A. Richard Newton Distinguished Innovator Lecture Series Dr. Jun Wu

Thanks, Chuck, and thanks to Victoria. So if you see on the back row and you can hear me, please raise your hand. Okay, wonderful. So I was very, very happy to be here and to be given the chance to share with you some information. After I worked for many, many years in industry in the Valley and also have some experience in China, instead of telling you something new, I really like to share what I thought. I really hope that instead of just giving you some ideas, I really like to learn from you and get some feedback from you. Okay, so before I started, I talked to some guys here and they asked me whether I should join a small company with like a thousand people or some big companies with big names with like 10,000 or even more people, which seems to me very stable. I really couldn't give you advice which one is better because I know you and your expertise and your personality, because it really depends. Some people really like companies with 1,000 people and then help that company to grow from 1,000 to 10,000 people, but some people really like small startups. But one fact I can tell you is that for 400 to 500 companies, the average lifetime for those companies is only 39 years old, which is about half of a person's life. We can live about on average 80 years. So that means those big companies will lose their essence probably one day suddenly in five years. Some big names like Sun Microsystems, well it's part of Oracle now, and they have a competitor. I think also got benefit from some Berkeley employees and students actually. It's called SGI, Silicon Graphics, and it's actually basically founded by people from Stanford and many engineers graduated from Berkeley, and you probably haven't heard that name. There's also a big name, Fairchild. Is there anybody who never heard this name, Fairchild? Well, we are in Berkeley and engineering school. In the morning I gave a talk to another group of people and I found 80% of people never heard of Fairchild. It was a great company, maybe one of the greatest companies historically, but it's a very, very small semiconductor company today. So what I want to tell is that everything got changed. Some companies got risen, some companies fell down, but Silicon Valley is still Silicon Valley. We are not another Detroit, we are not another Pittsburgh, even though we don't have silicon, we don't have much silicon here today. I would like to say it's the value of innovation, or value of software, or value of internet, or value of whatever, but not a value of silicon. But still people call it Silicon Valley. So what happens inside Silicon Valley? That's some secret I want to share with you today. So let me start from some story of Google. Let's start from how Google grew from a small company to a bigger one. I would split Google into three stages: very small one, medium size one, and the huge one. During the early stage of Google, I joined when Google had about 300 people. The only thing the management shared with the information from the management point of view. At that time, the challenge for Google was that the company was so small and we have so many things to do. We have so many things to do but we have so few people. So the management only worried about whether we do the right thing. So what they did is that they spent one day every week. We have a small group of management with Larry, Sergey, Eric, and Urs Hölzle, who was the first engineer VP at that time, and Jeff Dean, who is the guy in charge of infrastructure today, and also the first Google Fellow at that time. These five people stay together in a small conference room with about 10 people and ask engineers to present their work one by one every week. Not every week they work, every week you present maybe every three or four weeks, so everybody can have like only five minutes to present his work. What they want to make sure is that you do the right thing, and that's it. They don't care about how you implement anything, and they even don't care about the progress you have done. So they trust on your skills and your judgments on engineering and on products. They have to trust on you because they have no choice. There are so many things to do and they just want to make sure that you pick the right thing to do, and that's it. Why? Because they know that your future has been bundled with the future of the company, so you have to work hard. They don't worry about that. When I was in Google at that time, I soon went back home before midnight. Usually the time I arrive home is about 2:00 on average. Sometimes I went home at like 5, 6:00 in the morning and then get back to company at about 10:00 in the morning. So everybody worked like that. We worked like a dog, about 7 days a week and like 15 at least 15 to 18 hours every day. That's our life. Because we have no hope if the company died. So the management don't worry whether you work hard, whether you cooperate. They don't worry about that. They just want to make sure that around 500 important projects, we picked maybe 10 most important ones to do, or even five. We thought about image recognition to recognize people's face, we thought about speech recognition, and we thought about machine translation, but we never did that. We know that they really will change the world actually, but we cannot afford to waste time to do that. So we focused on search, all kinds of search. We focused on advertisement on both search and content targeting advertisement. And we focused on something you never think about: email. We focused on email at very early stage starting from 2001, even though we launched Gmail service I think 2003 or 2004. We didn't want Yahoo to know that we focus on email because we still partnered with Yahoo and they probably would want to terminate our cooperation. But we prepared for that because we want to have a glue to people to our services. Without Gmail, we cannot do that. So we focused on a couple things and eventually Google grew to about 1,000 people and went to IPO. At that time, many people got rich. What management worried about? Many of them grew from a fresh grad to a senior engineer or even a staff engineer. They manage a team of 10 people, and instead of doing only five or 10 projects, we did about maybe 50 projects or even 100 projects. So at that time, what management really worried is whether people can still cooperate with each other, whether we still have the same goal. At early stage, they don't worry about that, everybody has the same goal. When company has 2,000 people, whether we can still do that, we can still cooperate. So the company found an organization, a small group of management we call OC, Operation Committee. It has CEO, CFO, chief... we don't have CTO, but we had people in charge of engineering at Google, and he has some senior VP to report to him. We have some kind of chief product manager even though we don't call that name, we call like CVP product. We have the top legal guy. This group of about 10 people is called Operation Committee, OC. So then Engineer Award becomes the OC Award. We have the Founders Award also. So if you did something wonderful and with great impact to the world, you got award. So that's the company at that time. The organization of that time is really important. Google is over there. We founded lots of project groups that have very structured report structure. Like Amazon, they have this very massive structure basically every has two bosses. One is the boss in charge of your product, so what your project, he takes care of technical issues or product issues. And you also have a people manager which gives you bonuses and if you want to have some promotion and want to talk to somebody, he will take care of that. But he only works for your HR issues instead of your product issues and your project issues, he didn't care about that. So the organization became quite messy, but people cooperate very well. We always talk to each other before we start some new project and make sure that we don't do two same things inside the company. You don't need to pay attention to Microsoft or Apple. Apple Jobs is like the king of the company, right? And we don't have such king inside company. And Microsoft, they have some branches and they fight with each other. So that's what management really cares: make sure that the company still wants to cooperate. The advantage is very clear, and the people still work hard and cooperate. The bad thing is the cost of cooperation is so huge. So give you an example. Suppose you want to build an application for Google Maps, a new application, and everybody likes it. How could you build a team? You need 20 people, but unfortunately probably you cannot find 20 people in Mountain View in three months. If you want to hire people, it really takes time. And then the management tells you, well, so in New York, there are five engineers, they just finished their previous projects. Whether you like to take over these five people? You say yes, why don't I take over these five people, that's cool. And you got five people from New York, you got five people maybe from India, and 10 people from Kirkland. Then to cooperate these three teams inside three locations, and you are in Mountain View for example, it's really painful. Eventually, after some change and reorganization, it becomes that in New York office, I had 20 projects with 30 maybe offices in 30 countries. Then in Mountain View, there are some projects that need have five engineering groups from four countries. So it's totally messy. Then there becomes some projects because it's so messy, there are some projects with engineers from many different areas, and even the HPM and the engineering director don't know whether some small offices they have really have contribution to this project. So the management becomes really painful, and the bureaucracy increase dramatically just in two or three years. From 2007 to 2009, suddenly Google found that even though competing with Microsoft is still okay, competing with Facebook is almost impossible at that time. And also Google faced the challenge from Apple or mobile applications, mobile operating systems, and cell phones. So then Google made another change. The current change localizes Google. They don't have OC anymore, no operating committee. So there's only one guy in the middle called CEO, and because his name is Larry Page, he has a team called our team, L Team, so direct report to him. So Alan Eustace, who used to play role as like CTO inside Google, he actually become like the CVP or one we call product area. We don't want to call it business unit because inside BU you have to take care of your budget, you have to earn money. He's from a company that definitely knows that. So in Mobile BU, you have to sell your cell phones to get money, but in Google you don't need to do that. We call it product area, but otherwise it's more like BU. So he's in charge of product area called Knowledge. It includes machine translation, speech recognition, and Knowledge Graph, which we think is an important engine behind all Google products, which will put all the concept together using a graph. And he's also in charge of Google Maps and GE services. And Sergey, he had Google X and Innovation. I single was my boss, and he's in charge of Google Search, the oldest and the biggest branch inside Google. And Susan was the only female inside OC, and now she's in charge of many things. Right now she's in charge of YouTube. YouTube used to report to Salar. Actually, I would credit him as the inventor of Google AdWords. Without Salar, Google don't know how to... well, Google knew how to earn money but cannot be that profitable. So Salar was in charge of Google Ads for a long time. During the early days, even Larry Page during some time, he actually got confused whether Google can deliver ads as a service, and Salar worked really hard to make it happen. As Sundar actually took over the... today, Urs is still in charge of global infrastructure. If you think of Google data center, he is the guy who built the data center. Before he joined Google, actually he was a professor, very famous on object-oriented design, that's his expertise, not data center. How to build a data center? And Sundar took over Android and Google Chrome, and that branch was reported to Andy Rubin before, but Andy wants to do something new, he wants to build like a Google, the next Google. So right now we have some product area. Each product area is like a small Google. They have their own power to give for example somebody some stock options, and they can make a decision, and they have lots of powers inside. So right now we have instead of having one Google, we have some product area. So the border is very clear, so the cooperation between product area becomes very few. So inside product area, there also connections of outside comparatively is independent. So that becomes the third or fourth stage of Google. So I just want to tell you some story from the management point of view. Now let me talk about something about Silicon Valley in general. I was asked by many, many people why Silicon Valley become so successful. You know that we are so successful, we have almost one company listed in NASDAQ or New York Exchange every week, and 40% of the biggest 100 companies in NASDAQ are headquartered in Silicon Valley. There are many people give us the reasons. We have the top two best universities in the world according to the number of best majors. Both Berkeley and Stanford are better than Harvard or Cambridge in number of best programs in the world. And the way we had culture, I like hippie. Actually, if I have time, I will tell you that there's a very strong correlation between hippie and the innovation of Silicon Valley. I'm not kidding you, actually it's true. So I would say this is a secret of Silicon Valley. There are many explanations why Silicon Valley becomes so successful. One story I heard is because we had Stanford in the Valley. You know that many companies were incubated inside Stanford, right? Like Cisco, Sun Microsystems, Yahoo, and Google. But I always ask myself whether it was true. I really thought about that because if one university is so important, why doesn't it happen in the Boston area? In Boston, they have almost twice as many universities as they have here, and they have almost one quarter million of college students. It's a big number compared to the area with only about 4 million people. And they have Nobel Prize laureates, one every 10,000, which is also a very high percentage. But they don't generate any great companies since the 80s. I checked 12 Bay companies inside Fortune 500. The last one was incorporated in 1979. Why they don't have great companies in 30 years? This is a good question. So somebody says weather is important, but why not in Florida? Silicon Valley happens. And somebody told me that venture capital was really important, but the truth was early venture capitals came from New York area. We didn't have venture capital in the 50s and even in the 60s. Most of the famous venture capitals in this area were incorporated in the 70s, like Sequoia, KPCB, and NEA. Actually, they were founded in the 70s, but Silicon Valley started before that. So why is that so important? Also, if you look at the history of venture capital investment, even though they were not called venture capital, they did a similar thing like JP Morgan 100 years ago, more than one century back. He invested on Edison and Tesla just like a venture investment. But why Silicon Valley was not in New Jersey or in New York? I even don't agree that Stanford has so much contribution to the Valley as many Stanford students claimed. If you look at the status of Stanford University, it almost went bankrupt at that time. After World War II, Berkeley has a much bigger name than Stanford. I give three pictures here: Oppenheimer, Lawrence, and Compton. They basically led the entire project, the Manhattan Project. Without these three guys, we cannot have Manhattan Project done. But if you list 50 most important people inside Manhattan Project, you cannot find anybody from Stanford. So there must be something else. Let me skip all this. So other possible reasons. I put some Chinese characters here. To decode the secret of Silicon Valley, I probably briefly introduce the history of Silicon Valley. Let me still briefly review the history of Silicon Valley. I put some Chinese characters there. The first one means timing, so we need a good timing. The second one needs location. The third one needs people. So if you put these three things together, something will happen. Silicon Valley started from a good time. In the 50s, IBM wanted to build their storage. They wanted to build a new branch instead of building another office near New York. They thought, well, because it's something new, let's found a new location to build this new branch. So they eventually found in San Jose there are still some area they can purchase. They purchased a land and built a so-called San Jose Research Center, but now called Almaden. Before IBM entered this area, nobody here knew how to build a computer. Because computer was so new at that time, they eventually brought thousands of experts to this area. So that started the computer industry in this area. So IBM is important. The other one you heard his name, Fred Terman, who was credited as the father of Silicon Valley. I just told that after World War II, Stanford almost got bankrupt. So he read the will of Leland Stanford and found that even though Stanford cannot sell the land to the public, they can lease the land and get some money so that they can still fund education. They didn't want to get rich at that time. Their dream was just make sure the university can run. So they leased the land to HP, to Kodak, to Lockheed Martin, and the companies with big labs and got some income. That small park called Stanford Scientific Park now become the center of Silicon Valley. So that's for the location and for the land. The third thing is really important. You must have heard the story of the Traitorous Eight. Is there anybody who never heard of the Traitorous Eight? Okay, so the Traitorous Eight. To make a long story short, Shockley, a research scientist from Bell Labs, he invented the transistor and got Nobel Prize. He came from actually California and his mother was old, so he decided to get back to this area to take care of his mother. Because he was so famous, he had very bright people, very smart people. But he had two problems. Number one, he's not a good coworker. Number two, he was not a good boss. Inside his company, used to be called Shockley Transistor, and then called Shockley Semiconductor. Inside that company, eight people decided to rebel. Those people include Gordon Moore, right? Gordon Moore was one out of eight. And another one called Noyce, he was the inventor of the integrated circuit. And another one called Kleiner, with no, Kleiner Perkins the VC. So Kleiner was one out of eight. These eight people decided to quit Shockley and they founded another company with the help of a venture investor from New York. That company was called Fairchild. It was named by the investor, so formed Fairchild family. Many people probably don't even want to work with Fairchild today, but in the 60s, it's such a great company that almost everybody in semiconductor area once worked for Fairchild. Suppose we have a summit of semiconductor companies in this room with all your guys here, about 100 people. Then you will find about 60% of those giants in semiconductor once worked for Fairchild. It's such a great company during the 60s. We don't know how many companies it spun off afterwards, but we know that there are about 92 public companies spun off from Fairchild. Also, if you sum up the market cap of those 92 companies, it's 2.1 trillion, which is huge. It's more than the GDP of India, more than the GDP of Brazil, more than the GDP of Canada, and also more than the GDP of Russia. We don't have another IBM, AT&T. We never have in this area, but we have Fairchild. So the Traitorous Eight is number one secret I want to tell you about Silicon Valley. So if you want to get success here, you have to be a traitor. There's no hope here if you work for a company for many, many years, for 20 years. You had to betray your company. Then Moore and Noyce became traitors again from Fairchild and they founded a great company called Intel. And then there came some other traitors from Fairchild saying, well, why don't we found another company compete with Intel? Then that company became AMD. So that can explain why Fairchild couldn't be another IBM because all those smart people left the company, but it really benefited our area, our Silicon Valley area. So if you look at the 60s, 1968, when was hippies? I think it's about that time. Basically, if you want to do some homework, you will find that those people have a very tight correlation in their sources. They don't want traditional stuff. They want to try something new even though they may fail. The entire area is tolerant to the failure, to the rebel, to the betray, and to the traitors. So that's something Silicon Valley is so different from other areas. So I would credit this one other reason. In other countries in the world or in other states inside US, we couldn't have another Silicon Valley. So that also explains why in Silicon Valley we can incubate one business from existing businesses. If you take a study on the history of Sun Microsystems, Cisco, Yahoo, or Google, you found that they all were incubated in existing institutions. They're not starting from scratch. So the idea is if you want to do your own business in the future, don't start from scratch. You probably want to start from existing business and work for some companies and then become the traitor, and then you will get much easier to get success. This is the secret of Silicon Valley. This is number one conclusion I want to tell you. So let me talk about non-compete agreements. This is true for most areas inside the states, but this is really hard to enforce in Silicon Valley. Why? Because if you have some disagreement between the employee and the employer, most time the law will favor the employee. So that's the advantage I think you guys really want to take in the future because you don't want to start anything from scratch and you want to take this advantage. This one happens in New Jersey. When I was in New Jersey for Bell Labs, they told me that not only everything you are doing inside your office within the eight hour working hours belongs to the company, everything you thought at home even though during your vacation when you sleep on the beach in Hawaii becomes AT&T's property. Even though you work on software, if you invented some automobile, that also becomes the property of AT&T. So during that area, you cannot imagine they could have another Silicon Valley. It's impossible. So this is also I think the secret of Silicon Valley and why we can incubate so many great companies. So this is one thing I like to share with you. So this is top one secret: betrayal and also tolerance to betrayal. I give you a concrete example. Our fund, Z Park Venture, invested on a company. It's a small company working on mobile security. Those three engineers worked for another public company called Palo Alto Networks. You know, a very famous company in security today. Do you know who was their first investor? Their CEO. Very interesting. All their colleagues know that these small guys work for their own companies inside the existing company, and nobody stopped this. And the CEO became the first angel investor to that company. Why? Because if they got died, this small company, that's fine. It doesn't hurt Palo Alto Networks. If eventually these three people got success, they had the priority to purchase it back. This prevents the existing company from pushing these three guys to their competitors. Even though maybe one day they will get very successful, they become another Symantec, but still Palo Alto Networks will get benefit from that. So this is the tolerance to betrayal. This is really important. I would say this is top one secret of the Valley. Does anybody know this is actually a jar of blue China? Does anybody know how much does it cost during the auction? It's roughly right, it's like 30 million British pounds, which is about 2.5 tons of pure gold at that time. So there are some products like iPhone or like blue China that are liked by almost everybody in the world. Some people like some product, some people don't like that one, but for some product almost everybody likes. I found that all those products, like iPhone, basically it's not a product of one person or one company, but it actually has a few because it was the product which almost embedded all the culture inside. So the hybrid culture is really important. So many companies or many countries compete with Silicon Valley to find good people to build great products and they fail. Why? Because I would say this small Silicon Valley actually reflects the entire world because we have the people from the entire world. This is really important. So this is I would say the number two secret inside Silicon Valley. Internationalization is really, really important here. Even for the small company with like 50 people or maybe only 20 people, they are international companies here in Silicon Valley. This doesn't happen in Europe, doesn't happen in China, doesn't happen in India, and doesn't happen even in East Coast, but it happens in Silicon Valley. So I would say this is probably the number two secret in Silicon Valley. The third one is rejection of mediocrity. I would say this is really important. Why? Because you cannot make a living unless you are excellent. Housing price is so expensive here, everything is so expensive here. A company cannot make a living if they are not in top three in this section or in the States. Because everything is so expensive. Let me show you this chart. The green one is the increase in housing price in Silicon Valley. You can see it's increased very fast. Who can afford that? People in these companies can afford this housing. Why? Because the market cap of these companies increased even faster. So if you want to be an entrepreneur or if you want to join a company, join a great company. If you want to be an entrepreneur, be the top three in your area. This is the growth of the housing and this is the growth of your wealth. This is the only way you can get success in Silicon Valley. So Silicon Valley is for great people and for you guys. Whenever a company is losing their profit margin, it goes down. They will either get died, get acquired by some other companies, or gradually will lose their masters, even just like SGI and HP. I would say this is not a good example. So this is conclusion I have. These are the four words I give you: have some dreams, don't be so loyal to your previous employee, you have to be excellent, and if you are boss, be tolerant to traitors inside your companies. Thank you.

So the question was to compare big company in China, Tencent, with Google. Actually, whenever you want to build some global product or international product, it's quite different to build a product to fit a particular country or particular area. So that's quite different. Give you a concrete example. Whenever you use Google Maps, one thing to do is to understand the address. That should be a quite easy program to write, right? I think you guys can do that. You can use some like Python match or your writer maybe finite state machine to do that. It's quite easy if only for English. But if you want to build a product for all countries in the world, it's quite different, and the design of the program completely different. So you have two choices. One is to build one product only for one country and then grow from there. You say, well, whenever I occupy the market in China and then go globally, this is one solution. Or from the first day, your design product for the entire world. They are different. Whenever you make decision, your fate is decided already. So what Google did was actually they designed product for everybody in the world. At the beginning, this design is very complicated. It probably even takes longer time to implement, but eventually it got benefit from that. Whenever Google launches a service in one country, in US usually, it takes only six months to go to about eight or nine major languages and configure fixes in Europe: French, Italian, German, and Spanish, and plus CJK, Chinese, Japanese, Korean, in just six months. So internationalization becomes so easy. In Tencent, it's quite different. They did an extremely good job in China in almost all aspect related to internet, but it takes them so hard to become a global company because the design of product was just to fit one country or one culture, and it's almost impossible to change the structure behind and make the global product. So that's the difference I observed between these two companies. I would say they are all great companies, but they are much different in the way they are doing things.

Back in the 1980s, actually Japanese put much effort to become multinational companies and they got success in automobile and electricity and in some area, but not in all areas. But if you look at US companies, US companies do a much better job to go globally. If you count the revenue of S&P 500 companies or US companies, more than 50% came from overseas actually. For China, I don't think this generation of companies can do an easy job to become global companies. Maybe from next generation of companies, maybe the companies funded today instead of funding 10 years back, can do the job. That's my conclusion actually. Because their culture was already there, the product were designed for only one country or one culture. It's really so hard to change it. There are only two exceptions. One is Huawei, and it's like Cisco in China, but they have global revenue, 70% of revenue came globally, only 30% came from China. Another was Lenovo actually, and no, I think by Lenovo, instead of building many international offices, so that's the only two exceptions I found. Thanks.

Sure, sure. I'll call.