ok-To The Moon - Tiger Brokers Online Investing Community

1
Comment
Favorite

wormsccc
2021-05-18

What is the story of a new Baidu?

百度回港上市现场，遇到跟踪百度多年的极客公园的创始人张鹏，问他怎么看。张鹏说：“这是百度一直通过技术创造价值的第二次体现，可以理解为百度在互联网搜索技术之外，AI第二技术曲线完成了一次上市。” 当晚，

Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.

Report

Comment

Top
Latest

No comments yet

Most Discussed

{"i18n":{"language":"en_US"},"data":{"magic":2,"id":194883026,"tweetId":"194883026","gmtCreate":1621353587465,"gmtModify":1704356353768,"author":{"id":3571822975434375,"idStr":"3571822975434375","authorId":3571822975434375,"authorIdStr":"3571822975434375","name":"wormsccc","avatar":"https://static.tigerbbs.com/3aa76875645103bf18f2faa96838defd","vip":1,"userType":1,"introduction":"","boolIsFan":false,"boolIsHead":false,"crmLevel":11,"crmLevelSwitch":0,"individualDisplayBadges":[],"fanSize":14,"starInvestorFlag":false},"themes":[],"images":[],"coverImages":[],"extraTitle":"","html":"<html><head></head><body>ok</body></html>","htmlText":"<html><head></head><body>ok</body></html>","text":"ok","highlighted":1,"essential":1,"paper":1,"likeSize":1,"commentSize":0,"repostSize":0,"favoriteSize":0,"link":"https://ttm.financial/post/194883026","repostId":1184519409,"repostType":4,"repost":{"id":"1184519409","kind":"news","weMediaInfo":{"introduction":"纸上谈兵，分辨科技史信号与噪声。","home_visible":1,"media_name":"乱翻书","id":"1084516151","head_image":"https://static.tigerbbs.com/1c1c24ff5bb34d338f5ba94644774bba"},"pubTimestamp":1621346747,"share":"https://ttm.financial/m/news/1184519409?lang=en_US&edition=fundamental","pubTime":"2021-05-18 22:05","market":"hk","language":"zh","title":"What is the story of a new Baidu?","url":"https://stock-news.laohu8.com/highlight/detail?id=1184519409","media":"乱翻书","summary":"百度回港上市现场，遇到跟踪百度多年的极客公园的创始人张鹏，问他怎么看。张鹏说：“这是百度一直通过技术创造价值的第二次体现，可以理解为百度在互联网搜索技术之外，AI第二技术曲线完成了一次上市。”\n当晚，","content":"<a href=\"https://laohu8.com/S/BIDU\">Baidu</a>When I returned to the listing site in Hong Kong, I met Zhang Peng, the founder of Geek Park, who has been following Baidu for many years, and asked him what he thought. Zhang Peng said: \"This is the second manifestation that Baidu has been creating value through technology. It can be understood that Baidu has completed a listing in addition to Internet search technology, the second technology curve of AI.\"That night, Pan Luan made a live broadcast with several friends to talk about Baidu's secondary listing and AI stories, about mobile ecology, cloud business and<a href=\"https://laohu8.com/S/5RE.SI\">smart</a>Drive and talk about the rhythm and path of business development.Questions include:How does Baidu search respond to short video + live broadcast expression and consumption mode?Where are Baidu's opportunities in cloud computing?Which will be the most important monetization business of Baidu's intelligent driving sector in the future?Baidu's autonomous driving and<a href=\"https://laohu8.com/S/TSLA\">Tesla</a>What are the different abilities of Didi?……<img src=\"https://static.tigerbbs.com/0311a26d3913800cbeae05cf8b3f0df8\" tg-width=\"1080\" tg-height=\"1311\" referrerpolicy=\"no-referrer\">At the end of April, I turned over books and participated in the Vientiane Conference of Baidu Mobile Ecology. Baidu's latest mobile ecological strategy appears in the form of X + Y strategic layout, that is, \"horizontally expanding the user scale and vertically deepening the industry vertical category\". At the meeting, nine support plans were released for founders and partners. Good-looking videos should help settled creators realize their cash, Baidu Health should support doctors, MCN institutions, public hospitals and businesses, and Baidu e-commerce should also support brands, businesses and anchors.As an \"X\" platform, Baidu APP has three strategies: content video, search personalization,<a href=\"https://laohu8.com/S/III\">Information services</a>The latter \"two modernizations\" also directly point to ecological supply.<img src=\"https://static.tigerbbs.com/d5a74a662cb78533cd7f518703cafe83\" tg-width=\"1080\" tg-height=\"695\" referrerpolicy=\"no-referrer\">In the later interview session, I flipped through the book and asked several business leaders questions:Nowadays, the quality of content in good-looking videos is uneven. How to improve the reality of the content?Many of the second creation clips are fake vocals. How to enhance the authenticity of the content?Baidu's X + Y strategic layout, how many types of Y does it have? What are you going to do?What might be the next Y?……This article is a shorthand collation of two activities, which is intended to answer a question:What is the story of a new Baidu?1. Mobile ecosystem1.1 Is the good-looking frame video a basic information structure similar to the barrage of station B? Everyone can tag or a specific creator? What kind of thinking is the whole?Song Jian(General Manager of Baidu Short Video Ecological Platform)：The structure of frame video is a video that hopes to be the standard of knowledge, from product innovation to truly pan-knowledge. Short videos have become entertainment-oriented in recent years, and their products are more suitable for entertainment. Compared with barrage, frame videos are more objective and rational. They are knowledge supplementary information content and do not require so many emotions. We want to face a larger user group, so we are thinking about how to define other videos around our videos, which is our starting point.We definitely prioritize creators to do it, but UGC we are also trying. Now our video information structure, such as medical and health videos, says that this is diabetes. If the frame happens to be in the symptoms of diabetes, he doesn't know this thing particularly clearly by watching only one video. If this point is well supplemented to the video information, it will be very good. Whether the creator can do this or not, and whether UGC can do this depends on a set of strategies. Baidu will also provide many tools to help do this information.The law of information evolution is from low information to high information. Whether it is text recognition or from long video to short video, information is getting higher and higher. However, this \"high\" is too high in the network now, and it will catch the eye more and more through emotions. The density of information must go high, and whether it must go to the stimulation of extreme emotions is a question we are thinking about. We want to supplement the information density through frame videos, break it up, and increase the information density, but it's not all emotions, and add some knowledge. On the one hand, it caters to the trend, and on the other hand, it moves towards a more standardized degree of video.1.2 Nowadays, the quality of some content in good-looking videos is uneven. For example, some of the second creation edits are AI-synthesized vocals. How do you view this matter and how do you improve the real-life feeling of the content?Song Jian(General Manager of Baidu Short Video Ecological Platform）：Let me briefly introduce two points, how to improve this problem.One is that you should have a better community atmosphere. Many products are scientific when selecting seed authors from the beginning, that is, it is easy to form a community atmosphere slowly, but this can't go back.The second is to say, we should do more strategies for the author or UGC, so whether we are doing Duka or doing frame videos, we need new tools to have new things, so as to achieve differentiated things. The second one is done by us on our own initiative, otherwise it is impossible to make great changes in the ecology, for example, when there is no advantage in volume, so consider it from the perspective of tools (product innovation).1.3 To attract more users or expert teams to choose Baidu platform, what are the unique advantages of Baidu Health?<img src=\"https://static.tigerbbs.com/38e7a2b4434acc3103770af068b02bf6\" tg-width=\"512\" tg-height=\"341\" referrerpolicy=\"no-referrer\">Yang Minglu(General Manager of Baidu Health): The first mass base. Who came to mind first when encountering this situation? Baidu has this mind, which is mainly manifested in the volume of users. We have a very important mass base.The second is technical strength. Baidu's technology competition is very strong. Based on big data, it makes accurate crowd portraits and empowers AI. I believe that soon, everyone will be able to see some very obvious changes, especially in precise doctor-patient matching. Recently, it has accumulated many patents. This is a powerful technical ability.The third is the strong doctor operation ability based on the user base. Why do doctors come to the platform? The big reason is that there are users, but will he stick, because he has to operate many scenarios. What Baidu Health provides doctors is actually a full-scenario platform, including detailed doctors' personal encyclopedia business cards, suggestions on what videos to do in Baijiahao, online consultation, live broadcast, etc. Baidu is a one-stop all-round online practice venue. Therefore, it is very important for doctors to form a link between these things in one place. Therefore, operational capabilities and full-link services are very important points for us to attract more cooperative institutions and doctor groups. The most important thing is to grasp C and B, which is a very important sword for Baidu Health to do business today.1.4 What are the manifestations of the service-oriented Baidu mobile ecosystem that we can perceive?Li Tianhua(Former mobile Baidu architect)：Among Baidu's more than 30 million daily active services, there is one network. Let's talk about four networks: information network, e-commerce network, industrial network and social network. Baidu is also deepening the industry network. Their own love purchasing has already gained income. There is also a lot of room for independent listing when it is dismantled, and its annual income is more than double digits. This unit is 100 million. Therefore, I think if Baidu is really willing to give talents more room for growth and release more equity for them to take the lead independently, I think Baidu can do a lot.Love purchasing, the B to B ecosystem, before<a href=\"https://laohu8.com/S/BABA\">Alibaba</a>Billions of revenue, at least 3.5 billion of revenue is brought by Baidu's traffic acquisition. In the past history, Baidu only paid 5.6 billion advertising fees. Later, Baidu felt that I might as well do this service myself. In fact, this logic is right. B2B Alibaba also provides an intermediate service logic. This logic is very suitable for Baidu, so Baidu did it itself. Now Baidu has done a good job itself, which has also brought double-digit growth and over 100 million revenue.1.5 Is it too late for Baidu to make efforts in e-commerce now? How does Baidu plan to make efforts and how to differentiate?Ping Xiaoli(Vice President of Baidu, General Manager of Baidu APP)：First of all, let me answer why I cut into e-commerce. I think there are a few changes in business trends that we have observed here.First, the infrastructure of the Internet is becoming more and more mature. In terms of payment process, it is no longer exclusive to e-commerce. Second, the supply side of e-commerce platforms is constantly enriching, which should be said to be extremely enriched. Now the supply chain side is actually completely more open and open, and it is also very open to all platforms. Third, in terms of users' purchasing habits, today's purchase, in fact, many consumption decisions are stimulating and rational, and the scene is stimulated. Users may have shopping needs anytime and anywhere. This kind of shopping demand, with the improvement of Internet infrastructure and the great enrichment of supply side, makes it possible to buy at any time on all occasions.Secondly, we think it is not too late for Baidu to cut into e-commerce, because we have these three basic conditions, not only Baidu, but I think many scenarios may become a possible purchase channel. That is to say, in fact, the phenomenon that China's Internet e-commerce companies are greatly concentrated at the top will gradually change to some extent.Furthermore, whether other platforms have advantages or not, when we do such things, we not only see whether the industry has advantages, but also look at it in combination with our own products. In fact, a large number of users search for product category words and brand words every day on \"search\", such as which washing machine brand is good to search for. There are so many such words. Why do we see so much demand? It hasn't died out for so many years, because we see what the user wants to check in the whole process. In fact, he needs to know better about this brand, this product and how it is, including some of its main parameters, official evaluation and user experience. That is, he will go to Baidu to search before the whole consumption decision-making pre-link. Let me give you a very simple example. You couldn't buy medicines online before, but now you can. It turns out that everyone can buy medicine. We see that many users actually want to search on Baidu, such as searching for a certain product for a long time. Why can't they place an order directly? This search is more like three C products and furniture products, and some decision-making users with higher unit prices will be more willing to search, so we think we have certain characteristics here.Therefore, for example, when we search for e-commerce, we focus on building two scenarios, one is the search scenario and the search for e-commerce. In fact, this is also based, which is the basis just mentioned. The second is that from the perspective of the key direction, the key will be more, which may not be the same as other platforms, that is, we are all positively doing some homogeneous competition, and we are doing more of our venues. Our purpose is not to compare with others. Today's e-commerce is a service-oriented strategy. Baidu has no battlefield. It is not this logic to be e-commerce. The whole logic means that from \"content to service\", the service-oriented strategy is a very important part, and buying goods is a part of the service. The purpose of our e-commerce is more about how to better meet users' needs in this place, dig out users' needs, and do a good job in users' real shopping needs. If we do it in this kind of scenario, we are more inclined to the user's relatively rational role, and let him choose better products, brands, and explosions. These things are our main tasks. When we do it, we will aggregate more product content, go to service purchase channels, and give more information to users, so that they can make a comprehensive decision regardless of seeing the price or protecting their rights and interests.Another scene is \"live broadcast\". Live broadcast e-commerce is also a product form that we are seeing growing very fast now, because Baidu actually has search, recommendation and information flow, and the information flow has also changed greatly from graphics and texts to live broadcast. change. It is relatively typical, and it is more suitable for the precipitation of popular products in various activities. If live broadcast and e-commerce are precipitated together, it will be very smooth. For example, some merchants put their brand stores directly in Baidu, search for one-click direct access, put their most complete stores here, and then they can do some activities with some goods, cooperate with Baidu stores, and broadcast two live broadcasts. Let users have some node-like cognition, such as the live broadcast with wine some time ago. In fact, after we finished the live broadcast that day, the gmv was very high that day. After that, many people searched there and then went back to that store to buy. Therefore, it is a scene that can be completely taken together, so this is an opportunity we see, and it is also some of our ideas that we have been exploring and verifying in the past period of time.1.6 Baidu's X + Y strategic layout, how many types of Y does it have? What are you going to do? What might be the next Y?<img src=\"https://static.tigerbbs.com/e1162a7bad64ba3e26a0d37fbb04772e\" tg-width=\"1080\" tg-height=\"462\" referrerpolicy=\"no-referrer\">Ping Xiaoli(Vice President of Baidu, General Manager of Baidu APP)：There are probably three types of Y, but good-looking is different from healthy. Good-looking is X, and healthy is Y. However, these three categories are basically summarized into two categories, which are internally called Y1, Y2 and Y3. Y1 and Y2 are strongly related to X. Y1 and Y2 take advantage of the advantages of X platform to give full play to their industrial advantages, and the space market is relatively large. Here, choose the platform with key depth, and the first choice is the strategy of extending the content of X platform to service. Second, on the Y track, you can have a certain market share and some of your own positions, which is Y1+Y2. The internal difference between Y1 and Y2 is the difference in size, the health is Y2, and the small Y1 does not need special explanation here. Y3 doesn't have much strong connection with X. For example, network disks and games are placed in Y3, which are all separate industries.The vertical category of Y is how to do better and more prosperous, including how to do poorly. In fact, we have a set of shutdown and transfer mechanism in the whole incubation, that is, to see the progress. Of course, if we can't do it, we will definitely shut down some decisively. We do more than a dozen or twenty vertical categories internally, and some teams only have a few people to explore first. The general mechanism is like this.How to be better and more prosperous, I think it is very important that every vertical category should find someone who understands this industry. In Baidu, we are on the platform ourselves. I understand the ability of the platform very well, but each vertical category has different styles of play. However, you can do it yourself. If you don't rely on Baidu's advantages, you may not be able to do it according to the entrepreneurial method. Because you will compete with others in a homogeneous way, you need a combination of the two. Just like when I had a discussion with my colleagues earlier, I told him that in fact, you don't know how big a gold mine search is. Outside, everyone thinks that the traffic should be quite good and the value is quite high. This is still a very open cognition.Search can gain insight into consumer demand. For example, it is also a trend for fashionable products now, and many of these things can be fed back in users' minds in a timely manner. How to undertake this kind of user's needs better? If you only search for products, it turns out that it literally matches the corresponding answers, but the industry is not. You have to see what these questions mean to the user, what he needs, and then combine the products of the industry you designed, how to undertake and connect with him. Here, two teams of X and Y need to collide with the design. So we have some cooperation here, and we have some teams here that specialize in combining with them. Why is health doing well? In fact, health is that the whole search team was hatched, and later some people who understand these industries were joined.Therefore, this is also the experience of past practice. The capabilities and genes of the X platform actually helped this industry a lot in the early stage. However, we also need the input of professionals in the industry, so we will adhere to this experience and principle in human organization construction and product design. What really reflects is the differentiated competition path, not what others have done. I have to do the same as others. This is not the only idea for Baidu to do this, so we chose a lot of Y, which is strongly related to X.In addition to health, the next step may be to start the in-depth layout of Y in multiple fields of \"education, medical beauty, and automobiles\". The logic of the selected vertical category itself must first conform to the strategic logic of the big Baidu APP, which is \"content + service\" The connection point is relatively clear. First of all, the whole decision-making and logic users need a lot of content as requirements, and second, the whole back link has services in it. The third is that the needs of users in its industry itself are equivalent to that part of a certain link happens on Baidu APP, or we have certain advantages. Therefore, when choosing vertical categories, it will be selected according to Baidu's genes. Under the logic of choice, it is determined that the main task of these vertical categories is to strengthen Baidu's overall capabilities in this respect. At the same time, when comparing in other markets at the same time, I also have to look at other categories in other markets. In this case, it provides a strong spur and the requirement of survival consciousness.2. Intelligent cloud2.1 Which businesses of Baidu are likely to grow independently in the future?Li Tianhua (former mobile Baidu architect):At present, Baidu accounts for several relatively large categories in the Top80% traffic, such as education, health, games and reading, which are all relatively large categories that Baidu can make in-depth efforts. I think a skinny camel is bigger than a horse. Don't listen to the outside who always spit out that Baidu is bad and bad. There are more than 200 million daily living apps, and there are several daily living apps in Baidu with more than 30 million daily living apps, including maps, network disks, etc., which are all very good. Of course, the most basic search is how to fight the defensive battle well, and at the same time make use of the new opportunities in the market in the next 3 to 5 years, seize them, make a second spring, and AI can better serve various scenarios, drive others, whether it is automobile Industry or service industry, or smart industry, education, make greater upgrades and build some infrastructure facilities. There is also a small platform with large infrastructure facilities within 10 years, all of which are thoughtful. To realize the trend of the talent market, a large number of talents enrolled in schools every year enter Baidu's system for incubation in large quantities.As we all know, Baidu has been searching for so many years, and the mobile phone Baidu has about 200 million daily activities. Baidu's most valuable businesses, such as Baidu Netdisk, have about 31 million to 33 million daily users and about 3.3 billion annual revenue. As you can see, when Baidu went public, Ali issued a clarion call again, saying that they would also be Ali's network disk. In fact, everyone has seen Baidu's persistence in the network disk for so many years, and there is an episode in the middle, when 360 gave up the 360 network disk. This matter is a long-term value investment matter. This business is still the number one in the industry at present, and it will definitely be split and listed independently.There is huge room for several core businesses of Baidu in the future. For example, Xiaodu speakers, which are growing well now, are in charge of Jing Kun. This business is a better product in the future era of instinctive interactive experience, and there is a lot of room. The essence of the speaker business is a product in the RTC era. RTC is the logic of Real time clock, a real-time connection service. One of the biggest opportunities from today's service to the future is the RTC era. A valuable business in the RTC era. In fact, Tesla's autonomous driving is essentially an RTC service. There is a<a href=\"https://laohu8.com/S/300024\">Robot</a>Help you calculate in real time, help you see the road conditions, and help you control the car. In fact, this is the same. Duer OS business is also a product in the RTC era. It has the same interactive experience as Clubhouse. I get up every morning and ask Xiaodu Xiaodu how the weather is. Xiaodu Xiaodu, open the curtains, this in itself is the future of interactive experience. In this case, Xiaodu's independence is conducive to its growth, and it will also help it drive Baidu's stock price to continue to rise.In addition to Baidu Netdisk and Xiaodu, as we all know, Baidu's health service revenue has been good for so many years. Personally, I think Baidu Health also has the opportunity to split up, which has a huge value-added space for Baidu's stock price. My suggestion is that you can pay attention to these core businesses.Here, I would like to mention Baidu Map in particular. I think Baidu Map has done very well in the last year, and it has been upgraded than before. I don't know if you have used Baidu map to take a taxi. Everyone found that recently, Didi took a taxi, and Didi changed a new version. Everyone complained and couldn't find out how to use Didi. But you will find that taking a taxi with Baidu Map is cool and cheaper than Didi. You can take a taxi with Baidu Map, which is really easier to use than Didi. My experience these days is the same.2.2 Where is the technical threshold for cloud disk products?Pan Luan:In fact, network disk is a very high-tech thing. I can tell you one first<a href=\"https://laohu8.com/S/XNET\">Thunder</a>Examples of. In fact, the founding team of Thunder also went out of Baidu, and co-founder Cheng Hao and CTO Li Jinbo both went out of Baidu. Where is the technical cow of Thunder? Some of its capabilities are very similar to those of the network disk. Thunder needs to be crawler, transmission and storage. In fact, each of them is a very hard technology. If you have an impression, those who made network disks in the early years were not as powerful as Baidu.In fact, Baidu was the first to do distributed computing in China. Baidu's internal operation and management servers should be very strong, because it has to carry the computing power of peaks and valleys. Moreover, 10 years ago, Baidu's servers should have the largest number in China, involving machine management, resource scheduling and the whole network management. In fact, these three core technologies are relatively advanced. At first, Alibaba Cloud's architecture system was exactly the same as Baidu's service system.Li Tianhua (former mobile Baidu architect):In the middle, we should first talk about the value of cloud disk products and services, that is, the server bandwidth cost that everyone jointly dilutes. The bandwidth cost per megabyte of CDN of normal Internet companies is more than that of 7 yuan, and the cheaper ones are more than that of 6 yuan. Alibaba Cloud can generally buy 20% off in China, which is considered cheap, but few people can buy 65% to 30% off. In fact, the bandwidth cost per megabyte is six or seven yuan, which is normal, and it is also true for large companies like Bilibili. same. There is also a problem of storage cost for individuals. If there is no problem of storage cost, there is no need for sharing logic. In fact, sharing is still everyone's joint efforts to dilute the cost of bandwidth servers. There is such a logic in essence.2.3 How to define intelligent cloud? Where is the room for improvement in the future development of intelligent cloud?Li Tianhua(Former mobile Baidu architect)：In fact, the recent growth of Baidu Smart Cloud is very high, and part of the reason is that Baidu used to buy Ali for many internal businesses. Later, the boss wanted everyone to buy their own services, so he also brought a lot of income increment.Strictly speaking, Baidu Cloud still has a lot of room for improvement in its business. I look at this from both sides. Should we give some space or give each business independent development, otherwise it will also be a problem to be run by Baidu Cloud. Of course, if all the income is given to Baidu Cloud, Baidu Cloud can have a better valuation, a better cake when listed, and it can be better empowered if it is taken back from the secondary market, and the boss can also earn More, this is normal.Shi Zhong(Light Black Technology)：The entry point of Baidu Smart Cloud was chosen when there was not much space left for him. And this is also based on its own advantages, that is, AI.Many times, I don't know if you still remember the point he cut in. When Baidu first launched ABC, it actually gave an example. Using AI to help Shougang do the quality inspection of steel plates is actually a very heavy thing. It won't be like the website. I sell you resources. When you go online, I can make money like rent collection, which requires strong service and strong adaptability. The system of quality inspection in steel plants itself is very complicated, and it needs a lot of adaptation and adjustment. Including some wind power cooperation later, predicting which units will have problems, including intelligent customer service.Frankly speaking, the industry itself has many categories, so what it does is scattered. But the advantage is that once there is a benchmark in the industry, this benchmarking effect is actually very strong. It's not that easy to make money, but I think this track is the broadest, so I can only say this. But to do it well, it actually takes a lot of effort. Can this matter be done as imagined? It remains to be seen. In short, there is imagination.Pan Luan:Someone asked, many people are doing AI, and there has never been an exit visible to the naked eye. They grit their teeth, find someone, hire a team of smart speakers, and find an exit at once. I think Baidu has always been practicing dragon slaying in technology, and has always wanted to find a suitable scene. Making smart speakers, like Hangzhou Ruoqi, there is another company that makes smart speakers. I think their latest industrial-grade solutions are now using Baidu's cloud services to make their solutions. These teams that used to be speakers began to borrow Baidu's cloud services.Brother Qiushi(Smart Speaker PM)：There will be this process. These general cloud services, or general NLP services, include<a href=\"https://laohu8.com/S/002230\">IFlytek</a>, Alibaba Cloud provides it, it depends on who is more diligent in sales. In fact, Baidu has invested a lot in this area, because of its external expansion in this area, and some ISVs have profit-sharing, coupled with the policy of interest incentives. Including Alibaba Cloud, I feel the same. In essence, I feel that it is still at the stage of pre-sales selling concepts to spread the number of users. There are various demands, some of which are to improve efficiency, some of which are that you can't do it if others do it, and some places can get more government subsidies if you do such things. Therefore, the coverage ability of local push still has a relatively large impact. This kind of sales-oriented work actually accounts for a relatively large proportion in the whole, which is not particularly the same as the logic of the whole Internet. My own feeling is that the vertical field is different from the consumer Internet. The cognition of the interests of all parties in the industrial Internet and the upgraded modules are very important. This kind of knowledge reserve is not necessary for a large company. If you are willing to do it, a lot of it is dirty work, and the investment in this reserve, accumulation and collation is very large in every field. The friends of Light Black Technology have a good chat, and they are still very particular about choosing fields.3. Autonomous driving & car manufacturing3.1 How fast or slow does Baidu build cars?Li Tianhua(Former mobile Baidu architect)：Now the chairman of Baidu's car is my most trusted big brother in Cover. I think this matter should have a very good result, so everyone still has some small expectations for Baidu's car. Especially when it comes to the stock price pull, there should be a small expectation. There are generally two types of companies. In the stock market, one is to make games, and the other is to be real. If you finally make the game, then you are a very good company, and this is the logic. So don't worry too much about the short term. Many people complain that this CEO is unreliable, and that one is unreliable either. It doesn't matter, the market has enough time. The Internet generally incubates a new business. Whether it is incubated from a large platform within the system or from outside, this cycle takes about two years. Within two years, as long as he takes the key steps right, this problem is not big.3.2 What are the different capabilities of Baidu's autonomous driving and Tesla and Didi?Shi Zhong(Light Black Technology)：In fact, there are two evolution routes of autonomous driving, all of which are similar represented by Internet manufacturers. They take the route of pure autonomous driving, starting with L4, and starting with no one. You can limit the scene first, and unmanned driving with limited scenes is their start. But like Tesla, its starting point from the beginning is L2 or L3, which is assisted driving. These are two different routes. But in the end, they all want to get to L5. L5 means unmanned driving without limited scenes, that is, any scene and any road condition can be reached. But now no one knows when this matter will be achieved. As far as we can see now, what we can see is L4. L4 is everyone's goal, but L4 is also marching towards it. There are two routes, one is the Tesla route and the other is the Baidu route, which can be understood so simply.The difference is actually relatively simple. Just like cultivation, there are different methods, including the Sword School and the Qi School. It is the butt that determines the head. If Tesla has a car, it must first sell the car. It will not take driving itself as its own responsibility, but selling the car as its own responsibility. Because no one can do L5 level autonomous driving now, but they can sell cars first. Use relatively low-level assisted driving to hone my autonomous driving technology, and one day maybe I can become a pure autonomous driver. But Baidu didn't have a car when it started. When it started without a car, the final route was not to sell cars, but to sell services. That is to say, its ultimate situation may be like Didi, which is to operate a service system. He doesn't have to own a car. He only needs to buy services, and I cover all the underlying technologies. Therefore, there is no need to have a stage where users drive themselves, in the form of services such as Robotaxi.Pan Luan:Baidu's car making now should be the most concerned thing about Baidu in this round, that is, a new Baidu. The mobile ecosystem we are seeing at present, and what we are seeing further ahead are things related to the cloud. The story that is close to being realized should be the car part. The car part is also divided into several lines, such as providing services to these car manufacturers and helping the government build the Internet of Things system. Just like Baidu's smart clouds now, many of them should be tied to the government's large transportation system. Of course, there is another piece of self-driving taxi service in the future. But when we discussed Baidu Maps earlier, consumers actually didn't care whether your technology was good or not. What they cared about was how to send me there. At that time, they had to consider whether the vehicle density was enough or not.You can think of it as a service provider for car manufacturers. Does everyone think this is my core interest? It was rumored before that Baidu wanted to vote<a href=\"https://laohu8.com/S/LI\">Li Auto</a>, must use<a href=\"https://laohu8.com/S/APOL\">Apollo</a>If you don't have the ideal system, you choose to build it yourself. I think at present, powerful car manufacturers may choose their own systems that can have stronger control, and invest in or acquire autonomous driving<a href=\"https://laohu8.com/S/V03.SI\">Startups</a>, but they will not hand over the entire autonomous driving solution to Baidu. They may purchase some specific services of Baidu, such as high-precision maps.3.3 Is there really an opportunity for the big Android system that Baidu once envisioned in the automotive industry? Can Apollo become the Android of the automotive industry?<img src=\"https://static.tigerbbs.com/d4f13dcf2d14ce5aec837d5bddc0197d\" tg-width=\"1080\" tg-height=\"483\" referrerpolicy=\"no-referrer\">Classmate Chen(Baidu Autopilot Business)：In the intelligent transformation of the automotive industry, Baidu Apollo hopes to use more deterministic technologies and solutions to accelerate the boarding of intelligent technologies for OEM customers in a limited time window, so as to prevent everyone from repeatedly building wheels.First of all, from the perspective of enterprise R&D investment costs, if we consider developing a multi-model platform that supports intelligent driving and includes E/E architecture, the investment cost is around 10-20 billion; The investment in technology research and development of autonomous driving software is also more than tens of billions, and a cumulative sales of 100-200W units are required to achieve cost amortization.Secondly, in addition to the high cost investment, the time cost of autonomous driving software development is also huge, and it needs to go through the process of capacity building, forming effective R&D capabilities, and then productizing output. It usually takes at least 6 years for companies in the industry to independently develop autonomous driving software. Most importantly, research and development costs do not guarantee the availability of available technical results, including test licenses, maps, V2X road test resources, intelligent clouds, etc., which will affect the level and landing time of autonomous driving technology.Apollo hopes to use its eight years of technology accumulation to help car companies build competitive smart cars and promote the development of the entire industry in the form of open cooperation.","collect":0,"html":"<!DOCTYPE html>\n<html>\n<head>\n<meta http-equiv=\"Content-Type\" content=\"text/html; charset=utf-8\" />\n<meta name=\"viewport\" content=\"width=device-width,initial-scale=1.0,minimum-scale=1.0,maximum-scale=1.0,user-scalable=no\"/>\n<meta name=\"format-detection\" content=\"telephone=no,email=no,address=no\" />\n<title>What is the story of a new Baidu?</title>\n<style type=\"text/css\">\na,abbr,acronym,address,applet,article,aside,audio,b,big,blockquote,body,canvas,caption,center,cite,code,dd,del,details,dfn,div,dl,dt,\nem,embed,fieldset,figcaption,figure,footer,form,h1,h2,h3,h4,h5,h6,header,hgroup,html,i,iframe,img,ins,kbd,label,legend,li,mark,menu,nav,\nobject,ol,output,p,pre,q,ruby,s,samp,section,small,span,strike,strong,sub,summary,sup,table,tbody,td,tfoot,th,thead,time,tr,tt,u,ul,var,video{ font:inherit;margin:0;padding:0;vertical-align:baseline;border:0 }\nbody{ font-size:16px; line-height:1.5; color:#999; background:transparent; }\n.wrapper{ overflow:hidden;word-break:break-all;padding:10px; }\nh1,h2{ font-weight:normal; line-height:1.35; margin-bottom:.6em; }\nh3,h4,h5,h6{ line-height:1.35; margin-bottom:1em; }\nh1{ font-size:24px; }\nh2{ font-size:20px; }\nh3{ font-size:18px; }\nh4{ font-size:16px; }\nh5{ font-size:14px; }\nh6{ font-size:12px; }\np,ul,ol,blockquote,dl,table{ margin:1.2em 0; }\nul,ol{ margin-left:2em; }\nul{ list-style:disc; }\nol{ list-style:decimal; }\nli,li p{ margin:10px 0;}\nimg{ max-width:100%;display:block;margin:0 auto 1em; }\nblockquote{ color:#B5B2B1; border-left:3px solid #aaa; padding:1em; }\nstrong,b{font-weight:bold;}\nem,i{font-style:italic;}\ntable{ width:100%;border-collapse:collapse;border-spacing:1px;margin:1em 0;font-size:.9em; }\nth,td{ padding:5px;text-align:left;border:1px solid #aaa; }\nth{ font-weight:bold;background:#5d5d5d; }\n.symbol-link{font-weight:bold;}\n/* header{ border-bottom:1px solid #494756; } */\n.title{ margin:0 0 8px;line-height:1.3;color:#ddd; }\n.meta {color:#5e5c6d;font-size:13px;margin:0 0 .5em; }\na{text-decoration:none; color:#2a4b87;}\n.meta .head { display: inline-block; overflow: hidden}\n.head .h-thumb { width: 30px; height: 30px; margin: 0; padding: 0; border-radius: 50%; float: left;}\n.head .h-content { margin: 0; padding: 0 0 0 9px; float: left;}\n.head .h-name {font-size: 13px; color: #eee; margin: 0;}\n.head .h-time {font-size: 12.5px; color: #7E829C; margin: 0;}\n.small {font-size: 12.5px; display: inline-block; transform: scale(0.9); -webkit-transform: scale(0.9); transform-origin: left; -webkit-transform-origin: left;}\n.smaller {font-size: 12.5px; display: inline-block; transform: scale(0.8); -webkit-transform: scale(0.8); transform-origin: left; -webkit-transform-origin: left;}\n.bt-text {font-size: 12px;margin: 1.5em 0 0 0}\n.bt-text p {margin: 0}\n</style>\n</head>\n<body>\n<div class=\"wrapper\">\n<header>\n<h2 class=\"title\">\nWhat is the story of a new Baidu?\n</h2>\n<h4 class=\"meta\">\n<a class=\"head\" href=\"https://laohu8.com/wemedia/1084516151\">\n\n<div class=\"h-thumb\" style=\"background-image:url(https://static.tigerbbs.com/1c1c24ff5bb34d338f5ba94644774bba);background-size:cover;\"></div>\n\n<div class=\"h-content\">\n乱翻书 \n2021-05-18 22:05\n</div>\n</a>\n</h4>\n</header>\n<article>\n<a href=\"https://laohu8.com/S/BIDU\">Baidu</a>When I returned to the listing site in Hong Kong, I met Zhang Peng, the founder of Geek Park, who has been following Baidu for many years, and asked him what he thought. Zhang Peng said: \"This is the second manifestation that Baidu has been creating value through technology. It can be understood that Baidu has completed a listing in addition to Internet search technology, the second technology curve of AI.\"That night, Pan Luan made a live broadcast with several friends to talk about Baidu's secondary listing and AI stories, about mobile ecology, cloud business and<a href=\"https://laohu8.com/S/5RE.SI\">smart</a>Drive and talk about the rhythm and path of business development.Questions include:How does Baidu search respond to short video + live broadcast expression and consumption mode?Where are Baidu's opportunities in cloud computing?Which will be the most important monetization business of Baidu's intelligent driving sector in the future?Baidu's autonomous driving and<a href=\"https://laohu8.com/S/TSLA\">Tesla</a>What are the different abilities of Didi?……<img src=\"https://static.tigerbbs.com/0311a26d3913800cbeae05cf8b3f0df8\" tg-width=\"1080\" tg-height=\"1311\" referrerpolicy=\"no-referrer\">At the end of April, I turned over books and participated in the Vientiane Conference of Baidu Mobile Ecology. Baidu's latest mobile ecological strategy appears in the form of X + Y strategic layout, that is, \"horizontally expanding the user scale and vertically deepening the industry vertical category\". At the meeting, nine support plans were released for founders and partners. Good-looking videos should help settled creators realize their cash, Baidu Health should support doctors, MCN institutions, public hospitals and businesses, and Baidu e-commerce should also support brands, businesses and anchors.As an \"X\" platform, Baidu APP has three strategies: content video, search personalization,<a href=\"https://laohu8.com/S/III\">Information services</a>The latter \"two modernizations\" also directly point to ecological supply.<img src=\"https://static.tigerbbs.com/d5a74a662cb78533cd7f518703cafe83\" tg-width=\"1080\" tg-height=\"695\" referrerpolicy=\"no-referrer\">In the later interview session, I flipped through the book and asked several business leaders questions:Nowadays, the quality of content in good-looking videos is uneven. How to improve the reality of the content?Many of the second creation clips are fake vocals. How to enhance the authenticity of the content?Baidu's X + Y strategic layout, how many types of Y does it have? What are you going to do?What might be the next Y?……This article is a shorthand collation of two activities, which is intended to answer a question:What is the story of a new Baidu?1. Mobile ecosystem1.1 Is the good-looking frame video a basic information structure similar to the barrage of station B? Everyone can tag or a specific creator? What kind of thinking is the whole?Song Jian(General Manager of Baidu Short Video Ecological Platform)：The structure of frame video is a video that hopes to be the standard of knowledge, from product innovation to truly pan-knowledge. Short videos have become entertainment-oriented in recent years, and their products are more suitable for entertainment. Compared with barrage, frame videos are more objective and rational. They are knowledge supplementary information content and do not require so many emotions. We want to face a larger user group, so we are thinking about how to define other videos around our videos, which is our starting point.We definitely prioritize creators to do it, but UGC we are also trying. Now our video information structure, such as medical and health videos, says that this is diabetes. If the frame happens to be in the symptoms of diabetes, he doesn't know this thing particularly clearly by watching only one video. If this point is well supplemented to the video information, it will be very good. Whether the creator can do this or not, and whether UGC can do this depends on a set of strategies. Baidu will also provide many tools to help do this information.The law of information evolution is from low information to high information. Whether it is text recognition or from long video to short video, information is getting higher and higher. However, this \"high\" is too high in the network now, and it will catch the eye more and more through emotions. The density of information must go high, and whether it must go to the stimulation of extreme emotions is a question we are thinking about. We want to supplement the information density through frame videos, break it up, and increase the information density, but it's not all emotions, and add some knowledge. On the one hand, it caters to the trend, and on the other hand, it moves towards a more standardized degree of video.1.2 Nowadays, the quality of some content in good-looking videos is uneven. For example, some of the second creation edits are AI-synthesized vocals. How do you view this matter and how do you improve the real-life feeling of the content?Song Jian(General Manager of Baidu Short Video Ecological Platform）：Let me briefly introduce two points, how to improve this problem.One is that you should have a better community atmosphere. Many products are scientific when selecting seed authors from the beginning, that is, it is easy to form a community atmosphere slowly, but this can't go back.The second is to say, we should do more strategies for the author or UGC, so whether we are doing Duka or doing frame videos, we need new tools to have new things, so as to achieve differentiated things. The second one is done by us on our own initiative, otherwise it is impossible to make great changes in the ecology, for example, when there is no advantage in volume, so consider it from the perspective of tools (product innovation).1.3 To attract more users or expert teams to choose Baidu platform, what are the unique advantages of Baidu Health?<img src=\"https://static.tigerbbs.com/38e7a2b4434acc3103770af068b02bf6\" tg-width=\"512\" tg-height=\"341\" referrerpolicy=\"no-referrer\">Yang Minglu(General Manager of Baidu Health): The first mass base. Who came to mind first when encountering this situation? Baidu has this mind, which is mainly manifested in the volume of users. We have a very important mass base.The second is technical strength. Baidu's technology competition is very strong. Based on big data, it makes accurate crowd portraits and empowers AI. I believe that soon, everyone will be able to see some very obvious changes, especially in precise doctor-patient matching. Recently, it has accumulated many patents. This is a powerful technical ability.The third is the strong doctor operation ability based on the user base. Why do doctors come to the platform? The big reason is that there are users, but will he stick, because he has to operate many scenarios. What Baidu Health provides doctors is actually a full-scenario platform, including detailed doctors' personal encyclopedia business cards, suggestions on what videos to do in Baijiahao, online consultation, live broadcast, etc. Baidu is a one-stop all-round online practice venue. Therefore, it is very important for doctors to form a link between these things in one place. Therefore, operational capabilities and full-link services are very important points for us to attract more cooperative institutions and doctor groups. The most important thing is to grasp C and B, which is a very important sword for Baidu Health to do business today.1.4 What are the manifestations of the service-oriented Baidu mobile ecosystem that we can perceive?Li Tianhua(Former mobile Baidu architect)：Among Baidu's more than 30 million daily active services, there is one network. Let's talk about four networks: information network, e-commerce network, industrial network and social network. Baidu is also deepening the industry network. Their own love purchasing has already gained income. There is also a lot of room for independent listing when it is dismantled, and its annual income is more than double digits. This unit is 100 million. Therefore, I think if Baidu is really willing to give talents more room for growth and release more equity for them to take the lead independently, I think Baidu can do a lot.Love purchasing, the B to B ecosystem, before<a href=\"https://laohu8.com/S/BABA\">Alibaba</a>Billions of revenue, at least 3.5 billion of revenue is brought by Baidu's traffic acquisition. In the past history, Baidu only paid 5.6 billion advertising fees. Later, Baidu felt that I might as well do this service myself. In fact, this logic is right. B2B Alibaba also provides an intermediate service logic. This logic is very suitable for Baidu, so Baidu did it itself. Now Baidu has done a good job itself, which has also brought double-digit growth and over 100 million revenue.1.5 Is it too late for Baidu to make efforts in e-commerce now? How does Baidu plan to make efforts and how to differentiate?Ping Xiaoli(Vice President of Baidu, General Manager of Baidu APP)：First of all, let me answer why I cut into e-commerce. I think there are a few changes in business trends that we have observed here.First, the infrastructure of the Internet is becoming more and more mature. In terms of payment process, it is no longer exclusive to e-commerce. Second, the supply side of e-commerce platforms is constantly enriching, which should be said to be extremely enriched. Now the supply chain side is actually completely more open and open, and it is also very open to all platforms. Third, in terms of users' purchasing habits, today's purchase, in fact, many consumption decisions are stimulating and rational, and the scene is stimulated. Users may have shopping needs anytime and anywhere. This kind of shopping demand, with the improvement of Internet infrastructure and the great enrichment of supply side, makes it possible to buy at any time on all occasions.Secondly, we think it is not too late for Baidu to cut into e-commerce, because we have these three basic conditions, not only Baidu, but I think many scenarios may become a possible purchase channel. That is to say, in fact, the phenomenon that China's Internet e-commerce companies are greatly concentrated at the top will gradually change to some extent.Furthermore, whether other platforms have advantages or not, when we do such things, we not only see whether the industry has advantages, but also look at it in combination with our own products. In fact, a large number of users search for product category words and brand words every day on \"search\", such as which washing machine brand is good to search for. There are so many such words. Why do we see so much demand? It hasn't died out for so many years, because we see what the user wants to check in the whole process. In fact, he needs to know better about this brand, this product and how it is, including some of its main parameters, official evaluation and user experience. That is, he will go to Baidu to search before the whole consumption decision-making pre-link. Let me give you a very simple example. You couldn't buy medicines online before, but now you can. It turns out that everyone can buy medicine. We see that many users actually want to search on Baidu, such as searching for a certain product for a long time. Why can't they place an order directly? This search is more like three C products and furniture products, and some decision-making users with higher unit prices will be more willing to search, so we think we have certain characteristics here.Therefore, for example, when we search for e-commerce, we focus on building two scenarios, one is the search scenario and the search for e-commerce. In fact, this is also based, which is the basis just mentioned. The second is that from the perspective of the key direction, the key will be more, which may not be the same as other platforms, that is, we are all positively doing some homogeneous competition, and we are doing more of our venues. Our purpose is not to compare with others. Today's e-commerce is a service-oriented strategy. Baidu has no battlefield. It is not this logic to be e-commerce. The whole logic means that from \"content to service\", the service-oriented strategy is a very important part, and buying goods is a part of the service. The purpose of our e-commerce is more about how to better meet users' needs in this place, dig out users' needs, and do a good job in users' real shopping needs. If we do it in this kind of scenario, we are more inclined to the user's relatively rational role, and let him choose better products, brands, and explosions. These things are our main tasks. When we do it, we will aggregate more product content, go to service purchase channels, and give more information to users, so that they can make a comprehensive decision regardless of seeing the price or protecting their rights and interests.Another scene is \"live broadcast\". Live broadcast e-commerce is also a product form that we are seeing growing very fast now, because Baidu actually has search, recommendation and information flow, and the information flow has also changed greatly from graphics and texts to live broadcast. change. It is relatively typical, and it is more suitable for the precipitation of popular products in various activities. If live broadcast and e-commerce are precipitated together, it will be very smooth. For example, some merchants put their brand stores directly in Baidu, search for one-click direct access, put their most complete stores here, and then they can do some activities with some goods, cooperate with Baidu stores, and broadcast two live broadcasts. Let users have some node-like cognition, such as the live broadcast with wine some time ago. In fact, after we finished the live broadcast that day, the gmv was very high that day. After that, many people searched there and then went back to that store to buy. Therefore, it is a scene that can be completely taken together, so this is an opportunity we see, and it is also some of our ideas that we have been exploring and verifying in the past period of time.1.6 Baidu's X + Y strategic layout, how many types of Y does it have? What are you going to do? What might be the next Y?<img src=\"https://static.tigerbbs.com/e1162a7bad64ba3e26a0d37fbb04772e\" tg-width=\"1080\" tg-height=\"462\" referrerpolicy=\"no-referrer\">Ping Xiaoli(Vice President of Baidu, General Manager of Baidu APP)：There are probably three types of Y, but good-looking is different from healthy. Good-looking is X, and healthy is Y. However, these three categories are basically summarized into two categories, which are internally called Y1, Y2 and Y3. Y1 and Y2 are strongly related to X. Y1 and Y2 take advantage of the advantages of X platform to give full play to their industrial advantages, and the space market is relatively large. Here, choose the platform with key depth, and the first choice is the strategy of extending the content of X platform to service. Second, on the Y track, you can have a certain market share and some of your own positions, which is Y1+Y2. The internal difference between Y1 and Y2 is the difference in size, the health is Y2, and the small Y1 does not need special explanation here. Y3 doesn't have much strong connection with X. For example, network disks and games are placed in Y3, which are all separate industries.The vertical category of Y is how to do better and more prosperous, including how to do poorly. In fact, we have a set of shutdown and transfer mechanism in the whole incubation, that is, to see the progress. Of course, if we can't do it, we will definitely shut down some decisively. We do more than a dozen or twenty vertical categories internally, and some teams only have a few people to explore first. The general mechanism is like this.How to be better and more prosperous, I think it is very important that every vertical category should find someone who understands this industry. In Baidu, we are on the platform ourselves. I understand the ability of the platform very well, but each vertical category has different styles of play. However, you can do it yourself. If you don't rely on Baidu's advantages, you may not be able to do it according to the entrepreneurial method. Because you will compete with others in a homogeneous way, you need a combination of the two. Just like when I had a discussion with my colleagues earlier, I told him that in fact, you don't know how big a gold mine search is. Outside, everyone thinks that the traffic should be quite good and the value is quite high. This is still a very open cognition.Search can gain insight into consumer demand. For example, it is also a trend for fashionable products now, and many of these things can be fed back in users' minds in a timely manner. How to undertake this kind of user's needs better? If you only search for products, it turns out that it literally matches the corresponding answers, but the industry is not. You have to see what these questions mean to the user, what he needs, and then combine the products of the industry you designed, how to undertake and connect with him. Here, two teams of X and Y need to collide with the design. So we have some cooperation here, and we have some teams here that specialize in combining with them. Why is health doing well? In fact, health is that the whole search team was hatched, and later some people who understand these industries were joined.Therefore, this is also the experience of past practice. The capabilities and genes of the X platform actually helped this industry a lot in the early stage. However, we also need the input of professionals in the industry, so we will adhere to this experience and principle in human organization construction and product design. What really reflects is the differentiated competition path, not what others have done. I have to do the same as others. This is not the only idea for Baidu to do this, so we chose a lot of Y, which is strongly related to X.In addition to health, the next step may be to start the in-depth layout of Y in multiple fields of \"education, medical beauty, and automobiles\". The logic of the selected vertical category itself must first conform to the strategic logic of the big Baidu APP, which is \"content + service\" The connection point is relatively clear. First of all, the whole decision-making and logic users need a lot of content as requirements, and second, the whole back link has services in it. The third is that the needs of users in its industry itself are equivalent to that part of a certain link happens on Baidu APP, or we have certain advantages. Therefore, when choosing vertical categories, it will be selected according to Baidu's genes. Under the logic of choice, it is determined that the main task of these vertical categories is to strengthen Baidu's overall capabilities in this respect. At the same time, when comparing in other markets at the same time, I also have to look at other categories in other markets. In this case, it provides a strong spur and the requirement of survival consciousness.2. Intelligent cloud2.1 Which businesses of Baidu are likely to grow independently in the future?Li Tianhua (former mobile Baidu architect):At present, Baidu accounts for several relatively large categories in the Top80% traffic, such as education, health, games and reading, which are all relatively large categories that Baidu can make in-depth efforts. I think a skinny camel is bigger than a horse. Don't listen to the outside who always spit out that Baidu is bad and bad. There are more than 200 million daily living apps, and there are several daily living apps in Baidu with more than 30 million daily living apps, including maps, network disks, etc., which are all very good. Of course, the most basic search is how to fight the defensive battle well, and at the same time make use of the new opportunities in the market in the next 3 to 5 years, seize them, make a second spring, and AI can better serve various scenarios, drive others, whether it is automobile Industry or service industry, or smart industry, education, make greater upgrades and build some infrastructure facilities. There is also a small platform with large infrastructure facilities within 10 years, all of which are thoughtful. To realize the trend of the talent market, a large number of talents enrolled in schools every year enter Baidu's system for incubation in large quantities.As we all know, Baidu has been searching for so many years, and the mobile phone Baidu has about 200 million daily activities. Baidu's most valuable businesses, such as Baidu Netdisk, have about 31 million to 33 million daily users and about 3.3 billion annual revenue. As you can see, when Baidu went public, Ali issued a clarion call again, saying that they would also be Ali's network disk. In fact, everyone has seen Baidu's persistence in the network disk for so many years, and there is an episode in the middle, when 360 gave up the 360 network disk. This matter is a long-term value investment matter. This business is still the number one in the industry at present, and it will definitely be split and listed independently.There is huge room for several core businesses of Baidu in the future. For example, Xiaodu speakers, which are growing well now, are in charge of Jing Kun. This business is a better product in the future era of instinctive interactive experience, and there is a lot of room. The essence of the speaker business is a product in the RTC era. RTC is the logic of Real time clock, a real-time connection service. One of the biggest opportunities from today's service to the future is the RTC era. A valuable business in the RTC era. In fact, Tesla's autonomous driving is essentially an RTC service. There is a<a href=\"https://laohu8.com/S/300024\">Robot</a>Help you calculate in real time, help you see the road conditions, and help you control the car. In fact, this is the same. Duer OS business is also a product in the RTC era. It has the same interactive experience as Clubhouse. I get up every morning and ask Xiaodu Xiaodu how the weather is. Xiaodu Xiaodu, open the curtains, this in itself is the future of interactive experience. In this case, Xiaodu's independence is conducive to its growth, and it will also help it drive Baidu's stock price to continue to rise.In addition to Baidu Netdisk and Xiaodu, as we all know, Baidu's health service revenue has been good for so many years. Personally, I think Baidu Health also has the opportunity to split up, which has a huge value-added space for Baidu's stock price. My suggestion is that you can pay attention to these core businesses.Here, I would like to mention Baidu Map in particular. I think Baidu Map has done very well in the last year, and it has been upgraded than before. I don't know if you have used Baidu map to take a taxi. Everyone found that recently, Didi took a taxi, and Didi changed a new version. Everyone complained and couldn't find out how to use Didi. But you will find that taking a taxi with Baidu Map is cool and cheaper than Didi. You can take a taxi with Baidu Map, which is really easier to use than Didi. My experience these days is the same.2.2 Where is the technical threshold for cloud disk products?Pan Luan:In fact, network disk is a very high-tech thing. I can tell you one first<a href=\"https://laohu8.com/S/XNET\">Thunder</a>Examples of. In fact, the founding team of Thunder also went out of Baidu, and co-founder Cheng Hao and CTO Li Jinbo both went out of Baidu. Where is the technical cow of Thunder? Some of its capabilities are very similar to those of the network disk. Thunder needs to be crawler, transmission and storage. In fact, each of them is a very hard technology. If you have an impression, those who made network disks in the early years were not as powerful as Baidu.In fact, Baidu was the first to do distributed computing in China. Baidu's internal operation and management servers should be very strong, because it has to carry the computing power of peaks and valleys. Moreover, 10 years ago, Baidu's servers should have the largest number in China, involving machine management, resource scheduling and the whole network management. In fact, these three core technologies are relatively advanced. At first, Alibaba Cloud's architecture system was exactly the same as Baidu's service system.Li Tianhua (former mobile Baidu architect):In the middle, we should first talk about the value of cloud disk products and services, that is, the server bandwidth cost that everyone jointly dilutes. The bandwidth cost per megabyte of CDN of normal Internet companies is more than that of 7 yuan, and the cheaper ones are more than that of 6 yuan. Alibaba Cloud can generally buy 20% off in China, which is considered cheap, but few people can buy 65% to 30% off. In fact, the bandwidth cost per megabyte is six or seven yuan, which is normal, and it is also true for large companies like Bilibili. same. There is also a problem of storage cost for individuals. If there is no problem of storage cost, there is no need for sharing logic. In fact, sharing is still everyone's joint efforts to dilute the cost of bandwidth servers. There is such a logic in essence.2.3 How to define intelligent cloud? Where is the room for improvement in the future development of intelligent cloud?Li Tianhua(Former mobile Baidu architect)：In fact, the recent growth of Baidu Smart Cloud is very high, and part of the reason is that Baidu used to buy Ali for many internal businesses. Later, the boss wanted everyone to buy their own services, so he also brought a lot of income increment.Strictly speaking, Baidu Cloud still has a lot of room for improvement in its business. I look at this from both sides. Should we give some space or give each business independent development, otherwise it will also be a problem to be run by Baidu Cloud. Of course, if all the income is given to Baidu Cloud, Baidu Cloud can have a better valuation, a better cake when listed, and it can be better empowered if it is taken back from the secondary market, and the boss can also earn More, this is normal.Shi Zhong(Light Black Technology)：The entry point of Baidu Smart Cloud was chosen when there was not much space left for him. And this is also based on its own advantages, that is, AI.Many times, I don't know if you still remember the point he cut in. When Baidu first launched ABC, it actually gave an example. Using AI to help Shougang do the quality inspection of steel plates is actually a very heavy thing. It won't be like the website. I sell you resources. When you go online, I can make money like rent collection, which requires strong service and strong adaptability. The system of quality inspection in steel plants itself is very complicated, and it needs a lot of adaptation and adjustment. Including some wind power cooperation later, predicting which units will have problems, including intelligent customer service.Frankly speaking, the industry itself has many categories, so what it does is scattered. But the advantage is that once there is a benchmark in the industry, this benchmarking effect is actually very strong. It's not that easy to make money, but I think this track is the broadest, so I can only say this. But to do it well, it actually takes a lot of effort. Can this matter be done as imagined? It remains to be seen. In short, there is imagination.Pan Luan:Someone asked, many people are doing AI, and there has never been an exit visible to the naked eye. They grit their teeth, find someone, hire a team of smart speakers, and find an exit at once. I think Baidu has always been practicing dragon slaying in technology, and has always wanted to find a suitable scene. Making smart speakers, like Hangzhou Ruoqi, there is another company that makes smart speakers. I think their latest industrial-grade solutions are now using Baidu's cloud services to make their solutions. These teams that used to be speakers began to borrow Baidu's cloud services.Brother Qiushi(Smart Speaker PM)：There will be this process. These general cloud services, or general NLP services, include<a href=\"https://laohu8.com/S/002230\">IFlytek</a>, Alibaba Cloud provides it, it depends on who is more diligent in sales. In fact, Baidu has invested a lot in this area, because of its external expansion in this area, and some ISVs have profit-sharing, coupled with the policy of interest incentives. Including Alibaba Cloud, I feel the same. In essence, I feel that it is still at the stage of pre-sales selling concepts to spread the number of users. There are various demands, some of which are to improve efficiency, some of which are that you can't do it if others do it, and some places can get more government subsidies if you do such things. Therefore, the coverage ability of local push still has a relatively large impact. This kind of sales-oriented work actually accounts for a relatively large proportion in the whole, which is not particularly the same as the logic of the whole Internet. My own feeling is that the vertical field is different from the consumer Internet. The cognition of the interests of all parties in the industrial Internet and the upgraded modules are very important. This kind of knowledge reserve is not necessary for a large company. If you are willing to do it, a lot of it is dirty work, and the investment in this reserve, accumulation and collation is very large in every field. The friends of Light Black Technology have a good chat, and they are still very particular about choosing fields.3. Autonomous driving & car manufacturing3.1 How fast or slow does Baidu build cars?Li Tianhua(Former mobile Baidu architect)：Now the chairman of Baidu's car is my most trusted big brother in Cover. I think this matter should have a very good result, so everyone still has some small expectations for Baidu's car. Especially when it comes to the stock price pull, there should be a small expectation. There are generally two types of companies. In the stock market, one is to make games, and the other is to be real. If you finally make the game, then you are a very good company, and this is the logic. So don't worry too much about the short term. Many people complain that this CEO is unreliable, and that one is unreliable either. It doesn't matter, the market has enough time. The Internet generally incubates a new business. Whether it is incubated from a large platform within the system or from outside, this cycle takes about two years. Within two years, as long as he takes the key steps right, this problem is not big.3.2 What are the different capabilities of Baidu's autonomous driving and Tesla and Didi?Shi Zhong(Light Black Technology)：In fact, there are two evolution routes of autonomous driving, all of which are similar represented by Internet manufacturers. They take the route of pure autonomous driving, starting with L4, and starting with no one. You can limit the scene first, and unmanned driving with limited scenes is their start. But like Tesla, its starting point from the beginning is L2 or L3, which is assisted driving. These are two different routes. But in the end, they all want to get to L5. L5 means unmanned driving without limited scenes, that is, any scene and any road condition can be reached. But now no one knows when this matter will be achieved. As far as we can see now, what we can see is L4. L4 is everyone's goal, but L4 is also marching towards it. There are two routes, one is the Tesla route and the other is the Baidu route, which can be understood so simply.The difference is actually relatively simple. Just like cultivation, there are different methods, including the Sword School and the Qi School. It is the butt that determines the head. If Tesla has a car, it must first sell the car. It will not take driving itself as its own responsibility, but selling the car as its own responsibility. Because no one can do L5 level autonomous driving now, but they can sell cars first. Use relatively low-level assisted driving to hone my autonomous driving technology, and one day maybe I can become a pure autonomous driver. But Baidu didn't have a car when it started. When it started without a car, the final route was not to sell cars, but to sell services. That is to say, its ultimate situation may be like Didi, which is to operate a service system. He doesn't have to own a car. He only needs to buy services, and I cover all the underlying technologies. Therefore, there is no need to have a stage where users drive themselves, in the form of services such as Robotaxi.Pan Luan:Baidu's car making now should be the most concerned thing about Baidu in this round, that is, a new Baidu. The mobile ecosystem we are seeing at present, and what we are seeing further ahead are things related to the cloud. The story that is close to being realized should be the car part. The car part is also divided into several lines, such as providing services to these car manufacturers and helping the government build the Internet of Things system. Just like Baidu's smart clouds now, many of them should be tied to the government's large transportation system. Of course, there is another piece of self-driving taxi service in the future. But when we discussed Baidu Maps earlier, consumers actually didn't care whether your technology was good or not. What they cared about was how to send me there. At that time, they had to consider whether the vehicle density was enough or not.You can think of it as a service provider for car manufacturers. Does everyone think this is my core interest? It was rumored before that Baidu wanted to vote<a href=\"https://laohu8.com/S/LI\">Li Auto</a>, must use<a href=\"https://laohu8.com/S/APOL\">Apollo</a>If you don't have the ideal system, you choose to build it yourself. I think at present, powerful car manufacturers may choose their own systems that can have stronger control, and invest in or acquire autonomous driving<a href=\"https://laohu8.com/S/V03.SI\">Startups</a>, but they will not hand over the entire autonomous driving solution to Baidu. They may purchase some specific services of Baidu, such as high-precision maps.3.3 Is there really an opportunity for the big Android system that Baidu once envisioned in the automotive industry? Can Apollo become the Android of the automotive industry?<img src=\"https://static.tigerbbs.com/d4f13dcf2d14ce5aec837d5bddc0197d\" tg-width=\"1080\" tg-height=\"483\" referrerpolicy=\"no-referrer\">Classmate Chen(Baidu Autopilot Business)：In the intelligent transformation of the automotive industry, Baidu Apollo hopes to use more deterministic technologies and solutions to accelerate the boarding of intelligent technologies for OEM customers in a limited time window, so as to prevent everyone from repeatedly building wheels.First of all, from the perspective of enterprise R&D investment costs, if we consider developing a multi-model platform that supports intelligent driving and includes E/E architecture, the investment cost is around 10-20 billion; The investment in technology research and development of autonomous driving software is also more than tens of billions, and a cumulative sales of 100-200W units are required to achieve cost amortization.Secondly, in addition to the high cost investment, the time cost of autonomous driving software development is also huge, and it needs to go through the process of capacity building, forming effective R&D capabilities, and then productizing output. It usually takes at least 6 years for companies in the industry to independently develop autonomous driving software. Most importantly, research and development costs do not guarantee the availability of available technical results, including test licenses, maps, V2X road test resources, intelligent clouds, etc., which will affect the level and landing time of autonomous driving technology.Apollo hopes to use its eight years of technology accumulation to help car companies build competitive smart cars and promote the development of the entire industry in the form of open cooperation.\n</article>\n</div>\n</body>\n</html>\n","type":0,"thumbnail":"https://static.tigerbbs.com/0cb3c6c3c55603d00a14ee3389850b7f","relate_stocks":{"09888":"百度集团-SW","QNETCN":"纳斯达克中美互联网老虎指数","BIDU":"百度"},"is_english":false,"share_image_url":"https://static.laohu8.com/e9f99090a1c2ed51c021029395664489","article_id":"1184519409","content_text":"百度回港上市现场，遇到跟踪百度多年的极客公园的创始人张鹏，问他怎么看。张鹏说：“这是百度一直通过技术创造价值的第二次体现，可以理解为百度在互联网搜索技术之外，AI第二技术曲线完成了一次上市。”\n当晚，潘乱与几位朋友一起做了场直播，聊聊百度二次上市和AI故事，关于移动生态、云业务和智能驾驶，聊业务发展的节奏和路径。\n问题包括：\n短视频＋直播化表达和消费模式下，百度搜索如何应对？\n百度在云计算上的机会在哪里？\n百度智能驾驶板块未来最主要的变现业务是哪一块？\n百度的自动驾驶和特斯拉、滴滴的不同能力项是什么？\n……\n\n4月底，乱翻书又参加百度移动生态的万象大会。百度最新的移动生态战略在以X+Y战略布局的形式出现，即“横向开拓用户规模，纵向深耕行业垂类”。会上针对创始者和合作伙伴发布了九大扶持计划，好看视频要帮助入驻创作者变现，百度健康要扶持医生、MCN机构、公立医院和商家，百度电商也要扶持品牌、商家和主播。\n作为“X”平台，百度APP有三个战略，内容视频化、搜索人格化、信息服务化，后面“两化”也都是直接指向生态供给的。\n\n在后来的采访环节，乱翻书向几位业务负责人发问：\n现在好看视频里内容质量参差不齐，怎么去提升内容的真人感？\n二创剪辑很多都是假的人声，怎么增强内容的真实性？\n百度的X+Y战略布局，有几种Y？打算怎么做？\n下一个Y可能是什么？\n……\n本文是两次活动的速记整理，意在回答一个问题：\n一个新的百度的故事是什么？\n一、移动生态\n1.1 好看的帧视频是类似B站弹幕那样的基础信息结构？所有人都可标记还是特定创作者？整体是什么样的思考？\n宋健（百度短视频生态平台总经理）：帧视频这个结构是希望能够做成知识的标准，从产品创新变成真正泛知识的一个视频。短视频这几年都走娱乐化，它们的产品都是更适合娱乐，帧视频相对弹幕是围绕更加客观更加理性，是知识补充信息含量，不需要那么多情绪。我们要面向更大的用户群体，所以我们在思考怎么围绕我们的视频来界定其他的视频，这是我们的出发点。\n我们肯定是优先创作者来去做，但UGC我们也在尝试。现在我们的视频信息结构，比如说医疗健康的视频，说这个是糖尿病，如果帧正好在糖尿病的症状，只看一个视频他对这个东西了解的不是特别清晰，如果对这个点非常好的补充到视频的信息上，就会非常好。这里面创造者能不能做这个事情，UGC能不能做这个要靠一套的战略，百度也会提供很多工具，帮助做这些信息。\n信息演化的规律就是从低信息到高信息，无论是文字识别还是从长视频到短视频，信息越来越高。但是，这个“高”现在在网络中太高会越来越通过情绪来抓眼球。信息密度一定往高处走，是不是一定往极致的情绪这种刺激上走，这个是我们在思考的一个问题。我们想通过帧视频补充信息密，掰一掰，把信息密度增加，但不全是情绪，补充一些知识的。一方面迎合趋势，一方面往视频更加规范的程度上走。\n1.2 现在好看视频里有的内容质量参差不齐，比如二创剪辑有一些是AI合成人声，怎么看这个事情，怎么去提升内容的真人感？\n宋健（百度短视频生态平台总经理）：我简单介绍两个点，怎么去改善这个问题。\n一个是说你要有比较好的社区氛围，很多产品从一开始它在选种子作者的时候是比较科学的，就是慢慢的很容易会形成一个社区氛围以后，但是这个没法再往回倒腾。\n第二个就是说要给作者或者是UGC做更多的攻略，所以在做度咔也好，在做帧视频也好，要有新的工具出现才能有新的东西，这样才能达到差异化的东西。第二个是我们主动做的，要不然在体量没有优势的情况下，比如说把生态进行很大的变化是不可以的，所以从工具角度考虑（进行产品创新）。\n1.3  想要吸引更多的用户或者专家团队来选择百度平台，百度健康独特的优势是什么？\n\n杨明璐（百度健康总经理）：第一群众基础。就是遇到这种情况第一时间想到的是谁？百度这个心智是有的，主要表现在用户体量，我们有很重要的群众基础。\n第二是技术实力。百度技术竞争很强，基于大数据做精准人群画像，进行AI赋能，相信不久，大家就能看到非常明显的一些变化，特别是在精准医患匹配上面，近期也积累了不少专利，这个就是强大的技术能力。\n第三就是基于用户基础的强大医生运营能力。医生为什么来平台？很大原因是有用户，但是他会不会黏得住，因为他要运营非常多的场景。百度健康给医生提供的其实是一个全场景的平台，包括细致到医生个人百科名片，开设百家号里面做什么视频的建议，在线咨询问诊，直播等，百度是一站式的全方位的在线执业场地。所以对医生来说，把这些东西在一个地方形成链路是很重要的，所以运营的能力和全链路服务是我们吸引更多合作机构和医生群体非常重要的一点，最重要的还是抓C和B，这个是我们今天百度健康做业务非常重要的利剑。\n1.4 百度移动生态的服务化有哪些可被我们感知的体现？\n李天华（前手机百度架构师）：百度在3千万以上的日活服务里面，其中还有一张网，我们讲四张网：信息网、电商网、产业网、社交网。产业网里面百度也在深入，他们自己做的爱采购，收入也已经起来了，它拆出来独立上市空间也很大，一年也是两位数以上的收入，这个单位是亿。所以我觉得百度如果真的愿意给人才更多的成长空间，愿意给他们释放更多的股权让他们独立挑头，我觉得百度有很多可以做的事情。\n爱采购，B to B这个生态，以前阿里巴巴大几十亿的收入，至少有35亿的收入是在百度获取流量带来的，之前历史上只给百度交过五六亿的广告费。后来百度觉得我还不如自己做好这个服务，其实这个逻辑是对的。B2B阿里巴巴提供的也是一个中间的服务逻辑，这个逻辑很适合百度，百度就自己做了，现在百度自己做得很好，这方面也带来了两位数以上的增长，过亿的收入。\n1.5 百度现在发力电商是不是为时已晚，百度打算怎么发力以及如何差异化？\n平晓黎（百度副总裁，百度APP总经理）：首先先回答一下为什么切入电商，我觉得这里有几个我们观察到的商业趋势变化。\n第一，互联网的基础设施越来越成熟了，从支付这套流程来讲，已经不是电商的专属了。第二，电商平台商品供给侧在不断丰富，应该说是极大的丰富，现在供应链端其实是完全更open和开放的状态，对所有的平台也都非常open。第三，用户的购买习惯上来讲，今天的购买，其实很多的消费决策有激发的也有比较理性的，场景的激发也好，用户随时随地都可能有购物的需求。而这种购物的需求，随着互联网基础设施的完善和供给端极大的丰富，让在所有场合随时有需求随时购买成为一个可能。\n其次，我们认为百度切入电商为时不晚，因为我们具备这三个基础条件，不光是百度，我觉得很多场景可能都会成为一个可能的购买通道。也就是说实际上原来中国互联网电商极大聚集在最头部的现象，慢慢会有一定的变化。\n再者，其他的平台来比有没有优势，我们做这样的事情不光看到行业有没有优势，还在结合自己的产品在看。“搜索”上其实每天有大量的用户搜索商品品类词、品牌词，比如说搜哪个洗衣机品牌好，这种词非常多，这种需求我们看到为什么会这么多？始终这么多年都没有消亡掉，因为我们看到用户在整个过程中他要查什么，他其实要更好的了解这个品牌也好，这个商品也好，它怎么样，包括它主要的一些参数，官方的评价，用户使用的体验。就是他在整个消费决策前链路之前会去百度搜查。我举一个很简单的例子，原来在线上不能购药，现在可以了。原来大家可以买药这个事，我们看到很多用户实际上在百度要搜，比如搜某个产品搜半天，他为什么不能直接下单。这个搜索更多的像三C类产品、家具类产品，相对单价更高的一些决策用户会更愿意搜索，所以我们觉得在这里我们是有一定特色的。\n所以，比如说我们在搜索电商的时候，着重构建两个场景，一个是搜索场景，搜索电商，其实这个也是有依据的，就是刚才讲到这个依据。第二个就是重点方向上来看，重点会更多的可能不会说跟其他平台一样，就是我们都是正面的做同质化的一些竞争，我们更多做好我们的场地，我们的目的不是和别人比做多大的场景。今天的电商是做服务化战略，百度没有战场，要做电商不是这个逻辑，整个逻辑是说从“内容到服务”，服务化战略里面很重要的一部分，购买商品是服务的一部分。我们做电商的目的，更多的是在这个地方如何把用户的需求满足的更好，把用户需求挖掘出来，把用户真正购物的需求做好。我们在这种场景做的话，更偏向于用户相对理性的角色，让他挑选到更好的产品和品牌、爆款，这些东西都是我们主要的工作。像我们做的时候，我们会聚合更多家的商品内容，到服务购买渠道，把更多的信息给到用户，让他不管看到价格还是保障权益，让他可以综合做一个决策。\n还有一个场景就是“直播”，直播电商也是我们现在看到增长非常快的一个产品形态，因为百度其实除了有搜索有推荐有信息流，信息流从图文现在到直播也起了非常大的变化。是比较典型的，比较适合各种活动爆款沉淀，如果把直播和电商沉淀在一起是非常顺的。比如说一些商家把他品牌店直接入驻在百度，搜索一键直达，可以把自己的最全的店铺放在这儿，然后可以再拿一些商品做一些活动，和百度的店合作，再播两场直播。让用户产生一些节点式的认知，比如说前段时间做的直播带酒，其实我们当天直播完之后，当天gmv非常高，之后还有很多人在那儿搜索，再回到那个店铺购买。所以，它是一个完全可以承接到一起的场景，所以这是我们看到的机会，也是我们在过去一段时间不断摸索和验证了的一些我们的设想。\n1.6 百度的X+Y战略布局，有几种Y？打算怎么做？下一个Y可能是什么？\n\n平晓黎（百度副总裁，百度APP总经理）：大概有三类Y，不过好看和健康不一样，好看是X，健康这个垂类是Y。但是这三类基本总结成两大类，内部叫Y1、Y2、Y3，Y1和Y2和X有强关联的，Y1和Y2借助X平台的优势发挥产业优势，空间市场比较大，在这里选择重点纵深的平台，第一选择X平台内容到服务化延伸的战略。第二，在Y的赛道上，能够拥有一定的市场份额和自己的一些阵地，这是Y1+Y2。Y1、Y2的内部的区别是大小的区别，健康是Y2，小的Y1这里不用特别解释。Y3没有和X有特别多强联系，像网盘和游戏放在Y3，这都是单独的产业。\nY的垂类，就是怎么能做的更好更繁荣，包括做的不好怎么样。其实我们在整个孵化上有一套关停并转的机制，就是来看进展，当然实在做的不行，我们肯定果断的会关掉一些。我们内部做的垂类十几二十个很多，有的团队就几个人先摸索，大概的机制是这样子。\n怎么样更好更繁荣，我觉得很重要的一点就是每个垂类要找懂这个行业的人，在百度，我们自己在平台，我对平台的能力是非常理解的，但是每个垂类打法都有不同。但是垂类自己干就能干成，不依靠百度的优势的话，按照创业的方法也不一定能干成，因为会和别人同质化竞争，所以需要两个结合。就像前段和同事做一个讨论的时候，我给他讲说其实你们不了解搜索有多大的金矿，搜索在外面大家觉得流量应该挺不错的价值挺高的，这个还是很开放的一个认知。\n搜索可以洞察消费需求，比如说现在潮品起来也是一个趋势，很多这些东西在用户心中可以很及时的反馈出来。把这种用户的需求之后怎么承接的更好，如果只做搜索产品，原来是字面匹配相应的答案，但是行业不是，你要看到这些问题对用户意味着什么，他需要什么，再结合你所设计的行业的产品，怎么跟他承接，跟他对接，这里是需要X和Y的两个团队碰撞设计的。所以我们在这里是有一些合作的，我们这里有一些团队和他们专门做结合。健康为什么做的不错？实际上健康就是整个搜索团队孵化出来了，后来加入了这些行业里懂的一些人。\n所以，这也是过去实践出来的经验，X平台的能力、基因，其实在前期对这个行业有很大的帮助。但是，我们又需要行业的专业人输入进来，所以在人的组织搭建上，在产品设计上都会秉承这样的经验和原则。真正体现的是差异化的竞争路径，而不是说别人做了什么，我非得和人家做的一样，这也不是百度做这件事情唯一的思路，所以我们选择了很多Y都是和X有很强的关联的。\n除了健康，接下来可能在“教育、医美、汽车”多个领域上都开始Y的纵深布局，所选的垂类本身的逻辑就得首先符合大的百度APP的战略逻辑，它是“内容+服务”衔接点比较清晰的。首先整个决策和逻辑用户需要大量的内容是需求，第二个整个的后链路是有服务在里面。第三是它行业本身用户的需求相当于一部分某个环节是在百度APP上发生的，或者我们有一定优势。所以在选择垂类的时候，会根据百度的基因去选择。在选择的逻辑之下，就决定了这些垂类肩负的主要任务就是要加强百度整体在这方面的能力。同时放在其他市场同时比的时候，我还要看其他市场其他的品类，在这种情况之下，提供了既有很强的鞭策和生存意识的要求。\n二、智能云\n2.1 百度哪些业务未来有可能独立成长？\n李天华（前手机百度架构师）：现在百度占的Top80%流量里面比较大的几个品类，教育、健康、游戏、阅读，对百度来说都是可以深入发力的几个比较大的品类。我觉得瘦死骆驼比马大，大家不要听外面老是吐槽百度这个不好那个不好，2亿多日活的APP，手百一直在，百度3千万以上的日活以上的APP有好几个，有地图、网盘等等，都是非常不错的。当然最基本的搜索，如何把防守战打好，同时再利用市场未来3到5年新的机遇，把握住，做出来第二春，以及AI更好的服务各个场景，带动其他，不管是汽车产业还是服务产业，还是智慧产业、教育，做更大的升级，做一些基础化设施。还有小度做10年以内基础化设施大的平台，都是考虑得比较周到的。人才实现市场的趋势，每年大量校招生的人才，大批量的进入百度的体系内孵化。\n大家知道，百度做了这么多年的搜索，手机百度大概两个多亿的日活。百度现在最有价值的几个业务，比如百度网盘，大概有3100万到3300万的日活，一年大概33亿的收入。大家也看到了，百度上市的时候阿里又发出号角，说他们也要做阿里的网盘。其实大家看到了百度在网盘里这么多年的坚持，中间还有一个小插曲，当年360放弃了360网盘。这个事情是一个长周期价值投资的事情，这个业务在行业内目前还是属于第一，它绝对会拆分出来独立上市的。\n在百度未来几个核心业务都有巨大的空间，比如现在增长比较好的小度音箱，是由景鲲负责的。这个业务在未来的本能交互体验时代里面是一个比较好的产品，有很大的空间。音箱业务本质是一个RTC时代的产品，RTC是Real time clock的逻辑，实时连接的服务。当下这个服务到未来一个最大的机遇，其中之一就是RTC时代。RTC时代里面有价值的业务，其实特斯拉自动驾驶本质上也是一个RTC的服务，有一个机器人实时帮你去计算，帮你看路况，帮你操控车。其实这也是一样的，小度助手（Duer OS）业务也是一个RTC时代的产品，它跟Clubhouse的交互体验是一样的，每天早晨起来问小度小度，天气怎么样。小度小度，打开窗帘，这本身就是交互体验的未来。在这种情况下，小度独立出去有利于它的成长，也有利于它带动百度的股价继续往上涨。\n除了百度网盘、小度，大家也知道，百度这么多年来健康服务收入一直不错，我个人觉得百度健康也有机会拆分出去，这个对百度的股价都有巨大的增值空间。我的建议是大家可以关注一下这些核心业务。\n这里还要特别提一下百度地图，我认为百度地图最近这一年多做的特别好，比以前做的有升级。不知道大家有没有用过百度地图打车，大家发现最近用滴滴打车，滴滴改了一个新版本，大家都很抱怨，都找不到怎么用滴滴了。但是你会发现用百度地图打车比滴滴爽，还要便宜，大家可以用一下百度地图的打车，确实比滴滴好用，我这几天的体验也是这样的。\n2.2 云盘产品的技术门槛在哪里？\n潘乱：其实网盘是技术很高的事情，我可以先给大家讲一个迅雷的例子。其实迅雷的创始团队也是百度出去的，联合创始人程浩和CTO李金波都是从百度出去的。迅雷这个技术牛在哪里？它跟网盘的能力有些是非常像的，迅雷要做爬虫、传输、存储，其实每一项都是非常硬的技术。如果大家有印象的话，早年那些做网盘的都不如百度给力。\n事实上最早在中国做分布式计算的绝对是百度，百度在内部的运营和管理服务器应该都是很坚挺的，因为要承载波峰波谷的计算能力。而且10年前百度的服务器应该是中国最多的，涉及到机器的管理、资源的调度以及整个网络管理，其实这三项核心技术都是比较领先的。最初阿里云的架构体系跟百度的服务体系都是如出一辙。\n李天华（前手机百度架构师）：中间我们要先讲一下云盘产品服务的价值，就是大家共同摊薄的服务器带宽成本这个事情。正常的互联网公司大概每兆CDN的带宽成本在7元多，便宜一点的也得6元多。阿里云在国内一般能买到8折的就算便宜的，但是很少人能买到6.5折到7折的，其实每兆的带宽成本六七块钱很正常，像B站这样的大公司也是一样的。对于个人来说也有一个存储成本的问题，如果没有存储成本的问题，那就不需要共享的逻辑了。其实共享还是大家共同来摊薄带宽服务器的成本，本质上有这么一个逻辑。\n2.3 如何定义智能云？智能云未来发展的提升空间在哪里？\n李天华（前手机百度架构师）：其实百度智能云最近的增长很高，也有一部分原因，百度以前内部有很多业务都是买阿里的，后来老板想让大家买自己的服务，所以自己也带来了很多收入的增量。\n严格意义上，百度云在业务上还有很大的提升空间，这一点我是两边看的，该给一些空间还是要给到每个业务独立发展，否则都被百度云跑着也是一个问题。当然如果都给了百度云这个收入，百度云就可以有更好的估值，上市也可以有更好的蛋糕，从二级市场拿回来，也可以更好的赋能，老板也可以赚得更多，这个也正常。\n史中（浅黑科技）：百度智能云的切入点，是留给他的空间不太多的情况下选择的。而且这个也是基于它自身的优势，就是AI。\n很多时候他切入的点，不知道你们还记不记得，百度刚推出ABC的时候其实给了一个例子，用AI帮首钢做钢板的质检，这其实是很重的一件事，不会像网站那样我卖给你资源，你上线了我就可以收租一样的赚钱，需要很强的服务，而且需要很强的适配性，钢铁厂质检这种情况本身系统是非常复杂的，需要做很多适配，做很多调整，所以这个钱说白了是一个非常辛苦的钱。包括后来做一些风电的合作，预测哪些机组会有问题，包括智能客服。\n坦率来说，工业本身的门类就比较多，所以做的是分散的。但是好处在于一旦在行业里有了一个标杆，其实这种标杆效应是非常强的。并没有那么好赚钱，但是我觉得这个赛道是最广阔的，只能这样说。但是要把它做好，其实还需要付出很大的努力。这件事情到底是不是能做成想像的那样？还是有待观察的，总之想像力是有的。\n潘乱：有人问，很多人都做AI，一直没有肉眼可见的出口，咬个牙，找个人，拉了一个智能音箱的团队，一下子找到了出口就成了。我觉得百度在技术方面一直是修炼屠龙术，一直想找到合适的场景。做智能音箱，像杭州若其，还有一家做智能音箱的，我看他们最新在工业级的解决方案里面，他们现在也是在用百度的云服务做他们的解决方案。之前做音箱的这些团队他们开始借用百度的云服务。\n求是哥（智能音箱PM）：会有这个过程。通用的这些云服务，或者通用的NLP的服务，包括科大讯飞、阿里云都有提供，就看谁的销售比较勤快。其实百度在这方面投入比较大的地方就是，它在这块的外拓，还有一些ISV是有分润的，再加上利益激励的政策。包括阿里云我感觉也一样，本质我上觉得现在还是在一个售前卖概念铺用户量的阶段，有各种各样的诉求，有的是为了提效，有的是别人做了你不做不行，还有一些地方是做了这样的东西能拿更多的政府补贴。所以地推覆盖能力还是影响比较大的，这种偏销售导向的工作，其实在整个里面比重比较大，跟整个互联网的逻辑不是特别一样。我自己的感受，垂直领域跟消费互联网不太一样，产业互联网里面各方利益关系的认知，升级的模块，这一类的知识非常重要，这一类的知识储备对于一个大公司来说不一定愿意做，很多是脏活累活，而且这个储备、积累和整理的投入每个领域都是非常大的。浅黑科技的朋友聊得很好，在选择领域这些地方还是很讲究的。\n三、自动驾驶&造车\n3.1 百度造车的快与慢？\n李天华（前手机百度架构师）：现在百度的车的董事长是我最信赖的老大哥在Cover，这个事情我觉得应该会有一个非常不错的结果，所以大家对百度的汽车还是要有些小小的期待。尤其在对股价拉动这个事情上，要有一个小小的期待。公司一般分两种，在股市里面一个是做局，一个是实打实。如果你把做局最终做成了，那你就是一个很牛的公司，就是这个逻辑。所以大家不要太担心短期，很多人吐槽说这个CEO不靠谱，那个也不靠谱。没关系，市场有足够的时间，互联网一般孵化一个新的业务，无论是体制内从一个大平台里面孵化还是从外面孵化，大概这个周期需要两年。两年的时间内，只要他走对了关键的几步，这个问题是不大的。\n3.2 百度的自动驾驶和特斯拉、滴滴的不同能力项是什么？\n史中（浅黑科技）：其实有两条自动驾驶的演进路线，以互联网厂商为代表的都差不多，他们走的是纯自动驾驶的路线，起步就是L4，起步就是没有人，可以先限定场景，限定场景的无人驾驶是他们的起步。但是像特斯拉这种，从一开始它的出发点是L2或者L3，是辅助驾驶，这是两条不一样的路线。但是他们最后都是为了走到L5，L5的意思就是不说限定场景的无人驾驶，就是任何场景，任何路况都可以到。但是现在这件事情没有人知道什么时候能达到，现在目之所及，我们能看到的是L4，L4是所有人的目标，但是L4也正在朝它行进。两条路线，一个是特斯拉路线，一个是百度路线，可以这么简单理解。\n不同点其实比较简单，就像修炼一样，有不同的法门，有剑宗，有气宗。就是屁股决定脑袋，特斯拉有车，首先必须要把车卖出去，不会以自己驾驶为己任，而是以卖车为己任。因为现在谁都做不了L5级别的自动驾驶，但是可以先卖车。用比较低级的辅助驾驶来磨炼我的自动驾驶技术，有朝一日也许我可以成为纯自动驾驶。但是百度起步的时候没有车，这种没有车起步的时候，其实最终的路线不是卖车，是卖服务。也就是说它的终极情况可能就像滴滴一样，是运营一套服务系统，不用拥有车，他只要买服务，底层所有的技术我全包了。所以根本不需要有一个阶段是用户自己开车的阶段，是以Robotaxi这种服务的形式。\n潘乱：百度现在做车这个事情应该是这一轮里面大家对百度最关注的事情，就是一个新的百度，我们目前看到的移动生态，再往前面看到的是跟云相关的事情。接近可以变现的故事应该就是车这一块了，车这一块也分几条线，比如跟这些车厂做服务，帮政府做物联网的系统。就像现在百度的智能云里面应该很多都是跟政府大交通体系绑在一起的，当然还有一块就是未来做出租车自动驾驶打车服务。但前面在讨论百度地图的时候，消费者其实不关心你的技术好不好，关心的其实就是怎么把我送过去，那个时候要考虑车辆密度够不够。\n你可以认为是给车厂做服务商，大家都觉得这是不是我的核心利益？之前传出来百度要投理想汽车，一定要用阿波罗的系统，理想没有就选择了自建。我觉得在当前，有实力的车厂可能都会选择自己能够更强把控的系统，投资或收购自动驾驶的创业公司，但不会把整个自动驾驶方案都交给百度来做，他们可能会去采买百度某些特定服务，比如高精地图。\n3.3 汽车行业真的存在百度曾设想的大安卓系统的机会吗？Apollo能不能成为汽车行业的安卓？\n\n陈同学（百度自动驾驶业务）：在汽车产业的智能化变革中，百度Apollo希望利用确定性更高的技术和解决方案，为OEM客户在有限的时间窗口，去加速智能技术上车，避免大家重复造轮子。\n首先，从企业研发投入成本看，如果考虑到要开发支持智能驾驶、包含E/E架构的多车型平台，投入成本量级在100-200亿左右；而自动驾驶软件的技术研发投入也在数百亿级别以上，需要累计销售100-200W台才能实现成本摊销。\n其次，自动驾驶软件开发除了高额成本投入之外，时间成本同样巨大，需要经历能力构建、形成有效研发能力再到产品化输出的过程。业内自主研发自动驾驶软件的企业，通常至少需要6年以上的时间。最重要的是，研发成本，并不能保证得到可用的技术成果，包括测试牌照、地图、V2X路测资源、智能云等方面，都会影响自动驾驶技术的水平和落地时间。\nApollo希望利用自己8年的技术积累，用开放合作的形式，帮助车企打造具有竞争力的智能汽车，推动全行业的发展。","news_type":1,"symbols_score_info":{"BIDU":0.9,"09888":0.9,"QNETCN":0.9}},"isVote":1,"tweetType":1,"viewCount":2308,"commentLimit":10,"likeStatus":false,"favoriteStatus":false,"reportStatus":false,"symbols":[],"verified":2,"subType":0,"readableState":1,"langContent":"EN","currentLanguage":"EN","warmUpFlag":false,"orderFlag":false,"shareable":true,"causeOfNotShareable":"","featuresForAnalytics":[],"commentAndTweetFlag":false,"andRepostAutoSelectedFlag":false,"upFlag":false,"length":2,"xxTargetLangEnum":"ORIG"},"commentList":[],"hasMoreComment":false,"orderType":2}