log. So my task was to optimize the performance to make it run faster. We went from 24 or more hours to 15 hours, to 10 hours, to 8 hours, to 5 hours. In the end, I remember we actually got it down to 1.5 hours. That little story actually captures the fun of being a programmer. Nothing was more thrilling than seeing how you trim off seconds with better algorithm, better programming, a smarter way of dealing with your code to get to that point. For those of you not in computer science, you might think, 'My god, so boring,' but if you're in computer science, you know what I mean. I worked on that project for a little bit, and then I worked in the Windows team on the messaging platform. Many of the things I learned from the book of Douglas Comer helped with internet working with TCP/IP. I worked on messaging, and then I took a pretty big change by doing an internal transfer to Microsoft China in Beijing. With that change, I also took on a completely different role. I was tasked to pick up the Windows Mobile team in China. In the end, I had close to 150 very talented people working on various features for Windows Mobile. During those four years, I realized how fast the China market is moving, how big the China market was, and how fast the mobile internet was coming along. That's also why I decided to move on to the next phase of my life, to co-found a new startup at the time, Xiaomi. Leaving Microsoft was a natural move because I felt I cannot miss this wave. I've already missed the internet wave. Going back in perspective, I graduated in 96. If I was super smart, I would have gone into the internet field right away. I should be the one that co-founded Facebook or Netscape, but I completely missed the internet wave. We had access to the internet, of course, but we were just using it to do fun things, to talk to each other, to do free downloads, the wrong content. But it didn't come to the idea of, 'Hey, why can we do e-commerce?' If we did, we could have been Amazon. So we completely missed the internet wave for no good reason. But when the mobile internet wave came, I was in a position to see it coming. I just had a gut feeling that we cannot miss this again.

It's really interesting. People around me all think that it was such a big decision that I had to make, leaving the comfort zone, leaving behind the good pay. Microsoft paid me very generously as I moved up the rank. But at the same time, I had this concept of a safety net. I spent 15 years with Microsoft already, and I had done pretty well. My reviews at Microsoft were as good as my GPA at Purdue. I would like to call that my safety net. If I don't use my safety net, I'm actually wasting it. I knew that if I were to take a venture and fail miserably, I could still always come back and Microsoft would gladly take me. I still could code like when I was young. But when I was observing this huge opportunity of the booming China market and the booming mobile internet market, this perfect storm was hitting in a very obvious way in my mind. It was really obvious to take the leap of faith to capture this opportunity. So to me, it was actually not a very difficult decision at all. I had been thinking about capturing that wave for a while. The last straw that broke the camel's back was meeting our CEO, Lei Jun. I don't know how many of you know about him. He was already a well-known serial entrepreneur before Xiaomi. He did multiple startups, and all of them were fairly successful. I met him via a second panel firm, and my ex-coworker at Microsoft. I met him in a coffee shop. We basically sat there for five and a half hours. What did we talk about? We basically talked about products for five and a half hours. He would be carrying with him a bag that contained six or seven smartphones at the time from different makers, from Apple, from Motorola, Samsung. Xiaomi at that time was actually shipping very shitty handsets. They didn't improve until we came. The first iPhone made by Google, he would tell me that he had taken apart all these phones the hot way. He would be telling me the things that he would have done differently for different products, for hardware, for software, for the design. We just talked about products for five and a half hours. To me, it was actually very unusual. When I was at Microsoft during my last couple of years, I had reviews and meetings with the senior VP of Microsoft on a regular basis. None of them would actually talk to me about products like that. None of them would actually show me their passion for products like that. I felt it was actually the core of building a startup. The core for a startup is really to create a good product. That was the last straw on the camel's back. I realized that the two of them were about to form a company. Toward the end of the coffee, which is probably the most important coffee I have ever had in my life, I just said to them, 'Okay, whenever you guys are up to something, count me in.' I was the first co-founder to join this startup. It was unnatural, no logical steps. We eventually had seven co-founders, myself included.

The same process of how you naturally recruited me actually worked really well for the other co-founders. But we realized how complicated this startup is. By the way, can I assume that you all at least know Xiaomi to some degree? In one sentence, Xiaomi is a company that does hardware products, software, and internet services. The core of the company was to bring cutting-edge technology, the premium of premium products, to the mass market at very affordable pricing. That's basically what the team was trying to do.

Exactly. Just to give you an idea, at the time, if you wanted to get a flagship phone from Motorola, Samsung, or Apple, it would cost about 6,000 RMB, not very affordable to most people. But when we launched our first phone, we actually used all the top components from all the makers that were providing components to these giant manufacturers. If Motorola is using the latest CPU from Qualcomm, we used the same CPU, we used the same top-end components for everything. When we launched the first phone, when they were selling their phone for 6,000, we sold our phone at below 2,000 RMB. That is what actually brought products to the mass market. That is what caused even the journalists to literally jump out of their chairs at our press release, couldn't believe that we were actually selling for that price. In the beginning, people were skeptical that we were cutting costs by using inferior components. I still remember when we first launched the first phone, people would actually take it apart to make sure we indeed used the components that we claimed we were using. They did it for many products for Xiaomi, to the point that eventually we decided to put so much focus on designing the interior of the hardware, knowing that they would take it apart. We wanted them to be really impressed by opening up the hardware, the inside looks better than the outside. But after a while, they believed that we meant what we said. I won't go into too much detail of how we achieve that, it could take two or three hours. But knowing that was what we set out to do in the beginning, we realized how complicated this endeavor was. That's why we had to find seven co-founders. If we could, we would be finding 70 co-founders, knowing how complicated this was, but we couldn't find that many people. One of us came from a hardware background, me and another co-founder came from software, two co-founders came from design, one from industrial design, another from software design, the look and feel, one came from program management. Lei Jun was the ringleader, he knows everything based on his previous startup experience, experienced in management, product development, everything. So that's how we looked at it. We're building a hardware phone, these are the people we need, and we included them as our co-founding team. That's what we planned from the beginning.

If you tell the VC that you were building a phone, they would think you're crazy because all your competitors are Fortune 500 level giants. But fortunately for us, Lei Jun was actually very well-connected in the VC world. He was a well-known angel investor in the industry. So people actually gave a lot of credit to our background. They took a leap of faith. They didn't even want to hear about our business plan because if we told them, they might have a hard time convincing themselves this is the right thing. They would just say, 'Okay, whenever you guys do, we're just blindly throwing the money for the angel round and go with you.' Of course, now they are very happy they made that decision. But at the time, I would say the thing we set out to do was so impossible that it gave us actually the best chance of success. Let me explain what I just said. If you actually go out and do things that everybody agreed would be successful, well, if the logic is so convincing that something is going to be successful, I can bet you ten other people have already done that before. You have to assume that this world is full of smart people. If something is easy, everybody's going to do it. If it's so logical and possible, people have done that already. The kind of startup that would be successful is the kind that would seem impossible at the beginning, a little bit crazy, to scare away the skeptics, to push away the big guys. If I were to pitch this idea inside Microsoft at the time, it would be shut down because it was crazy. But these are the kind of ideas that actually turn out to be the most likely to be successful, assuming we can actually execute it according to the plan, assuming we were lucky enough. It takes a lot of luck to be a startup. But I think sometimes, having the conviction to believe that what you're building, if you fail, you fail, but if you were to succeed, it's going to make a positive impact big enough to make you not only proud but willing to work on that problem 24 hours a day, seven days a week. That's literally what we have done in the past eight or nine years, especially in the early days when you really don't have enough to go through those sufferings. Comparing to my days at Microsoft, where I never went home before 2 AM, the startup is like ten times tougher.

You lose all your friends basically. The only time you get to spend is probably with your family on the weekend. But every day, you basically woke up and you would be working. Fortunately, we had enough funding that we didn't have to work in a garage. But we basically had to work. In the beginning, it was tough because although we could find some VC to support us, it was hard to recruit employees because they would think you're crazy too. Each recruit took so much time. We would be talking to the candidate we really wanted to get for one whole evening, another evening, another evening, to the point that they are so tired that they feel guilty of refusing to come. Then they would luckily agree, and you move on to the next candidate. But sometimes that's good because that's a high barrier to really find the people who believe in your vision and accept, 'Hey, I want to try to do the impossible as well.' You really have to find people who share your craziness at that time to take the leap of faith. But thinking back, that time was awesome. It was very productive. I personally learned a lot in this phase of my career.

The left picture was taken on the first day of our company, Xiaomi. Day one. Do you see the rice cooker in the middle? For those of you who don't understand Chinese, 'Xiaomi' literally means 'little rice' in Chinese. So as a symbol of the name of the company, we actually brought in a rice cooker, found little rice, each of us basically toasted and then got right to work. Unfortunately, that was the only picture we took that day. Apparently, we didn't think this would become a historical moment at the time. So we really just did it in a casual way. None of us did our hair, we didn't dress properly, taking the picture against the light. We didn't know we were changing the world on that day. We knew, but we didn't think the picture mattered. But now every time we celebrate our anniversary, people would share and circulate this picture. We all wish we had taken a proper picture. So now, I do a lot of investment, and I sometimes attend their first day of work, and I always remind them to take a good picture today.

We wanted to find something that's personable. We took notes from companies that are successful. They usually have a very simple but very personable name. Apple is something that everyone eats in the world, a healthy fruit. So we were thinking, 'What do people in China eat every day?' Rice. Originally, before the firm, it was something else. But as a rule of thumb for coming up with names for an internet company, you don't want names that are seemingly powerful, large, or strong because that's actually very impressionable and far away from the users. People like names that are very down-to-earth, particularly because that's the market segment you're going at. 'Xiaomi' means 'little rice.' Nothing is smaller than a small rice, nothing is more humble than that. We liked the name. After we started using that name and became reasonably successful, we saw a bunch of startups using different variations of the word 'rice' in their company names.

Before I answer the question, I have to call out that Xiaomi was not a typical startup. At the time we started, each of us was already carrying 15 to 20 years of work experience. So you can say that Xiaomi was founded by seven men, the average age of the founding team was 40 years old. I was the youngest at that time. Each of us was already fairly senior in our respective fields. That's what made the funding interesting. We could have funded the company on our own, especially Lei Jun. But we insisted we should take outside funding. Why? Because we wanted to give ourselves the extra pressure by taking other people's money. If you were to do a project out of your own pocket, if you fail, you can accept that because you're losing your own money. But if you have taken your best friend's money or money from a team, you couldn't sleep if you were going to lose this money. It gives you that extra pressure to do well. That was the reasoning seriously for us to take the funding. It makes you accountable to somebody else. The other reason is you want to take their money so that the valuation is a fair valuation. If we were to invest our own money, the valuation could be out of touch with reality. But when you take money from outside, believe me, nobody's smarter than those VCs. They will not give you more than what you're worth. So from then on, each round was based on the company's performance and projections. We rapidly grew in terms of valuation. We were really fast in reaching the billion dollar mark. I think we have skipped that one. I don't have the exact numbers, but we broke the record. Two years after the launch of our first phone, we reached 1 billion US dollars in revenue in sales. Four years after launch, we actually reached 10 billion US dollars. You probably have no idea because when I was in school, I didn't really know what 1 billion means. A million was a lot already. But to put that in perspective, it took me seven years to get to 10 billion. Actually, sorry, 100 billion RMB, which would be roughly 15 billion US dollars. It took me seven years. For some of the most successful companies on this planet, for Google it took nine years, Facebook 12 years, Apple 20 years, Alibaba 17 years, Tencent 21 years. That's what made it so incredible.

Xiaomi is in a funny position where ever since we launched our first phone, we have been on the map for the media and the press. Everyone knows about Xiaomi since the early days. It's almost like putting this startup under a spotlight or under a microscope. So we're already used to the voices that either think we can do anything, they ask us to build houses or cars, to the other spectrum where people are really pessimistic about the future of Xiaomi. They say things like, 'Do you think you would be dead by the end of this year?' They said things like that every year. Sometimes they're right, sometimes they're wrong. But what I know is that plain areas are actually very truthful. I cannot speak for all companies, but for Xiaomi, I can imagine we are constantly in a state of having this sense of urgency that if we don't overcome this problem, if we don't fix this challenge, we're dead in three months or less. Especially when we're dealing with the hardware business. We have to make very accurate estimates about how much inventory we need to incur. What if we actually ship a shitty product? Before you release it, you basically have to pre-order all the parts three months before. You're basically pre-paying for the phone. If you release a very bad product and nobody buys it, you end up with three months of very expensive inventory. One failure like that could easily kill the company when Xiaomi was small in scope. Now we're a little bit bigger, maybe two products could kill the company. For companies like Apple, maybe one bad product won't kill them, but two products is enough to really destroy the company. They are not as robust as you think they are. The other risk that we constantly remind ourselves of is to never forget what we set out to do at the beginning. Xiaomi set out to bring the best product, excellent in design, excellent in user experience, excellent in quality, at half the price, affordable to everyone. Do not overcharge your beloved customers. Do not take the margin just because you can. That is the fundamental principle of what makes Xiaomi. I have a lot of faith in Xiaomi because if we can uphold this principle forever, I think we can be very sustainable, just like Walmart, just like Costco. Those are companies that do not overcharge, do not take too much margin from the users, despite remaining competitive against their competitors. The moment when Xiaomi stops doing that, that would be the end of Xiaomi. That could end very fast. That's basically what keeps us awake at night, knowing that a big company could die faster than a small company. For those of you who have used older phones, how many of you remember Nokia in its prime days? Remember Motorola, Palm? It didn't take long for them to collapse. For Motorola, it actually died within one and a half years. Nokia died almost overnight. It's so scary, especially for our industry.

Because we had that original go-to-market, cell phone was really just the first product we used to test out our model. Because this model worked really powerfully. Just to give you an idea how powerful this idea was, when we launched our phone, it took us less than one and a half years to become number one in China. That's how powerful this model was. Now we're number two or three, second only to Huawei. But the model works really well even for the most competitive product in the industry. Once we have proven this model works, we have taken this model to do other things like TV, router, smart speakers. We gradually grow our portfolio much more. Even though I introduced you as a cell phone company, in fact you are much more than that. Today, we have over 120 million IoT devices connected, not including the phones. You probably don't have an idea how big that number is. 120 million IoT devices connected. That's actually the largest IoT platform in the world, not even Apple, not even Samsung, not even Huawei have that number of connected devices. That's just to tell you how powerful this Xiaomi model was in going into each vertical. When we take the same model in going into, say, the air purifier market, Beijing air was very polluted. Guess what? Before we went into the air purifier market, people would have to pay 7,000 or 8,000 RMB for a decent air purifier. If you have to change the filter, it costs another 1,000 RMB. When we go into the market, the cost of building a very good purifier costs less than 1,000 RMB. Those are just machines that have very big fans and a few filters. It goes back to your thought of companies charging margin blazes, they overcharged. So we managed to actually sell the air purifier for 799 RMB, one tenth of their pricing, one third of the comparator we measure against, with the same level of purifying performance. We literally gave it to a third party to do a test, and we were on tie if not better than these overcharging companies. We're talking about one tenth of the difference. With that, you can understand why we quickly became the number one air purifier manufacturer in the world in less than two years. When we do other devices like TV, router, we do the same thing. Years later, we finally released our rice cooker. We wished we had that on the first day of the company, then we would be doing marketing for our own rice cooker. But we do the same thing, we take this approach. That's why people are starting to see Xiaomi as more than just a cell phone company. It's an IoT company now. It's really exciting. We're also really fortunate that when we build the company, we came across the wave of mobile internet, we came across the wave of IoT, we came across the wave of AI. Just imagine, now that we have this many connected devices, these devices are generating a tremendous amount of useful data every second. With AI, with machine learning, many of you are studying data science now in school. That's wonderful because when you graduate, you're ready to solve some real world problems. Just imagine how much intelligence we could build into these IoT devices. Finally, we can actually turn your home into a smart home.

There's a lot of opportunity outside of this field, but within this field, I think 5G is going to be a big deal. You might think that your phone's connectivity and data is already fast enough. No, it's fast enough for the experience that we are giving you today. App programmers deliberately make sure that they operate decently given the current network bandwidth. If you give them 5G bandwidth, they will give you something that has an even richer experience. Finally, AR and VR, all those high-performance streaming experiences are going to be everywhere in the application. So I think 5G is going to be a big deal. AI this time around, we talked about AI in the past 40 years, it goes up and down. It goes up as hype and then people get disappointed and it goes down. This time, we have the material for good. We have strong reason to believe that it's going to be a big deal. I don't think it's going to achieve what we have seen in a movie like Skynet or things like that, but it is powerful enough to turn a lot of devices into intelligent devices with very little programming. No more decision trees we have to do. We could actually give them enough data so that they can become smarter. Outside of this industry, I think biotech is going to be a big deal. I think the four pillars that Purdue has picked for the 150 years are just spot on: space exploration, biotech, AI, sustainability of the earth. I think these four have tremendous future. I'm personally actively investing in fields outside of just the one I'm familiar with.

Too many. Do I have another two hours? Given the time limit, I will tell you what I think is the most important thing first. Do not rush into entrepreneurship too fast. Make sure you accumulate enough experience. Make sure you have the fundamental strength before embarking on an entrepreneurship project. Because it's really hard. Even if you have a brilliant idea, if you don't have the right fundamentals or experience or common sense, common sense usually comes after you fail 20 times. If you don't have enough of that, you're just going to work on a good idea and see it fail. You will see other people doing the same thing, but because their execution is much more well-planned and mature, they would get to the finish line before you. So number one is do not rush into entrepreneurship. It took me 13 or 14 years before I finally had the confidence to feel ready. The second thing, I shared this with someone over lunch today. They asked me this question. Just remember there are two things. Do not underestimate yourself just because you're going to go through difficult times. Even for the luckiest startup in the world, they go through many tough times. During the downtime, do not underestimate your caliber and your team's caliber. But at the same time, never overestimate your understanding of what the customers need. I have seen so many entrepreneurs fail because they have a pipe dream to think that this is a problem that needs solving, that this is the perfect opportunity for their startup to solve this problem. They believe that this problem is needed for everyone, only to find out later that it's a very narrow problem, only very few people actually have that problem. I've seen so many entrepreneurs overestimating their ability to truly understand the true demand of the market and the customer. How do we address that in Xiaomi? We were humble enough in not overestimating our understanding explicitly about the need of this younger generation. I cannot imagine myself understanding the needs of teenagers anymore. That's not my day-to-day life anymore. I don't know what they need. So from day one, we tried everything we could to be closer to the customer, to listen to their feedback. For example, our particular release of the OS, every time we upgrade, we actually pop up a survey and ask them, 'Tell us the top three things you like about this change, this build. Also tell us the three things you hate the most about this build.' We ask them to give us feedback, and we did it weekly. We did weekly updates on the OS. Can you imagine flashing your phone every week? But that's also where you get the feedback from your customers. We would give them, 'Among these five features, which one do you think is the most useful?' That's very helpful. They actually help us to spend our very limited resource on the most important problem. No matter how many engineers you have, you always have limited resource. So we realized this was the way to make sure we spend our very limited resource wisely on what matters. I would leave it at that. That's a lot to learn in entrepreneurship. I know where you are. I really wish when I was your age sitting in this room, someone with the experience of Lei Jun and myself were here to actually warn you of the things not to do and to do. So hopefully this is useful.