This article is produced by NetEase Smart Studio (public number smartman 163). Focus on AI and read the next big era!燑/font> From | Wired 燑/font> Authors | Jessi Hempel Compilation | Netease sees the outside compiles the platform 燑/font> Review | Song Yufei A company can have the most advanced technology in the world, the best talent and the most innovative product design. However, companies need data to build computer programs that are like intelligence that can change our city life. That is to say, a database-rich company will win. This is why, after resigning Microsoft earlier this year, Lu Qi went to Beijing to serve as the chief operating officer of Baidu. In his previous work, he was the chief assistant of Microsoft CEO Satya Nadella, helping Microsoft to develop artificial intelligence. Obviously, Lu Qi saw the opportunity on the other side of the Pacific Ocean: In China, the number of people online reached 7.31, which is almost twice the total population of the United States. Lu Qi said: "China has a structural advantage." On July 26, when Lu Qi visited Silicon Valley, he accepted an exclusive interview with Wired. Regarding “How Baidu Will Lead the Chinese Artificial Intelligence Marketâ€, Lu Qi gave an illuminating explanation. Lu Qi pointed out that many houses in many parts of the world have much in common with Chinese small houses, which is very different from the large-scale luxury villas in North America. Lu Qi thinks this will be China's biggest advantage in pushing artificial intelligence to the world. Of course, the U.S. tech giants may have an advantage in human resources. However, Lu Qi believes that Baidu has the ability to conquer the world. The following is an interview with Wired reporter Lu Qi quiz, compiled by NetEase smart (public number smartman 163): Jessi Hempel (Reporter, JH for short): Baidu has reorganized since you joined Baidu. Then I would like to ask you, as chief operating officer, what are you currently responsible for doing in the company? Lu Qi: In my work, I have a very good relationship with Li Yanhong. We are fighting side by side. First, I am primarily responsible for product development, sales, and marketing so that we can ensure that our overall strategy is fully synchronized. This is the first. Second, in terms of marketing strategies, we now have a clearer and more focused goal. This is really like two battles. One is to strengthen the foundation of our mobile communications and the other is to lead artificial intelligence. JH: How do you see Baidu's artificial intelligence technology? Lu Qi: We believe that the best way to commercialize artificial intelligence is to build an ecosystem. Essentially speaking, it is for our partners to accelerate the pace of innovation and use a healthy and stable economic model to build a long-term win-win situation for our developers and partners. The basis for achieving this is Baidu's brain (referring to Baidu's all AL resources). Currently, Baidu is much broader than Microsoft and Google provide because it is a platform. The Baidu brain has more than 60 different types of Al services. We are also the first big company to clearly distinguish the perception and cognitive layers. Perceived ability and cognitive ability are both related and completely different. Most Al platforms will confuse the two. JH: What are the advantages of Baidu compared with Apple Siri and Microsoft Cortana? Lu Qi: We focus on connecting customers and partners with two platforms. The first one we call DuerOS. It is an artificial computing platform based on natural language and dialogue. It is very similar to Alexa, Google Now, Siri and Cortana in the United States. The only difference is that DuerOS is ahead of other competitors. In China, DuerOS has accumulated more dialogue skills than other competing products. The intelligent dialogue technology we have developed includes 10 main areas and more than 100 sub-fields. We are building an emerging partner ecosystem. Our partners are also developing more and more smart combinations. In contrast, Amazon's smart portfolio is much larger than that of Baidu. Because they have a bigger partner ecosystem in the United States. But in China, Baidu is without doubt the market leader compared to most companies. Second, we are also the leaders of our partners. Today, DuerOS has hundreds of brands of home appliances, including refrigerators, air conditioners, televisions, children's stories and refrigerators. JH: How is voice technology in the U.S. market compared to the Chinese market? Lu Qi: The domestic environment is very different from the United States. Because we are talking about voice interaction. The sound environment and noise patterns of the two countries are quite different. Alexa Echo and Cortana are both upgraded and optimized for the living environment in the United States. In my opinion, this applies only to a small part of North America and Europe. Essentially speaking, this idea is based on the fact that you have a spacious house and there are several rooms in the house. But in China, this is not the case. Our goal, even for young high-income people, is generally only 60 square meters of houses, sometimes reaching 90 square meters. We have a better chance for DuerOS to face the global market. Why? Because Japan, India and Brazil are very close to China, they are far away from North America. So, this is the difference. So what do they have in common? That is technology. The core technologies are still speech recognition, signal processing, natural language understanding, and intelligent platforms. In many ways, our platform structure is very similar to Amazon. In my opinion, Amazon is doing very well now. Even if I worked at Microsoft, but to be honest, Amazon is leading. JH: Do you think that Amazon's weaknesses are on its back-end data and it's technically difficult to catch up with Google and Microsoft? Lu Qi: Four and a half years ago, my research project was Cortana. At the time, we all felt that Amazon's technology was backward. But in the artificial intelligence competition, I understand one thing: It is actually more important to have suitable application scenarios and ecosystems. Technically speaking, Google and Microsoft are far ahead of Amazon. But look at the current artificial intelligence and you will understand. In the United States, Amazon's Alexa is far ahead of other competitors. This is because Amazon found the right application scenario and standardized equipment. They repaired the equipment. Essentially, this is an Al-priority device. Microsoft and Google also made the same mistake. We put Cortana too much on mobile phones and personal computers, the most serious ones. In my opinion, in the short term, this phone will become a manual touch and mobile priority device. You need an Al First device to strengthen and consolidate the emerging ecosystem foundation. Now in China, what it means to Al First is much clearer than ever. This means that from the beginning, you have to interact in different ways and technologies. On the first interaction, this approach must be voice, image recognition, or facial recognition. You can use the screen or touch, but this is secondary. All systems at Baidu’s headquarters are face recognition. There is also a Baidu vending machine where you can buy goods through voice and face recognition. We are still working on a cafeteria. The goal is that you can take food and leave without manual charges. Technically speaking, this is feasible in many places, but this does not mean that people will accept it. I do not mean that all technologies are applicable. It is also related to the environment, culture and policy. This is why, in my eyes, artificial intelligence has great potential in China. JH: So, what are the consequences of the equipment we are developing now? People are now talking about Baidu, just like Microsoft? Do people have the same conversation with Microsoft on Baidu? Lu Qi: Actually it is almost. For us, protecting privacy is very important. Ultimately, our users will trust our technology. Therefore, we will discuss more in this regard. In terms of privacy protection, we will increase investment to ensure that users will trust our technology. For example, we discussed speech interactions. We are developing a technology that prevents accidental activation of smartphones. We understand that people do not want to let everyday private conversations reach the cloud. I also have private conversations in my living room at home. But sometimes smart speakers think you are trying to start them and they will send these voice messages to the cloud. JH: Do you think Chinese consumers will be concerned about this? Will the expectations of Chinese consumers be different? Lu Qi: We think they will be very concerned about this because I believe that people are rational after all. If the benefits are very attractive, people will weigh the pros and cons before making a decision. I think this is actually global. JH: This spring, Baidu announced an ambitious drone project "Apollo". Up to now, you have announced 50 partners. Why is Baidu investing so much in driverlessness? Lu Qi: If you want to really develop digital intelligence to acquire knowledge, make decisions, and adapt to the environment, you need to develop autonomous systems. In the autopilot system, the car will be the first large commercial application to be implemented. Just like today's smart phone systems, the mobile ecosystem is the largest Silicon Valley software ecosystem. I believe that the same thing will happen on the autopilot system. Baidu Motors will build a bigger ecosystem. The same combination of skills, hardware, sensors, chipsets, and software will be used to develop industrial robots and home robots. We hope to combine hundreds of companies and universities to study this matter and establish a very large ecosystem. Later we can develop robots, drones and all these autonomous systems. So, for me, autonomy is the core. JH: "Apollo" R & D, you help this very big? Lu Qi: I am the chief operating officer of Baidu, but I am directly responsible for product operations. In the past three months or so, I spent about 40% of my time talking to customers and partners on autonomous driving technology. From the current situation, to achieve fully automated driving, the basic technological path of autonomous driving technology is its iterative speed. JH: What is the speed of iterative auto-pilot technology? Lu Qi: Essentially depends on how much data you can get. Because if you want to drive on the highway, you must deal with different road conditions, such as light, weather, whether the road slips, tire pressure, and so on. In the Apollo project, we can integrate all resources, especially data resources, so that everyone can benefit. We wrote a statement about Apollo. It has four principles. Each one is important. The first is openness. At Baidu, we put technology—codes, services, and data—to our partners. This is very effective in China because China is a highly fragmented market. There are more than 250 automotive OEMs in China, unlike the United States, where the industry is highly concentrated. No original equipment manufacturer can completely independently develop. With our open code base on July 5th, a person can assemble a car in three days and can autonomously drive and develop in a limited mode. The second is resource sharing. In essence, Apollo plans to have two levels of cooperation. You can use Apollo's code and features, as well as some data sets, without any additional conditions. The second is to let you use all the data provided by Baidu - HD maps, training data, but we will also ask you to contribute the corresponding data. Here is a principle. The more data you contribute, the more data you get. The third principle is to accelerate innovation. Essentially, we have the ability to integrate more data and we can get more performance from the simulation engine. We can enable everyone to innovate faster. The fourth principle is continuous win-win. Baidu is the best example. Baidu will focus on providing high-end and high-value services, high-definition maps, and security services. We don't compete with anyone. We allow every automotive OEM to do more, whether Bosch, China or Nvidia. This is the reason why I founded overseas subsidiaries, Apollo USA and Apollo Singapore. The attitude of the Singapore government is basically: "Wow, come to Singapore, we are ready to invest!" JH: What needs to be done to achieve a fully autonomous car in China? Lu Qi: In terms of single technology, it is impossible to achieve driverless driving for a long time. I will give you a simple example. Suppose there was a traffic accident in one city. The police came and found no traffic signs on the road. So the police drew a sign on the paper and wrote "The speed must not exceed five miles per hour, and pay attention to changes in the road conditions." The police raised the paper to the passing vehicles. You need this technology to recognize handwriting and understand human language. This will take a long, long time to achieve. To make fully-automatic driving a reality, new traffic rules and laws and regulations need to be formulated. This is the first. Secondly, in the course of the “Apollo†project and collaboration with partners, we discovered that there are many opportunities for commercialization before fully automated driving. The Audi A8 is a good example. It can automatically follow the traffic flow in heavily jammed traffic. Traffic congestion is almost commonplace in Beijing and Shanghai. Now you can let the car drive itself. You can read books or do other things. In addition to following traffic, there are many other applications for driverless cars. JH: You were still at Microsoft when we met for the first time. Now you are working for Baidu. Why is this? Lu Qi: In October 2016, I broke my leg. I need to do two surgeries. In fact, Gates, Satya, and I are still in a very good relationship, so when I go to Seattle, I will go to Satya's home with Gates and I promise to be their personal advisor. JH: 2017 seems to be a trend indicator for the development of artificial intelligence in China. What is the significance of this year? Lu Qi: This is a combination of two factors, the maturity of the technology, and the number of vertical industries that can commercialize Al. From a global perspective, I can feel that China and the United States have great room for cooperation to jointly promote world development. I may be influenced by Bill Gates. He always said that for the purpose of practice, the current global economy can be regarded as a single-engine economy. The United States has a population of 5% of the world's population, but it has 24% of economic output and 60% of innovation capacity. Because there are 7 billion people in the world, this status is simply not enough to sustain economic growth. Of these, more than 3 billion people live a modern life. We have transportation, eating processed foods, and we have refrigerators. But beyond this 3 billion people, there is also a huge gap in the quality of life. The rest of the population lives in a completely different environment. Our task is to let everyone live a modern life. But how can this be done? This requires more innovation and better economic growth. In fact, China is the second largest innovation country. Bill Gates believes from the bottom of heart that a more innovative and more developed China is a good thing for the world. I believe this too. JH: A few years ago, when Baidu began to accelerate the development of Al resources, you focused on building the Silicon Valley lab. When Wu Enda left Baidu in the spring, instead of leading him to Baidu Al's research, he was in China. Has China's Al surpassed the United States? Does artificial intelligence in China catch up with the United States? Lu Qi: There is no doubt that the United States as a whole is much more powerful, but the gap between China and the United States is also continuously narrowing. There is no doubt about this. Since I have been in China for more than 6 months, honestly speaking, I read many newspapers and communicated with Al's developers. I can feel the talent reserve. Baidu will conduct more and more Al jobs in China. But at the same time, Baidu will continue to invest in the United States, San Francisco and Seattle. We just established a branch in Seattle because we acquired a company called Kitt.ai. For Al's senior staff, the United States still has an advantage. We want to make full use of this. Pay attention to NetEase smart public number (smartman163), get the latest report of artificial intelligence industry. Aspire Vape pen, Manufacture Aspire Vaporizer, Aspire Vape pod Shenzhen Xcool Vapor Technology Co.,Ltd , https://www.szxcoolvape.com
