Header image source | Set degree official
"Things under the iceberg, there is no easy thing."
Editor's Codex:
Here is the interview column "CEO Says" planned by Future Auto Daily. In 2022, the second wave of new car-making and entrepreneurship wave accelerated, and the first models of Youjia NV, Qingjie M5, and Jidu were successively launched; "Wei Xiaoli" frequently exerted force, Weilai ES7, Ideal L9, Xiaopeng G9 besieged the field of medium and large intelligent pure electric SUVs, and the new energy vehicle market entered the Spring and Autumn Warring States period of hundreds of controversies. Li Bin, the founder of Weilai, had expected that 2025 would usher in the decisive year of smart electric vehicles. Before that, the car companies secretly accumulated strength. With the theme of "Dormant Spring and Autumn", we will talk to the CEOs of new energy vehicle companies and how to prepare for 2025. In the first issue, we interviewed Xia Yiping, CEO of Jidu Automobile.
The day I went to interview Xia Yiping was Saturday. When we arrived at Jidu's office in Shanghai's Minhang district, Xia Yiping was pulling the middle management team to discuss in the conference room, and the atmosphere was quite warm. According to the Jidu staff, they are discussing the quality control of the first car, and discussions like this fill Xia Yiping every day.
When he was once again pushed to the front of the entrepreneurial stage, Xia Yiping became the CEO of Baidu CEO Robin Li.
In January 2021, Baidu announced the official formation of a smart car company to enter the automotive industry as a vehicle manufacturer. The new company is independent of the parent company system, maintaining independent operation, Geely Holding Group as its strategic investor and strategic partner; in March of the same year, the company was officially registered and established, named Jidu, Baidu and Geely share ratio of 55:45, with Xia Yiping as CEO.
From the perspective of secular origin, the combination of domestic Internet technology giants plus head independent car companies undoubtedly confirms that at the moment of change in the alternation of new and old energy, car manufacturing has been upgraded to another more inward war.
For Xia Yiping, the wheel of history turned around, from the automotive industry to shared bicycles, and then back to smart cars, going around for a few years, and stepping into the wind again.
Before joining Jidu, Xia Yiping was more widely known as the co-founder and CTO of Mobike. Further back, he was responsible for the intelligent networking business in traditional car companies such as Ford and FCA. From Xia Yiping's entire past experience, it seems to understand why Robin Li is so determined, Xia Yiping is the best candidate for the CEO of a new car company - understand the car, understand the Internet, understand technology, have entrepreneurial experience, neither lack of entrepreneurial enthusiasm, nor radical.
In the three years since leaving Mobike, Xia Yiping has experienced a fragmented investment career, but also experienced profound reflection. Trying to summarize the experience from a large number of books describing history and war, and reviewing the places where I did not do well in Mobike in the past.
Xia Yiping slowly confirmed one thing: from team management to corporate culture, entrepreneurship and war have a lot in common. Employees must have strong decision-making power and pay attention to the spirit of altruism. "To fight, the boss must first jump into the sea to build the bridge, and then let the brothers rush over." And never give up trying, "Mobike's early days went very badly, but we never gave up."
Although Mobike's experience Xia Yiping may not want to recall in detail, he was once in the hottest and most concerned industry in the venture capital circle, and his experience of endless battles still gave him the courage to not break the spirit.
For example, encourage automotive product manager teams to replace professional organizations to do user research. "Send them out and do it themselves, and each person talks about 1,000 users." From sitting in the office, wearing a suit and tie, to sweeping the street with a mobile phone, this is the most important Internet spirit xia Yiping learned in Mobike, "facing users can get the most direct feedback, which is much more efficient than sitting at home thinking and finding others to do user research."
After several years of entrepreneurship, the hot money handled by him is innumerable, and now although his back is against the "big tree", Xia Yiping is still accustomed to breaking a penny into two halves. More importantly, in the face of this protracted war, he realized that business is a battlefield made up of people, "the most important thing is that people are right, and the team is the greatest treasure and wealth." People are not right, and no amount of money is useful. ”
However, compared with the founding of Mobike, today's new car war, the complexity and difficulty of the exponential increase, in order to survive in this competition, you need to come up with a very competitive product, not just technology and business model innovation, and need to build a sustained competitiveness.
In the past year, Xia Yiping has raced with his team and time, with almost no respite. He travels from Shanghai to Beijing or Hangzhou every week to communicate with Jidu's Beijing-based software team and Baidu and Geely's teams. To save time, he only took early and last flights.
He lamented, "Building a car from 0 to 1 is just the tip of the iceberg." "Entering the stage of car building 2.0, Xia Yiping has to face more complicated affairs, such as coordinating production capacity, testing and verification, rationalizing the supply chain, and building a sales network... The follow-up to these cars is far more complicated than the car itself. These are the really "intimidating" icebergs hidden under the sea level. "Things under the iceberg, not a single one easy". If nothing else, Jidu's first concept car will be unveiled in April 2022 and delivered in mass production in 2023.
In 2021, when interviewed by the media, Xia Yiping said that he hoped that Jidu's first model could be produced at Geely's Hangzhou Bay plant. Now, the matter under the iceberg has finally come to a halt. Xia Yiping exclusively confirmed to Future Auto Daily that Jidu's first car will be manufactured at Geely's Hangzhou Bay plant.
Rationality is the most common way of thinking of Xia Yiping, who came from a technical background, but sensibility is far more deeply rooted in him than reason, and recalling the past can still see many idealistic moments.
Whether it was when he received the invitation of Hu Weiwei, he did not hesitate to go north to join Mobike. Still in the face of the entrepreneurial ending of being acquired, resolutely casting a "negative vote", we all saw more emotional things in Xia Yiping. "From a purely rational point of view, you can make a sum of money, but there are still feelings to start a company from 0-1." For Xia Yiping at this moment, the application of high-end automatic driving technology on household cars is the most worthy of fantasy and expectation.
The following is Xia Yiping's conversation with Future Auto Daily:
Destined to build a car
Q: You have worked for Ford, FCA, and Mobike, and your past experience is closely related to the travel industry. Why didn't you choose to join a car company after leaving Mobike, but did something that spanned a lot of time to become an investor?
A: The three years in Mobike were very tiring, and I wanted to rest for a while after leaving. Our family is in Shanghai, and the three years of doing Mobike have basically been in Beijing, and we have been away from home for too long.
In China, early investment is still relatively difficult, and the experience I accumulated in the process of starting a business in Mobike cannot be fully applied to others. So I invest, not for the sake of investing, but to get involved in these industries, to find a direction for myself, to think about what to do next.
Q: What industries were you looking at during your investment period?
A: From leaving Mobike in 2018 until 2020, I have been paying attention to two industries, one is the new energy automobile industry that I would like to return to, and the other is the application of AI and genetics and medicine.
Q: In addition to investment projects, what else did you do during the break?
A: Read a lot of books, study the places where I did not do well in Mobike in the past, and also find friends to do some review. Just like William (Weilai founder Li Bin) went to see the Long March, I also read a lot of books about history and war. From team management to cultural construction, entrepreneurship and fighting have a lot in common.
For example, in the Battle of Chosin Lake, when encountering enemy situation, we had a small detachment that directly killed it, without asking for instructions, and the decision-making power was very strong. Instead of going to war soon, you have to ask for instructions one level at a time before you can go to war. This inspired me that employees should also have strong decision-making power. I want them to be able to ask for my help when they need it, to let me solve the problems they have, I want to achieve them, our company emphasizes a culture of altruism.
Q: In 2018, the time you left Mobike coincided with the rise of the first wave of new car-making forces.
A: I was very busy every day at Mobike, basically devoted to my work, and did not pay special attention to what was happening outside. So after coming out of Mobike, it feels like the opportunity to build a new car is missed. Li Bin started Weilai in 2014, and by the time I came out of Mobike, he had been working for more than three years and had been very successful.
Q: How do you decide to build a car in the end?
A: Ren Xuyang (Baidu's chief consultant, Baidu's early entrepreneurial elder, and current chairman of Baidu Capital) found me first, if an AI pharmaceutical company found me first, maybe I would go to the pharmaceutical industry. I was recently reading a book about whether the world is random or not. Seemingly random things are actually predestined. Random events are also opportunities, just seize them. It's a lot like I joined the car-making camp.
Q: How did Ren Xuyang find you?
A: Xuyang contacted me through friends around me. The first time I found me was at the end of November 2020 and I talked two or three times later. At the time, I had no idea that Baidu was going to build a car.
In November 2020, I was planning what I was doing. By chance, One day Xuyang called me and just said he wanted to talk to me about new energy vehicles. A month later, I went to Beijing and talked to him again for 3 hours. Everyone talked quite passionately, until after the conversation, he did not mention the car building, but asked me if I wanted to meet Robin (Baidu chairman and CEO Robin Li) and communicate with him.
After that, Robin and I met to discuss the core capabilities required for the development of smart cars in the future, whether Baidu has such capabilities, and we even discussed whether Baidu should build cars, of course, he did not tell me that Baidu wanted to build cars. We probably talked about two things, the first is the future of the car, especially in the era of driverless driving, the ultimate battle is still the ecological battle. Second, we talked about why it is more suitable for car building now. If we take new energy vehicles as the electrification process of fuel vehicles to electric vehicles, I think the next stage of competition is intelligence.
It wasn't until the end of December 2020 that I saw Xuyang again, and he asked me if I would like to become the CEO of the newly formed company? I was passionate about the smart automotive industry, so I conformed to my original intention and agreed to it without hesitation. I didn't even give it to my wife, and I'm sure she'll support me.
Set degree for software validation and pre-development software integrated analog model SIMUCar
Source: Jidu Official
Q: Were there any other career options at the time?
A: After coming out of Mobike, many companies came to me. But with this entrepreneurial experience of Mobike, I really want to do something that has an impact on society.
I'm not a professional manager. For me, to do something, first of all, I have to be very passionate and have a original intention, if there is no motivation, I think I will definitely not do it well. In addition, I myself also like and appreciate companies that have a relatively large influence on society.
When Xuyang asked me if I would like to be CEO, I was still very confident in myself, why not give it a try? Just like when Hu Weiwei told me to do Mobike.
Q: There are still some doubts about Baidu's car manufacturing, and some people think that Baidu Apollo technology came out very early, but it has not been mass-produced. Do you find this a very challenging thing to do?
A: In the early days of Mobike's entrepreneurship, our families would question, "Why do we want to do shared bicycles, are they not stolen on the street", and even when we recruit people in the early days, we will be questioned by candidates. Mobike's early days were very unsmooth, but we never gave up.
Mass production of technology is a challenge. But where there are challenges, there are opportunities. Apollo and Jidu have different missions and solve different problems. Apollo's route is to verify unmanned autonomous driving capabilities and solve commercialization problems; while Jidu is for mass production and C-end, with the goal of high-end automatic driving product experience, re-integration and development verification of Apollo's high-order automatic driving capabilities, and output the high-order automatic driving experience to users. Apollo's technical ability is already very strong, and it is the core competitiveness of Jidu to implement this ability on the user side. It is worth imagining and pursuing, and it makes sense for everyone to do this.
Q: What adjustments do I need to make to Baidu Apollo technology to achieve mass production?
A: Baidu Apollo pursues fully driverless, mass production passenger cars are equivalent to a "safety officer" behind the steering wheel, and the safety level pursued by the two is not the same. Baidu Apollo is the highest configuration, applied to the mass production car needs to be cut, to see which functions are the most important to the user, how to achieve it, and the experience is better, in this range, the hardware configuration should be how. Mass production is really about finding a balance between function, cost and final implementation.
What's harder is under the iceberg
Q: There is still a period of time between receiving the INVITATION from the CEO and being officially announced. What are you doing during this time, and what are your plans for building a car?
A: I didn't immediately think about what model to make, but before March 2, 2021, the Baidu team and I started talking about the product. One thing, I was pretty sure of, was that we had to hire a designer to draw the product we wanted. Since February, I have been in contact with several designers.
I spent a lot of time thinking about what kind of designer I needed. According to the convention, most car companies will invite some more senior designers, or directly from Bentley, Ferrari and other foreign car companies to recruit people. But for us, with the blessing of Baidu's core AI technology, this designer must be able to understand the impact of technology on design, transform technology into design language, and be young and have a special understanding of Chinese users. Although some foreign designers can design very good, they will still carry the DNA of the previous brand.
Jidu design team at work Source: Jidu official
Q: After the designer is in place, what other aspects need to be completed?
A: I once drew three circles to connect the brand, the product, and the design. The idea conveyed by the brand, and the brand feeling conveyed by the product, as well as the design, must be one. These three positions are also the first ones I recruited.
Although I am from the CTO, I still have high requirements for design and branding. I've been thinking that I'm not simply looking for a designer, just building a car and sticking a logo, I hope to really build a brand.
The product definition is the upstream of R&D, and I will slowly build the R&D team. Let's do it first, and then I'll think about what kind of car I want to build.
Q: After Officer Lei announced the car, the first thing he did was to go to the major car companies to learn from the experience, have you ever done this?
A: After all, I have worked in a car company before, so I am still familiar with it, and for unfamiliar things, I rely on chat and interviews to make up for it. After entering March 2021, I interviewed more than 200 people in a month, sometimes from 9 a.m. to 9 p.m., and I couldn't even eat, that is, interviews or chat with people.
Q: When did the idea for a car robot come up with?
A: I thought about it on the first day of Jidu's establishment. When a car has a very strong ability to drive itself, voice and intelligence, it is essentially a robot.
Q: Some netizens commented that the concept of car robots is too advanced, and it is worried that the user market acceptance will be affected.
A: Someone has to be the "first person to eat crabs." Robots are the concept of the product, but it's not that I really want to be a robot like Optimus Prime and Bumblebee. At least, the concept of our entire product feels different from that of a traditional car.
What are the representations and characteristics of the products of the automotive 3.0 era? I think it's based on high-level autonomous driving capabilities, and the most typical ability of robots is to be able to walk on their own, which can distinguish us from other people's (products).
Q: Jidu's first car will be mass-produced in 2023. But the high-end L5 automatic driving, next year is certainly not possible. How to explain to the user that the design of The set degree is a car robot, but it cannot be fully realized, that is, it cannot be applied in life?
A: Very high-end does not refer to L5, but relative to the current market, we can do the relatively leading technology output and product experience.
In terms of software, the first version of our first car can already be done: in the city, when the user uses navigation, the vehicle recognizes that the red light can automatically stop; the navigation requires a left turn, and the car will drive to the left turn lane by itself (that is, automatic driving according to map navigation). This is much more complex than the user only using assisted driving on the highway, and the complexity of the scene is much higher.
Therefore, when the product is on the market, users can directly obtain our main product capabilities, including high-level autonomous driving experience, rather than being upgraded through OTA in the later stage, and our products also have continuous iteration capabilities.
Q: That is to say, there is no benchmarking product yet.
A: To say no, it's a little too proud. But I am still confident that I can make this product a benchmark smart car product.
In the past, many automotive companies' parts and intelligent systems would find external suppliers to develop, and there were not many AI capabilities placed in the car, including speech recognition capabilities and robotaxi capabilities.
This is because in the robotaxi under the emphasis on the level of security, the requirements for computing power are very high, and this computing power in the automotive field, with the current mass production chips, there is no combination that can meet their requirements. To achieve mass production, it is necessary to consider factors such as cost and function realization.
We have two of the most powerful chips in the automotive industry, and they can be mass-produced. One is the Qualcomm 8295 chip. Its AI hash rate is 8 times that of the 8155. The other is the NVIDIA Orin chip. Such computing power can help us achieve, take Baidu Apollo's algorithm, re-optimize and customize the development on the set of degree vehicles, which can be said to be a dimensionality reduction method.
In the past, many AI technologies could only run in the cloud, but now we have made localized deployments, and many AI problems have been solved, which has greatly improved the user's experience in the car.
Q: Now Tesla and Wei Xiaoli are playing the intelligent card, such as the Weilai supercomputing platform ADAM is also equipped with NVIDIA Orin chip.
A: I'm still focused on what I'm doing. They are all very good, and at the same time, I am very confident that Jidu's products will be done well, and in the field of automatic driving, Baidu's technical strength is still one of the best.
Q: What was the target user of the first car?
A: Our main focus is still intelligence, so we have to find that group of people - like technology, and design with a sense of technology, willing to use autonomous driving, willing to try new things. More like the early users of Apple's mobile phones and DJI drones, those who were the first to perceive smart technology products.
Q: The cost price of Apollo unmanned vehicle is 480,000 yuan, how do you price it?
A: The problem that Apollo wants to solve is not the same as the problem we want to solve, the customers we face are not the same, the pricing will definitely be lower than Apollo, and there will definitely be surprises when the price is announced. The price of the car actually affects sales to some extent, so we will also pay special attention to cost.
Q: Why is the pace of R&D so fast?
A: Because the two roads of sludge model (vehicle development) and SIMACar (software development) are running at the same time.
In this part of the sludge model, there are no shortcuts and follow the strict car-making process. What SIMUCar means is that we've advanced the validation of the software. From the time SIMUCar came out to the first closed loop, it took us 5 months. This is also the direction of smart car development, that is, the decoupling of software and hardware. If you follow the traditional car development method, when the vehicle is developed to a certain node, it will start software verification, and then do software and hardware integration, the whole cycle will be very long.
Now, in the process of vehicle development, we have begun to do software research and development and verification synchronously, and when the vehicle is developed to a certain extent, the maturity of the software will become better and better, and the verification of automatic driving capabilities will be advanced, which also means safer. This is also the reason why we can directly bring users a high-level autonomous driving experience when the product is on the market, because many capabilities and functions have been developed and verified in advance.
Clay model of the first model of Jidu Automobile Source: Jidu Official
Q: What do you find most difficult to do so many things at the same time?
A: The difficulties at the software development level are challenging but can still be carried. A while ago, a friend and I sighed, "The matter of building a car, it feels very difficult for laymen to see." But it seems to me that building a car from scratch is only the part of the iceberg that is exposed above sea level. Under the sea, including factories, capacity coordination, supply chain management, building sales channels, and delivery centers, these things are even more complex than the development of the car itself.
Source: Visual China
Q: "Things under the iceberg", which one is the hardest?
A: It's all very difficult, and there is no easy way to build a car. Even the choice of logo is difficult, we designed more than 500 logos, from the design, to take one by one to the car to see, to try, but also to consider the use of different scenes, we also found a group of post-90s employees internally, listen to their opinions.
Q: Has the "things under the iceberg" all started to move?
A: In 2021, my focus is mainly on building an automotive R&D team. This year's focus is on the backstage, including everything we just talked about.
Now that Jidu's first car is about to be manufactured, manufacturing and supply chain management are naturally the highlights. In the past two years, the supply chain of automobiles has been very tight, and we are also planning in advance, and even some parts need to be prepared in advance and ordered in advance. The same is true for opening a store, everyone is choosing a site to open a store, there are fewer and fewer places to choose from, and you must do some work in the front.
Q: Since 2021, the lack of core and battery is the norm in the automotive circle, how did you solve it?
A: Some preventive work will be done at present, after all, the first car will not be available until 2023. However, the lack of cores not only affects mass production, but sometimes also affects test parts. On some specific chips, I have a team dedicated to keeping an eye on the goods, updating the latest situation every day, such as a part A sample difference of three chips, they will look everywhere. At present, the battery supply situation for our 2023 models should not be worse than this year.
The CEO of the door-slamming
Q: Car building is a money-burning business, but behind Jidu there are two major shareholders of Baidu and Geely. Are you worried about funding?
A: It's too expensive to build a car. Inside the set, everyone knows that I am very picky, and I break a penny in half. My principle is that no matter how much money must be spent, I will spend it; I will not spend even a penny of money that is not necessary.
For example, many companies will spend money to find professional institutions to do user research, and the concentration will not be. I have so many product managers, sending them out to do it themselves, each talking about 1,000 users. This is the only way to do a good job.
The most important Internet spirit I learned at Mobike is "no face." On the first day of Mobike's shared bicycle launch, I want to take the App to the street to find people around Amway, please install one of our Apps.
Before are sitting in the office, wearing suits and ties people, at first there must be some embarrassment, but then began to become addicted, when you face the user, in the process of user installation App, you can observe the other party's operation process, what problems will occur. This is the most efficient and direct feedback, much more efficient than sitting at home thinking and finding someone else to do user research.
Q: Jidu has raised US$700 million so far, where has it been spent?
A: It's mainly the development of the first car.
Q: These two rounds of financing are still the investments of the previous two major shareholders, Baidu and Geely. Previously, you said "50 billion yuan invested in 5 years", where does the follow-up money come from? Will there be new investors coming in?
A: I haven't thought about the next thing yet, but the scale of 50 billion will definitely be there.
Q: Li Bin said that the threshold for building a car is 20 billion, and now that you are personally going to do this thing, do you have a personal feeling about this amount?
A: I think the first car was built, went public, and sold a little better, at least 10 billion yuan less. Automobiles are still a relatively asset-heavy industry.
Q: Recently, Wei Xiaolidu was successfully listed in Hong Kong. Jidu also has a company in Hong Kong, is this a plan to go public (to Hong Kong)?
A: For us, it still depends on the specific stage of the company's development, and whether it needs to do this thing. Step by step, when the company operates, there are some corresponding actions.
Q: Where was Jidu's first car made?
A: Basically in the Hangzhou Bay factory, it is Geely to help us to produce.
Geely Hangzhou Bay Production Base Source: Visual China
Q: Will the subsequent models be produced at Geely's factory?
A: I haven't discussed this detail yet.
Q: In addition to the vast architecture, how much does Geely participate in Jidu Automobile? How does it work? Will there be a special matchmaking team for them?
A: We don't just give the car to Geely to manufacture, we and Geely are fully connected in many aspects, including parts procurement, vehicle manufacturing, quality control, and production planning.
Our job is not just to dock, the two teams are deeply integrated, like a team working together. In terms of division of labor, all products, development, and user experience are all self-defined by the set degree. After the definition is completed, the vehicle part will be jointly developed with the Geely team; the software part, Jidu, also has its own autonomous driving and intelligent cockpit team, but will cooperate with the Apollo team. We are actually doing a good job of division of labor, which things are done in a concentrated degree, which they do, and everyone working together is a team, highly coordinated.
Q: What is the frequency of docking? What if there is a clash of cultures?
A: Every day is docking, and the driving force and goal of solving problems are the same. There are certainly conflicts in the way we work, but even if there are different ways of working between different teams within us, there will be problems of team integration. But people are still more open and inclusive.
Maintaining efficiency is a top priority
Q: Did you set a goal for HR on Programmers Day last year to have 1,024 family members (i.e. employee size) by the end of 2021? How many people are there at present? What is the proportion of technical talents?
A: It's a little bit worse, but not much. Technical personnel account for about 80%.
Q: Compared to last year, how often do you recruit people now?
A: For people of a certain level, I still have to interview in person, about 10 hours a week. I now spend relatively less time recruiting people externally, and most of my time is still on internal teams, solving some team efficiency problems.
Q: How many individuals are reporting directly to you now?
A: About 20 people reported directly to me. But in fact, my office is with everyone, there is no independent office, and everyone can find me directly.
Our company's structure is relatively flat, and I don't spend a lot of time managing it because we are very uniform in the implementation of OKR. The CEO reports to everyone every month, and everyone in the company knows that our biggest goal is to do something.
In 2021, every two months, I will pull a certain level or above of management, about fifty or sixty people, through OKR, align with each other. Basically there will be a day of meetings, from morning to night, everyone goes up to say their OKR, strong alignment. So I don't have to worry, after okr is finished, everyone can go down and do their own thing. When the company was first founded, we would come in about 100 people every month, and we had to force the pull, otherwise the company would be very inefficient. This year, we are very familiar with the operation and execution process of OKR, and are currently aligned once a quarter.
Jidu 1024 Programmer Day, Xia Yiping in a speech Source: Jidu official
Q: Is it the norm to pull mid-level meetings on weekends? What are the main communications?
A: It is the norm, mainly discussing the working methodology. For example, today (the day of the interview) we mainly solve two things, the morning meeting is to solve the test problem, and the afternoon discussion is the quality problem. Quality has its first principle, but how to go through the process, different companies are different, for us, we also need to grind out our own set of methodology.
I pay special attention to team effectiveness. At any time, as long as the efficiency of the local team is reduced, I will solve it as soon as possible. In the OKR of Jidu, it is written, "As the company's team continues to expand, the culture cannot be diluted, and the efficiency of the organization cannot be reduced."
Traditional car companies have a position called product marketing (product marketing position), the reason for this position is that the marketing department students do not understand the technology, and need to have such a role to communicate between the two teams. In order to improve organizational efficiency, in Jidu, we eliminated this position, and the product team can communicate directly with the marketing team what the product should be. The existence of the middle layer is an obstacle, and to improve efficiency, it is necessary to "go to the middle layer".
Q: If someone is inefficient, will they be fired? Are you a wolf manager?
A: I'm wolf-like, but I also focus on human nature. For people who don't fit the culture of the company, I will be ruthless. There are some people who can't adapt to the speed of the set team, and they just left after coming.
Many of our teams are very hard, like the design team and the vehicle development team last year do not know how many nights to stay up. But we don't force overtime, and our company's culture emphasizes a high degree of self-driving and efficient collaboration. My request is simple, goal-oriented, and don't delay the progress of the entire department because of individuals.
Q: Does anyone dare to argue with you in person?
A: One of the core cultures of our company is to dare to expose problems and dare to lead the fight. For example, if the budget is high, in other companies, employees may take the initiative to reduce it because the boss has objections, but in the set degree, employees will definitely run over to explain why it is high, as long as the explanation is clear, I can approve the budget.
I prefer to study war and military strategy, and pay attention to the spirit of altruism. To fight, the boss must first jump into the sea to build the bridge, and then let the brothers rush over.
Q: Jidu has set up its headquarters in Beijing. What is the function of the Beijing team?
A: The Beijing team is still dominated by the software team, compared with the software team in Shanghai, it is more inclined to the backend of the Internet of Vehicles, such as the system of the sales store, the construction of mobile apps, etc.
Q: How is the team building in Beijing? How often do you go to Beijing?
A: The Beijing team currently has nearly 200 people, and this year will form a large number of hundreds of people.
I go to Beijing almost once a week. He often flew from Shanghai to Beijing at 7 a.m. and back to Shanghai at 7 a.m. the next morning. I usually only take the first flight and the last flight, and I don't waste the intervening work time, so even if I leave from Beijing, I can still come to the Shanghai office at 10 a.m. for a meeting.
Xiaomi is not a race against
Q: Jidu and Xiaomi front and rear foot official announcement car, Xiaomi also gives priority to recruiting software talents, self-developed L4, Xiaomi is a competitor of Jidu?
A: From the perspective of product positioning, it should not be a competitor, we do not need to develop self-driving from 0 to 1, Baidu has been developed for many years, it is very mature, and our job is to mass-produce it. We very much welcome Xiaomi car.
In the era of smart car 3.0, mobile phone companies have inherent advantages in developing smart cars. From the perspective of technical architecture, the soft and hard decoupling of smart cars is like a mobile phone, which is essentially a smart phone. Under this trend, the biggest challenge is to integrate past experienced automakers into the (mobile phone) culture to create a team that is consistent in the definition of the product and the direction of technology.
Q: What other tech companies do you expect to build cars?
A: I very much wish Apple had come in very early. Apple will eventually estimate that they will still build cars, but they will definitely build a very perfect product. It is not ruled out that one day Google will also build a car.
I want more and more tech companies to build cars. After this type of company comes in, it will bring the competition of the entire industry to another dimension, not the PK of a simple car, but the competition of intelligent products.
Q: Are you afraid of stronger competitors?
A: No, the market for smart cars is very large, and there are not many players who build cars based on technology. In the future, there will be more and more such people, which will force the entire intelligent automotive industry to iterate and develop faster, which I think is a good thing.
Q: Will traditional car companies die in future competition?
A: Don't be too pessimistic, it doesn't have to be dead, it may exist in another way. Players in this industry are reluctant to do zero-sum games, and no company can build a product that will eventually sell 50 million vehicles. The manufacture of automobiles is still very complicated, there is no one who kills whom, and who does not integrate with the industry. We must jointly make the cake bigger, promote the evolution of high-level autonomous driving, and promote the construction of (related) regulations.
Q: In 2020, many new car companies have died, are you worried that Jidu will die?
A: Not really, I'm very confident now. I think the most important thing is that people are right, our team is the biggest treasure, the biggest wealth. People are not right, and no amount of money is useful.
The author | Wu Xiaoyu Wang Yan
Edit the | Li Huanhuan