Li Yanhong: Artificial Intelligence China does not lead the world and cannot explain

(Original title: Robin Li: Artificial Intelligence China does not lead the world and cannot be justified)

[Editor's note]

On June 29th, the first World Smart Conference with the theme of "Towards a Big Smart Age" was held in Tianjin. Baidu founder and Chairman Li Yanhong attended the conference and gave a keynote speech around "Artificial Intelligence? When Now." He pointed out that artificial intelligence is not the next scene, but this scene.

At the Baidu Union Summit last year, Li Yanhong used the “next scene, artificial intelligence” to define future trends. In just one year, the AI ​​era has now become the present. In the speech of the World Smart Conference, Li Yanhong pointed out that the world’s economic growth in the past 40 years mainly depends on the promotion of IT technology innovation, and the most important force driving the development of the world economy in the future is likely to be the progress of smart technology. He believes that China is "like a duck" in the AI ​​field and "may be naturally suitable for doing this." Li Yanhong also stated that the practical application of artificial intelligence has been accelerating to the ground. Intelligent voice interaction is becoming more and more common. AI has searched for achievements and the airport has been commissioned for boarding. The driverless driver is likely to be between 2021 and 2022. Become reality. Li Yanhong believes that all walks of life should grasp the broad prospects of AI. "We must work together to push smart technologies to a new climax."

The following is an account of Li Yanhong's speech:

Li Yanhong Oriental IC

Good morning, everyone is very honored to share some of my views with you at the World Smart Conference.

First of all, I would like to congratulate Tianjin. When the human race just entered the smart age, it was precisely at the right time that such a world smart conference was held. People with similar interests could be gathered together and discussed together. Whether it's the benefits smart technology brings to people in the future or the risks we may face in the future, these are very interesting topics, so it is very good for us to discuss together at this time.

Just like a few speakers just now, we all think that the era of artificial intelligence has come. In the next few decades, possibly 30 to 50 years, the most important force in promoting the development of the world economy is likely to be the progress of intelligent technology. Why do you say this way? If we look at history, we will basically reach the same conclusion. In the past 100 years, the growth of the world economy was mainly driven by technological innovation and innovation, not by population growth. In the past 100 years, most of the total economic growth was created by developed countries. The economic growth of developed countries was not brought about by population growth, but was caused by the improvement of people's production efficiency. If we look at the past forty years, one obvious feature is that technological innovations mainly take place in the IT field. I can say that the main driving force of world economic growth over the past four decades is the innovation of IT technology.

How to prove it? Let's take a look. In 1977, the top five companies in the U.S. stock market were mainly companies in the automotive and energy sectors, and only one related to IT was IBM. By 2017, the top five companies are now all IT companies. Why is this? Because so many years in the past, countless IT advancements have continued to advance the development of the world economy. Now, China's Ali and Tencent are gradually entering such a state.

As a technical person, I mainly want to see if we take the time closer. In the past two decades, where have the major IT innovations occurred? Someone may have heard a word called "go to IOE". I didn't understand it when I first heard it. What is "going to IOE"? It is to IBM, Oracle, EMC. Like Internet companies and search engine companies like us, IOE has not been used since its inception. The earliest time when the user is very large, we all need to go online to find things, but we can not buy so expensive server IBM, we can not buy Oracle so expensive, but still slow database management system. Because we are dealing with unstructured data, we have redeveloped a set of technologies to deal with unstructured data. We can support tens of millions of people and hundreds of millions at the same time using very inexpensive PC servers and software programs developed by ourselves. People search. So "go to IOE" solved this problem at the very beginning of the search engine company. From then on, search technology can be used not only in search but also in many applications. How to deal with unstructured data on a large scale in parallel? After these technologies generalize only for search, they are available later in other places. What is called after generalise? It is cloud computing. We all know that the concept of cloud computing was proposed by Google. They announced that the basics of cloud computing are GFS, BigTable, and MapReduce.

After entering the age of artificial intelligence, it is the search engine company that first used the original foundation and then continued to innovate, and pushed forward further, not only at the software level, but also at the hardware level. We all know that GPU was originally used to play games, and now the deep learning of artificial intelligence is basically done by the GPU. When Wu Enda had previously been at Google, it was said that he was very upset because Google did not allow him to buy a GPU. They did not believe this direction. Later, he could easily buy after Baidu. As a result, we have the world's largest GPU artificial intelligence deep learning computing system. FPGAs are also a new generation of system architecture that can be more inexpensive and more flexible to solve related problems. This explains why Baidu's network disk is still supporting, and other network disks are unsustainable, because the cost of the architecture we use is lower. Over the past two decades, search engine companies have made considerable contributions to computer science in terms of technology. Whether it is "going to IOE", cloud computing, or the widespread use of FPGAs and GPUs, it has always faced this problem with search engine companies, and has resolved It is inseparable.

Last year, I said at Baidu’s coalition summit that “the next scene is artificial intelligence.” In fact, after just one year, it was not the “next scene” but “this scene”. Now all people are aware that we are in the era of artificial intelligence.

There are new things in the age of artificial intelligence, and every time we see these things we are very excited. In 2009, Google began their autopilot project. Everyone knows that today both the Internet companies and auto manufacturers have realized that autopilot represents the future of the automotive industry. In January 2013, Baidu established the Institute of Deep Learning (IDL), the world's first enterprise research institute named after deep learning. Today, anyone who knows how to learn a bit of technology in depth will not dare to say that finding a high-paying job is no problem. In December 2014, Amazon began testing echo. This thing is very epoch-making. In the past, we all used mobile phones. Today, if we use something like echo, it can use far-field speech recognition. In the PC era, everyone uses the mouse and keyboard to interact with computers. In the era of smart phones, we use touch screens to interact with computers. In the era of artificial intelligence, it is possible to use voice and images to interact with computers. In December 2015, Baidu announced that the accuracy of our speech recognition technology has surpassed human human identification. In 2016 and 2017, Microsoft and Google also announced that their speech recognition accuracy exceeded human standards. If we get closer, we see that on December 5, 2016, Amazon launched Amazon Go, an unattended offline retail store. After you enter the store, whether it is picking things or making payments, you don't need to do it manually. You can rely entirely on machines to solve them. On May 4, we announced that we have changed Baidu's mission. Baidu was founded almost 18 years ago. During the first 17 years, we said, “Let people obtain information in the most equal and convenient way and find what they seek”. Everyone can feel that this is a mission with Internet characteristics. It is connected with people and information. But with the advent of the artificial intelligence era, we feel that we can do more than just these things. Smart technologies can change more and make the complex world easier, so we say "use technology to make a complex world simpler". Why do we say that the world is still complicated? When you go to the airport, you must also remember to bring your ID card and security check. Why can't you just leave your security card on the plane without a security check? We have to travel every day. Many people choose to drive, but they still have to learn to drive. Some people have to spend a month or even longer and spend thousands of dollars on driving school to learn how to drive. Why can't I sit down? Where do you want to drive in the car? For such things, we feel that smart technology can make the world easier. Including the TV remote control we use every day, there are dozens of buttons. Most people never actually use those buttons, nor do they know what the buttons are for. Why can't I just transfer to Tianjin TV? Tianjin Satellite TV, why can't I say that letting it be louder is bigger, I asked what the actress was called and it told me what it was. It could be solved with voice technology. The world will become simple in the future, and the way it becomes simple is to rely on smart technology. Last week, Facebook also announced that it had changed its mission. When Zuckerberg participated in the Facebook event in Chicago on June 22, he said that our past mission was "Make the world more open and connected." This is also true. Very characteristic of the Internet era is openness and connection. Now he also realizes that this thing is not enough. He said that we must "Bring the world closer together". How can we "closer together"? Still, through the user's portraits, through artificial intelligence techniques, people find the same interests and connect them together. The two people are separated by thousands of miles. If everyone is interested in peony, they can match each other with the user's portrait. There are more than 1,000 varieties of peony flowers, such as Yao Huang Wei Zi. However, when you say such a thing, no one around you can understand it, but there may be another person within a thousand miles. Interested in this thing, he can understand it, and two people can connect and be closer. It sounds a bit like a post, but it is indeed a new mission for Facebook.

Paying attention to artificial intelligence has become a global consensus. In the past year, if you talked with the leaders of any country in the world and talked about the topic of artificial intelligence, he could tell you a few words because artificial intelligence affects human society not only at the economic level but also at the political and cultural levels. The level has a very big influence. We have seen that the global technology giant’s investment in AI last year was US$30 billion, and people’s attention to AI is also unprecedented. We see that Baidu’s search for the word “artificial intelligence” has risen by 632% in 2016 compared to 2015, and in the first half of 2017 it has risen by 45%.

China still has a lot of advantages in artificial intelligence. We have a big market. We have a lot of talent. Everyone looks at artificial intelligence papers. Many of them are written by the Chinese people. We may be naturally suitable for doing this. We also have A lot of money. More importantly, if artificial intelligence technology is to advance, it needs a lot of data to accumulate and train. There is no market in the world that has more than 700 million netizens. They speak the same language and they follow the same culture. And ethical standards follow the same laws. You can't find such a market anymore. In such a market, you really have a lot of AI. We cannot lead the world. It is really impossible to say.

Therefore, foreign media also noticed that the "Washington Post" said that "China has led the United States in the field of artificial intelligence research." The "New York Times" said that "China is surpassing the United States in the field of artificial intelligence." Although the media has always been a bit unbelievable. Their words are always more extreme, but China’s achievements in artificial intelligence should be said to attract worldwide attention.

In terms of applications, artificial intelligence is emerging every day and is accelerating into various scenarios.

For example, face recognition, Yesterday, at Nanyang Airport has achieved a face brush plane, you can go directly without the boarding pass. In this way, things that we could hardly imagine in the past have actually happened.

And driverless? The global consensus is probably that between 2021 and 2022, driverlessness will become a reality. At present, Internet companies, automobile operating companies, automobile manufacturing companies, and auto parts companies have joined the ranks of unmanned research and development. This trend is unstoppable.

We also tried to use AI's technology and face recognition technology to help find lost loved ones. The most recent case was when a child in Chongqing lost at the age of four or five, and after twenty-seven years he lived in Fujian, through the recognition of face. He found his loved one. In another case, an elderly couple in Shaanxi brought his mentally handicapped son to Beijing for medical treatment. His son was lost because he had mental retardation and could not tell what his name was. Eight months after they had been in Beijing for eight months, they had been compared to our face recognition technology. When he found him, he was already full of faces, and he was completely different from what he usually saw. But the machine can recognize it.

The interaction of intelligent voices, such as talking to the TV and using voice to control TV, has now been realized. It has already landed in Changsha. With such a set-top box, it can be completely controlled by voice. You ask who this actor is and it really knows who it is.

The way I think about the artificial intelligence era is very different from the internet age. I have summarized some of them. First, we think that smartphones have been completely popularized and mobile phones will still exist for a long time. However, the opportunities for mobile internet are already running out. If you start a new business today, you need to rely on the mobile Internet. This is very difficult. Now. Second, we need to change the way of thinking from Think Mobile to Think AI. What is the way of thinking in the Mobile era? That is, everything slides by hand. When designing, it is very concerned about what the size of the font is. The navigation should be placed in what is pure software. But in the era of artificial intelligence, you must think The combination of software and hardware, when we meet in the company, it is clear that the product manager of the mobile product is very concerned about this function to use a few words to describe what the font size should be, what color; and artificial intelligence The product manager pulled out a chip from his pocket and said that my chip can now be made for 58 dollars. What's the function here? This is two completely different ways of thinking. So we also think that, thirdly, in the smart age, the combination of software and hardware will become more and more obvious. Fourth, many people talk about the data spike algorithm. Ma Yun has talked about the DT era for many years. Last month we met at the big data summit. What's particularly inconvenient is that we are a parallel forum. I felt stressful at the time. Everybody went to listen to Ma Yun’s speech. But fortunately, my field is also full. What I am talking about is an algorithm that really promotes social progress. At the end of the closing ceremony, Ma Huateng concluded that in fact they are not right. The most important thing is neither data nor algorithm. It is a scene. In fact, there is no need to do such an argument. Even I think that these are all routines. It is not a contradiction for everyone to say. When I spoke at the time, in addition to speaking about algorithms to promote social progress, I also said that past innovations came from universities and laboratories, and future innovations would come from data and scenarios. Fifth, because we have entered the age of AI and smart times, using AI thinking to make Internet products will have very big advantages.

So, I hope here that everyone will be able to grasp the broad prospects of artificial intelligence, because it really can change any industry, it will change the health care industry and the education industry. Just now, the Secretary of UNESCO’s Institute of Knowledge and Society also said that there are many We have accumulated knowledge, but how can people learn this knowledge more efficiently and more simply? Artificial intelligence has great potential for personalized learning and education. The financial sector now also uses artificial intelligence. Almost every industry, such as manufacturing, security, etc. will be very affected by artificial intelligence technology. Many forums this afternoon are discussing the influence of artificial intelligence one by one, so I will not go into details here.

What I want to say is that the age of artificial intelligence is not a patent of a certain company or a few companies. On the contrary, it is something that many companies do together. Recently I read Lee Kai-fu in an article in The New York Times. He said that the future of artificial intelligence will be cruel and the last chance will become a big company. There may be seven companies in the world that can truly benefit from the era of artificial intelligence to become large companies. Actually, I disagree with this view and want to do something. For example, Baidu has been doing driverless driving for several years and accumulated a lot of technology. However, we still feel that it is impossible for us to do this thing with our company. We need to introduce Tier 1 (Tier 1 suppliers) manufacturers such as Continental and Bosch. We must also bring in automakers, and we need to introduce operators who share travel. At the chip level, we cooperate with intel (Intel), and we also have to cooperate with a GPU company that depends entirely on the GPU, and cooperate with many companies such as HTC. Everyone can work together to push smart technology to a new climax.

I speak so much today. Thank you!

Capsule Slip Ring

What is a Capsule Slip Ring?

Capsule slip rings are a type of electrical connector that are used to transfer power and data between two rotating objects. They are commonly used in applications where there is a need to send or receive signals from a stationary object to a rotating one, or vice versa.

How Does it Work?

Capsule slip rings work by using a series of electrical contacts that rotate with the object on which they are mounted. These contacts are spaced evenly around the circumference of the ring, and as they rotate they make and break contact with corresponding contacts on the other object. This allows power and data to be transferred between the two objects without any interruption.

Advantages of Capsule Slip Rings

There are several advantages to using capsule slip rings in place of other types of electrical connectors.

For example, the rotation of the contacts causes them to be freely accessible and therefore free from dust and debris. This makes them ideal for applications where connector maintenance is often a problem.

Because the contacts are freely accessible, they can be readily replaced if a fault occurs. They also have very low electrical resistance, making them ideal for transmitting radio frequency signals. Because of the low resistance, this type of connector is often used in high-speed data transmissions.

This type of Conductive Slip Ring is similar to the shape of a hat. The brim part is a flange with a conductive slip ring. The function of the flange is to facilitate installation. It is connected to electrical equipment through the fixing hole of the flange. It can be firmly installed on the equipment. Therefore, the stator of the cap-type slip ring is generally fixed by the outer layer with a flange, and the inner part is the rotor with the inner shaft rotating. The stator and the rotor are not fixed, and the details will be different according to the installation method.

Capsule Slip Ring,Split Ring Motor,Slip Ring Motor Connection,Slip Rings In Synchronous Motor

Dongguan Oubaibo Technology Co., Ltd. , https://www.sliprobs.com