(Original title: Shen Xiangyang has no fear of BAT compulsion: Microsoft has cultivated the CTO of almost all IT companies in China, and 5,000 people hold up the AI's four major directions)
On May 10, 2017, Dr. Shen Xiangyang, executive vice president of Microsoft, Microsoft Artificial Intelligence and Microsoft Research Division, delivered a keynote speech at the Microsoft Global Developer Conference (Build) and accepted Xin Zhiyuan. Interviewed by several invited media.
In the interview, Dr. Shen Xiangyang believed that Cortana represented the future of AI. He mentioned Conversational AI several times, namely dialogue AI. In his view, the smart voice market is far from saturated, and all major giants are doing intelligent voice assistants, which shows that this is the right direction. In the face of Amazon products such as Amazon Alexa/Echo and Google Google Home, Microsoft's opportunities There are still many. Microsoft will use the advantages of Xiao Na and Xiao Bing to make efforts in terms of voice.
The difference is that Shen Xiangyang mentioned that he believes that "AI's understanding of human emotions will be a very remarkable thing." "Most people have yet to realize the importance of being able to continue to talk about this matter." . At this point, Microsoft's Xiao Bing is far ahead of other intelligent voice assistants, and the average number of rounds of conversations is currently the highest in the industry's 23 times. However, Shen Xiangyang also stated that Microsoft’s Conversational AI will “walk on two legsâ€. It will not only give full play to Cortana’s business assistant advantage, but also fully exploit Xiao Bing’s advantage in emotional chat.
When Xin Zhiyuan asked about the specific organizational structure of the 5000-person AI team that Microsoft organized last year, Shen Xiangyang did not directly answer. However, he stated that search engines, intelligent voice assistants represented by Cortana, the use of AI to help transform existing Microsoft products, and the development of new product lines using AI are currently the four major directions for the Microsoft AI team. In particular, the last point, Shen Xiangyang firmly believes that AI will subvert all commercial applications. In this case, it is Microsoft AI's most important thing to decide what to do and what not to do.
As for the loss of highly talented Microsoft AI talent, Shen Xiangyang seems to be not too worried about this. He told Xinzhiyuan: “I think the flow of talent is very normal. A large company has trained many talents, and the most important one is a company. You need to know the value and philosophy of your existence. This is reflected in why excellent employees choose to stay with you. Not only are you digging outside, what kind of amazing environment you provide to the staff here, let him be there. There is a very good development.†For other companies to come to Microsoft to “knock cornersâ€, Shen Xiangyang feels very proud that it reflects that Microsoft has cultivated excellent talents for the Chinese AI and even the entire IT industry.
Shen Xiangyang also said that after Microsoft reorganized and built a 5,000-person AI team last year, he set up a "Microsoft Artificial Intelligence College" to train a considerable number of Microsoft internal talents. "We hope that in this way, we can attract more outsiders." The talented people come to Microsoft. In doing so, we are not worried that someone will dug our AI talents. What's more important is to cultivate our talents as 'AI capable'."
Next, when answering questions from other media reporters, Shen Xiangyang stated that computer vision can identify everything in the next 10 years or so. He believes that in the next 3 to 5 years, vertical AI will develop and be optimistic about toB's business. In the next 5 to 10 years, he believes that "interpretable AI" is a very worthwhile research direction. breakthrough. In addition, Shen Xiangyang also said that cognitive computing will become Azure's "emphasis." It seems that in the Intelligent Cloud, Google and Microsoft are bound to have a wonderful match.
In the interview, Shen Xiangyang mentioned that he enjoyed the process of research very much. Unfortunately, he was “pulled out by Ballmer to make productsâ€. In addition, Shen Xiangyang believes that "you can learn a lot from the Android rise process and how you can do better in the next wave of deep learning. I don't think that a Microsoft company can do all these things."
How to respond to Tencent and other companies competing for Microsoft AI talent?
Xin Zhiyuan: There is a problem. How does Microsoft deal with brain drain? To be honest, like Tencent has already opened the Institute to Seattle.
Dr. Shen Xiangyang: Last time I told Ma Huateng, I said that you were too inconvenient and drove the Institute to our door. Pony brother is still a bit embarrassed, but also a little embarrassed. I think that the flow of talent is very normal. A large company has trained many talents. The most important thing is to say that any company you want to know the value and philosophy of your existence shows that why good employees choose to stay with you, not only When you go outside to dig people, what kind of amazing environment you provide to the staff here, so that he has a very good development here. When you ask Tencent, Tencent's talents are constantly losing. Tencent didn't have AI talent before, so he must go to dig people, go to Baidu to dig or dig elsewhere, and it is normal. Everyone to the Seattle Institute of Research and Development is very good for Seattle's local economy. In China, we also go out to attract businesses and pull people over, like Microsoft, Amazon, or other big companies. We also have the responsibility to drive the local economy in Seattle.
Xinzhiyuan: Just as you just said, how to retain AI talent is the strength of this company. However, from the current news, there are many people out of Microsoft Research.
Dr. Shen Xiangyang: Yes, when I last reported to Secretary Liu Yunshan in Wuzhen, I told everyone that Microsoft in China has trained countless talents for the IT industry in China. I said that you should not only see almost all IT companies in China. The CTO was developed by me. From Lenovo to Haier to a small company, I said that you need to see that Microsoft Research has trained 5,000 students in the past 18 years. Those are really amazing. A new generation of startup companies is coming out. - This may not be the case now - when computer vision, AI company just got up, those investors came to ask me some situations, I said you do not want to tell me, the computer vision company in China is either What my students have opened is either that of my students' students. This is a very glorious thing. Looking back, Microsoft Research and Microsoft have played a huge positive role in China's IT development in many aspects, and especially trained a lot of first-rate talent for China.
Xinzhiyuan: Then you also feel that Microsoft is also very proud to continue exporting talents.
Dr. Shen Xiangyang: We have always felt very proud. Microsoft Research Institute has always been very proud. We trained Li Kaifu. We trained Zhang Yaqin. We trained Zhao Feng. We trained Yong Yong, Yong Yong, or my brother. All of these were very good. We all feel very honored.
Shin Chih-yuan: If you continue to leave with senior research fellows, do you still feel very honorable?
Dr. Shen Xiangyang: You should specifically ask those who left. Are they still researchers? Do you understand what I mean? You ask them what they have studied recently and ask him to write an algorithm for you to see.
Let me add that Microsoft has recently done a very important thing, that is, talent. Shortly after the establishment of the AI ​​department I led six months ago, I established a "Microsoft Artificial Intelligence Institute" to train a considerable number of Microsoft internal talents. We hope that in this way, we can attract more outside talents to Microsoft. . In doing so, we are not worried that someone will dig our AI talent. More importantly, we must develop our talents into “AI capable,†so we have a prefixed lesson, a double-headed lesson, and a three-headed word. Class, until the six-headed class, similar to a class like graduate students. I recently completed a course on AI 611, specifically doing the hands-on project to learn more about specific projects. There were 10 projects. It was amazing. When they finished this class, I spent 2 hours listening to their reports. Very good! Therefore, we are now continuously developing AI talents, both internal talent training and attracting outside talents.
In the rankings of the best employers of global and Chinese technology companies released by LinkedIn, Microsoft ranks second in the world, second only to Google. But in China, Microsoft has surpassed Google to become the best employer. For details, please refer to the two reports of the Xinzhiyuan public number: "LinkedIn VP Wang Di's Interview | Global AI Talent Gathering, Top10 Sino-American Companies Ranking".
Microsoft 5000 AI Corps: Force Search, Cortana, etc.
Xinzhiyuan: Microsoft reorganized last year. You led a 5,000-person AI team. But we don't have much understanding of the specific organizational structure. I hope you can introduce this in this regard. The 5,000-strong team What is the specific focus, direction, and personnel structure? I believe that 5,000 people are still distributed around the world. What kind of contact methods do you take this time?
Dr. Shen Xiangyang: We usually do not talk about the specific structure and personnel of the outside world. Of course, there is an organizational structure within the internal organization. However, I can simply tell you that Microsoft's development of AI is mainly in four areas.
First, the search engine. You can agree, or disagree with me, that the world’s largest artificial intelligence may still be a search engine. Microsoft Bing has spent many years doing knowledge maps and more than 2.5 billion entities. There is a lot of knowledge here. The search engine itself is not just a business - of course now Bing is also very profitable. We account for 22.6% of the search market in the US, plus 11% of Yahoo! (Backstage technology is Microsoft's ), so we have a 1/3 search share in the United States, which recently rose to 16.5% in the United Kingdom. From the perspective of AI, search engines are mainly World Knowledge - the accumulation of knowledge in the world.
The second, very, very important thing is Cortana. This is a private intelligent digital assistant. I think Cortana represents the future of AI and people's understanding. To do a good job of AI requires three aspects of knowledge, one is your understanding of the world, one is your understanding of the work, and the other is your understanding of the user. These three things can be done very well. I think Xiaona is walking in this direction. Of course, there is a lot of investment in this.
Third, other companies cooperate with Microsoft, how to make AI (4) æšã„‹î‹„   ┳ è…‹ è…‹ è…‹ è…‹ è…‹ è…‹ è…‹ 岬绊 岬绊 岬绊 岬绊 岬绊 岬绊 也 也 也 也, also mentioned Cloud, also mentioned Windows, we do together . At the same time, we have put some of these technologies out to all other Microsoft developers, such as Cognitive Services, which is the focus of my speech. Cognitive services have been done by our AI department for so many years, and many of them are made from Microsoft Research.
Fourthly, I feel very excited about myself. I made a slight mention at the end of the keynote speech and I did not have the opportunity to speak. That is, Every Business App is going to be Disrupted by AI—“All commercial applications will be AI. subversion". Therefore, we are choosing which direction Microsoft is taking and what business AI opportunities we will explore. We hope to have the opportunity to share our progress with you in the near future.
Xin Zhiyuan: Are these directions still in the discussion stage?
Shen Xiangyang: We have already done a lot of product prototypes, and even a lot of products and services have already been used by many people. I also mentioned in my speech that we are all doing this from sales to marketing to after sales.
Conversational AI: Microsoft has some different opinions
Xinzhiyuan: There are many companies that use Conversation AI as an entrance, but others think that dialogue is a pseudo-need. From your introduction just now, can I understand this way, Microsoft is different from other companies, is that Microsoft's dialogue platform is to cut in from the perspective of emotion?
Shen Xiangyang: Dr. Microsoft not only helps you complete the task, Microsoft is also very concerned about the emotional development, I think we are two ways to go at the same time, we Xiaona is to help you, what do you have, I help you Finding the necessary problems, at the same time we Xiao Bing is more from an emotional point of view, chatting with you, trying to understand you, and continue to talk to you. Everyone's design concept is very different. For example, if you want to help people accomplish tasks, you want to be as fast as possible. If you want to chat with people, you have to be able to talk for as long as possible, these two things from product and design. The above is very different.
Most people have yet to realize the importance of being able to continue talking about this matter. I will give you a figure and we all do not know. One person speaks a few words a day? In fact, people can speak more than 6,000 sentences a day. You may not realize this. Why do people have to keep talking? It is definitely not because they want to accomplish one thing. It is the process of human self-learning. Through dialogue, through communication, and learning different things from different people, this is Microsoft's AI, In particular, conversational AI, a very important point, other companies may not necessarily pay attention.
The most important thing about Microsoft AI: What to do and what not to do
The following is an interview with Dr. Shen Xiangyang who interviewed the reporters to answer questions from other reporters.
Q: Microsoft is fully committed to AI. Will your department continue to expand?
Dr. Shen Xiangyang: You have to ask our CEO (laugh) for this question. I think it will happen.
Q: What is the next priority, and is there a process of development, such as short-term and long-term planning?
Shen Xiangyang: There are some. Any company or unit, especially after it is big, must think about what the short-term goal is, what the mid-term hope is, what the long-term vision is, and we must think from this perspective. What's special about my department is that I manage the Institute in addition to AI. At the Institute we have more than 1,000 scientists. Just now Xin Zhiyuan's colleagues also asked, I think the most important thing is to continue to train new generations of amazing researchers and do even more remarkable techniques.
For example, I just mentioned the integration of artificial intelligence (more symbolic processing in discrete space) and brain science (more continuous processing in Neural neurons). The relationship between the symbolic interpretable space and the continuous brain space, basically no one studied these directions three or five years ago. I think the more important thing is to cultivate a new generation and see such problems. If you study deeply, you have to explain it.
Back to what you just asked, what kind of goals does Microsoft have? I think it's pretty clear. In the short term, AI still needs to see short-term business and where there is real business value. That is, we can earn some money and set a small goal. This is very important. However, the more important thing is that after the establishment of the AI ​​department, we have to think clearly if we really think that AI will subvert more industry applications. In the process of subversion, where are our opportunities?
There are two main aspects: One is the combination of existing products, such as Office and AI, what disruptive content will be produced, what new products will come out, what new features come out, and we will make progress here. It's very good; the other you have to think about is where the new product line is, whether you have a new product line, and you can do billions of dollars in business after 3-5 years. To think about it, if you have such a business, five to ten years can be a business of 10 billion dollars. If there are, of course, we must put the horse to pursue such an opportunity. So the most important thing for us now in the entire AI department is to decide what to do and decide not to do anything.
In the next 10 years or so, computer vision can identify everything; in the next 5 to 10 years, interpretable AI is a worthwhile research direction, and there will certainly be important breakthroughs.
Q: Hello, Harry, the artificial intelligence research directions you mentioned earlier can be roughly divided into two aspects: perception and cognition. I would like to ask you to explain the trend of the two in a more technical language. Let me give you an example.
Dr. Shen Xiangyang: When it comes to artificial intelligence, everyone is very excited today. I also think that something is really happening. First of all, if you look back, why is there the word "artificial intelligence", which actually corresponds to "human intelligence"? Human intelligence.
Why do people think that people have intelligence? People's intelligence is basically divided into two parts, one is perception and the other is cognition. Perception is the greatest of all, and the greatest part of it is the perception of vision. Someone had done such research before. About 91% of a person's information was collected from the visual. I forgot what figure he used to figure out this figure, but I think everyone basically agrees that the vast majority of people perceive it. All from the visual. Then there is hearing, then there are other perceptions, such as touch.
I think the progress here is very large. I have been talking about this past year. The breakthrough in computer speech recognition is at most 5 years, that is, within 5 years of computer speech, no matter how you talk about it, you can recognize it. And certainly stronger than humans. Then, in the next 10 years or so, computer vision will reach this point. Today, many things in vision have surpassed people, specifically such as face recognition. But what I'm talking about is a very general, anything is recognizable, to a new place, AI sees a new thing can associate with other things, I think about 10 years or so can be achieved.
However, in terms of cognition, today we are far from achieving breakthroughs and we are not talking about human beings. The first is the problem of natural language processing, and then the problem of knowledge acquisition. More and more people should do more work in this area. Natural language processing, I have just mentioned that machine reading and language are relatively complicated. With today's existing methods, including deep learning methods, the results obtained are still not good enough. Of course, using deep learning can already help us with many things. For example, translation also uses many natural language things.
More importantly, today's definition of the whole "cognitive" thing is still in a relatively elementary stage. What is Common Sense? How do you know that after seeing this person, why would he be very close to him? We do not understand this yet, and this is a big problem. Slightly speaking, it is a very important issue. The combination of these things that we have done today with personal intelligence is not enough to integrate with the brain science and understanding is not enough. The main reason is that there are only a few examples of the "intelligence," such as the human brain. However, the human brain has a very special structure. Today we don't understand enough. As a science, brain science is still in an early stage today. We can't do too many experiments, and we can't open a person's head and stuff stuff into it at any time.
This is a long-term problem, and now more and more people are thinking about it. Joining artificial intelligence and brain science, we feel that there is a direction today that is very exciting. Now many people at the Microsoft Research Institute are doing things in this area. I also have some cooperation with many universities, which is called “interpretable AI. "(Explainable AI)." In my opinion, Explainable AI can certainly make great achievements in the next 5 to 10 years. If I have a graduate student today, I will let them do the work in this direction. The reason is very simple, because the biggest breakthrough of AI today is deep learning, but one of the biggest problems of deep learning is that the results are very good, but you Can not explain.
The best written article in this area, which I have seen myself, is a recent article in the New Yorker about medical AI. Why do you see the same picture, the doctor will tell you, you have no problem, the reason is one, two, three. But AI still cannot do this today. Deep learning can't do this. The big problem is that the space for everyone to solve the problem is not the same. The doctor is in such a continuous space as the so-called neural, brain. A lot of AI's understanding is done in the discrete space of symbols.
Therefore, how to link these things up and technically speaking, there are many areas that need to be broken through, and this is also one aspect of our current research institute that is very serious in scientific research.
Q: What are some of Microsoft's existing achievements or advantages?
Dr. Shen Xiangyang: I just mentioned that in terms of perception, from computer vision to computer speech, I think we are all leading.
Q: Is there any data in this area or a particularly strong case?
Dr. Shen Xiangyang: We are the first to achieve human recognition of speech recognition, especially in the most difficult data-level switch bot data set. I mentioned in my speech that we had already done this six months ago. In terms of computer vision, we achieved Object Recognition the same accuracy as humans for the first time two years ago. Next, I hope we can make breakthroughs in natural language understanding. The same is true for translation. The quality of translation, especially Spoken Language, is certainly the world's leading level. But there are so many languages ​​to do, there are many engineering problems to solve, and today Microsoft only provides 60 languages. There are more than 6,500 languages ​​in the world. How do we popularize artificial intelligence? considerate.
Microsoft Asia Research Institute is not a non-profit organization. Researchers are responsible for the company and must turn technology into products.
Q: How to actualize the research and commercialization of research in the Institute. Did you change from the research to the product process when you set up this team? Another is the development of new products, such as Demo on the first day of the conference, to help designers suffering from Parkinson's disease re-implement painting, how to expand such research cases?
Dr. Shen Xiangyang: This is a very good question. I don’t know how many people have done research. Scientific research is a very different thing. Many people do not understand, and say that so many people do scientific research, how have not yet achieved productization. Scientific research is a very long-term thing. You should either say that you do scientific research or you need to be patient with scientific research. For example, today we talk about the sky-rocketing quantum calculations, which specific days of quantum computers can be made, no one knows. Someone knows that they have to, VC will throw out like a wolf (laughs). At this point, the U.S. system is still worth our studying. From the university, to the institute, to the industry. Former AT&T and IBM, and now Microsoft, many companies are willing to spend a lot of money to do long-term research, and the vast majority of scientific research results do not only belong to the company, but their own company may benefit. In this regard, Gates made it very clear. For example, Apple and Microsoft's early success was very important to the graphical user interface. Graphic interface was first made by Xerox. We learned from them. Similarly, Microsoft has done a lot of amazing things today, but it may be normal for other companies and even some startup companies to do better.
Moreover, doing research is a very pleasant thing. The happiest thing to do in scientific research is to not worry about what others are thinking. You can pat your brain and think about it. There is an amazing idea of ​​how much I used to be. He enjoys doing scientific research and was eventually driven out of Ballmer to make products.
Recently, I talked with colleagues of Microsoft Asia Research Institute. In fact, we can provide the most amazing thing is an environment where not only everyone's treatment is also good, but more important is the side of people who are smarter than you, you can They learn. Second, we provide a free space at Microsoft Research. We never ask you to do anything. Researchers want to do what you want. What do you think should be done? Because you are already one of the smartest guys, you have Microsoft Research. When several people choose to leave, go out to do VC, change a company, this is normal. We will continue to train the next generation of people. What is important is that you have an environment where you can continuously cultivate talent.
Of course, because you are not a public welfare research institute, of course, you are responsible for the company, including the most important thing, that is, the transformation from technology to products. At present, AI itself is still studying a lot of things today, so today Microsoft reorganized and put AI and the research institute in the same department - I feel very honored to lead such a department - we see so many opportunities. How to transfer from technology to product, perhaps the best example today is cognitive services, of which about two-thirds of the technology was originally done by Microsoft Research, and it has been done for many years. We were not very clear before. Some computer vision How the technology is transformed into a product. However, because of Azure, there are opportunities for cognitive services. Many of the technologies of Microsoft Research have been transformed into products through cognitive services.
I will give you two examples. In Hawaii this July, I will go to CVPR, the largest international conference for computer vision. This year they asked me to give a keynote speech. The topic of my speech is "Commercialize Computer Vision - Success Stories and Lessons Learned". . I will give three examples here. One is cognitive services. Another example is that we have recently launched an amazing camera iOS app called Microsoft Pix. Everyone has a chance to try it. When the first edition came out, there were 12 previous papers based on it. Of course there are many other company camera applications that do very well, including the iPhone's own camera that does a pretty good job.
The third example is my favorite HoloLens. The process of R&D is the process of “research, development, research and developmentâ€. HoloLens's gang used to be Kinect. Kinect made Microsoft Research to do Kinect Fusion, and also made a project called Holodesk. If it is three-dimensional, how do you add some three-dimensional virtual objects? Afterwards, Microsoft had a number of great engineers who worked on products and design. They thought that HoloLens should be built on it. During this process, many of the computer vision and voice technologies were developed together by Microsoft Research. The process of research and development.
In the end, you still need to have a product, because for the general public, he will not read the paper, nor do you know how great your paper is. You want to explain to people, the easiest way to explain is that you give people a look at your product, HoloLens will be immediately seen, we hope to have the opportunity to do something even more remarkable, not only to consumers, but also more. Product for business users. Personally, I think that within 3-5 years, the biggest opportunity for AI is still in the corporate market.
What is the difference between Microsoft's Cortana and Google Home and Amazon Echo?
Q: You just compared Microsoft's Cortana, Google, and Amazon's related products. It is said that each service is very different. Can you give me an example? In addition, domestic AI companies have a point of view, if you do pure 2B developer's business, if you provide tools, it may not be a good business model in the domestic environment, I feel very hard not to make money. But your definition of Cortana is a 2B tool. I am more curious about what you think?
Dr. Shen Xiangyang: The first question is relatively simple. I would like to emphasize that other companies are doing very well in digital assistance, but they are very difficult to do. A recent survey of individuals on the Internet made it look like 75% of the time. If you ask Alexa, she will say "I don't have an answer." In fact, voice today is not because of how one has already been done. Others do not have to do it. [The market is still far from being [saturated] to this point, Google is doing it, and Amazon is doing it. , Apple is also doing, Microsoft is also doing, many domestic companies are doing, very good, at least we all see this direction very clearly. Like a driverless car, everyone is doing a driverless car. No one has ever discussed whether this will happen.
In the course of such a design, because it is itself called digital assistant, it helps you to get things done. What everyone thinks is how to find what you need as soon as possible, as soon as possible to help you complete such work, in such a design process, the shorter the better the task is completed. Another idea, a very special place, I think there is not much people do is from the chat robot: I am your companion, I am with you, I do not help you solve any problems. From Microsoft Research, we are walking on two legs. This is definitely the most different place for us.
Just now Xin Zhiyuan's reporter also asked, Microsoft is not only emotional? No, we are walking on two legs. This is like the world knowledge, work knowledge, and user knowledge I just talked about. It can help you solve problems. How can the other one get your trust, how can you continue to be friends with you, how can you continue to talk to you, and continue to accumulate knowledge in the process. We use the product design metrics, and you can talk back and forth with your agent. For Cortana and Siri, Google Now is about three round trips, which is equivalent to a half-way conversation. You asked Siri, how is the weather in Seattle? It said it was still raining and you had no questions asked. But Xiao Bing is not the same. Xiao Bing is the highest number of conversations we have today. It is 23 rounds. You have the opportunity to communicate with it and learn in exchange. What I want to emphasize is not that we only do one or the other, that is, the assistant (service) and the emotion (company) we have to do on both sides.
The second is the 2B question that you talked about. I also heard a lot of people say that for a start-up company that provides commercial AI services, in the face of companies like Microsoft that can provide good AI services, there is no such thing for small companies. Some business opportunities continue to live? This issue is my own view: Today, providing such a so-called API, I also think that the general small company may be more difficult to do, because there is a lot of investment, and all aspects of the charge will also have to compete with many large companies.
One thing that I think is very important, the next business opportunity lies in the so-called business AI. Because it has to use AI for every vertical, to do something like this, you have to really go deep into a vertical industry, you must have very special data, you must have such a user, can use Deploy the application to AI so that you can give feedback. Most AI's today's technology cannot be 100% or even 90%. There are many errors in it. Such as Alexa, Siri, Cortana, speaking today, many times it did not really answer your question. But you need to have enough users, he is willing to continue to use your AI application, give you feedback, provide you with data. I am very optimistic about the vertical industry.
Cognitive computing will become a top priority for Azure; deep learning can draw on the rise of Android
Q: There are two major pieces in your speech. One is how AI helps developers to develop more customized application scenarios. The other says redefine. I would like to ask, what is the plan for Microsoft to redefine the product through artificial intelligence?
Dr. Shen Xiangyang: This issue is very good. The three major areas I talk about today are three aspects. The three major aspects of AI are, first, that you have very strong computing power; second, you have to have very great algorithms; third, you must have your own data.
I use Microsoft's example to introduce it. At Microsoft, we certainly believe that all of Microsoft's products must be AI-enabled and define such products again. We are now focusing on two aspects. One is on all Office products. In the keynote, I showed you one of the functions of PowerPoint, which is translation. In fact, they also did another PowerPoint. I liked it very much. The so-called image caption: To come up with a photo, PowerPoint shows that the production will automatically give you a picture, which we have done quite well.
PowerPoint is used by many people. This data can help us continuously improve some algorithms. Not long ago, we also released Word, which uses AI technology in Word. This is also very important.
There are still many AI technologies here that have only just begun. I think that the most exciting technology for Office is the so-called machine reading. Not long ago, Microsoft bought a Canadian start-up company called Maluuba, faded the family and coerced the tomb, and evenly transferred to the bath (5). The influence of ffffice will be enormous, so our AI department colleagues and Office colleagues are doing together.
The other one is in the cloud. We have a lot of cooperation. You go to the Azure.com homepage and the content of Cognitive Services is placed in the most prominent position. This is Scott and I, after reviewing with our product team colleagues, Cognitive Services will become the top priority of Azure. Windows also has a lot of AI, like HoloLens also has a lot of AI technology, computer vision, computer voice.
Lu Qi went to Baidu. Will you cooperate with them?
Q: You just talked about Microsoft's cooperation with some service providers and application providers in the vertical domain. In terms of platforms, such cooperation as Google, Amazon, and Baidu, how to solve social responsibility and corporate business interests? In addition, after Dr. Lu went to Baidu, does Baidu and Microsoft have any cooperation in artificial intelligence in China?
Dr. Shen Xiangyang: You have asked these questions very well. Microsoft is special. We have a Microsoft Research Institute. The Institute itself is very open. We publish many papers every year for all scientific research. Almost all companies have Can't compare with us. To date, over the past 20 years or so, in the best position in all computer fields, the best paper award winner and Microsoft Research Institute are far ahead of the world. Today, Microsoft still has such responsibility for society and continues to do the greatest computer research, including AI, and to do it in an open manner.
From the perspective of a commercial company, we still have two aspects. You look very accurate. First, we provide a platform and tools that everyone can use, including researchers, university teachers, and students. Our company also does the vertical industry, office AI will do better and better, such as the cloud application, such as database, we can all continue to do these things. It is normal for you to compete with other products as a business. There may be several different aspects of how people can cooperate between these big companies.
At the tool level, I give an example. For example, TensorsFlow, there are some new tools out there, like deep learning to such a point, everyone feels reasonable, it will naturally emerge some tools. At the same time as TensorFlow appeared, there were at least four or five tools, for example, Amazon recently received a company to do MXNet, and each has its own strengths. TensorFlow Google pushed earlier, push more fierce, a lot of users. It is very important that we, as a community, should bring together some strength.
I think we can learn a lot from Android's rise and how can we do better in the next wave of deep learning. I don't think that a Microsoft company can make all these things.
Domestic companies mentioned that we are very open to domestic companies. You just mentioned Dr. Lu. I mentioned that many other colleagues went to other companies. The relationship between everyone is very good. You may know that in November last year, I went back to Microsoft Asia Research Institute to set up a community meeting with Yaqin, I, and Hong Xiaowen. We organized other people to come together. Everyone was a good friend for so many years, even if everyone was not in the same company. Working together, I think the most important thing is that everyone is still in an industry. Everyone should help each other. I certainly hope that every colleague will do well at Microsoft and I hope that my former colleagues can continue to do well.
Q: Will Microsoft consider more and more Open Source tools?
Dr. Shen Xiangyang: Certainly, Microsoft's implementation is already very open source. You see that CNTK is now fully open source. Open source is a culture and there are different ways to do open source. I myself believe that open source has a tremendous and far-reaching impact on the development of science and technology in this world. Microsoft is growing very fast. We are completely embracing open source.
燑br>
Finish And Wood Working Staple
Wood Working Staple,T-Head Staple,Hardware Decoration Staple,18 Gauge Brad Nail
Zhejiang Best Nail Industrial Co., Ltd. , https://www.beststaple.com