Author Yi Xiao Yi Not long ago, the 43 year old Microsoft took a solid step in the transformation, ending the dominance of Windows and embracing cloud and AI. Meanwhile, Microsoft Asia Research Institute will soon usher in its 20 birthday. Hong Xiaowen said it will open up AI technology and talents, and open cooperation mode. For Microsoft, what opportunities does architecture change bring? What are the new business plans for the Asian Academy? And, what are the frontiers in Microsofts recent data privacy, network security and even ethical issues? Come up and listen to Hong Xiaowens latest ideas. Talk about restructuring: Microsoft is a platform company, will seize the opportunity of digital transformation service. NetEase Intelligence: Microsoft recently announced the restructuring of its bet cloud and AI, and what opportunities are there to commercialize in the future? Hong Xiaowen: the new division of the Microsoft restructuring is called Cloud+AI, which is my favorite ABC model. A represents AI, B represents large data, C represents cloud computing, and the three are indispensable. Microsoft has been working as a platform company, and now all the actions are digitized, and everyone sees a huge opportunity for digital transformation, A+B+C, to help every company, everyone. I think there are opportunities for Microsoft in the four directions of digital transformation. The first is to improve the product and service. ABC can promote the product iteration, before we call Internet plus, now called AI+, are in fact the digital transition. The second is to close the distance between the company and the customer. Each company has its own customers, ABC can help the company to close to the customer distance, without waiting for customer complaints, can provide positive services to customers, make customers your fans, become part of the market propaganda. Third is to improve the operating efficiency of the company, each company can use ABC method to make the company more efficient, it is clear that the people will not say much. The fourth is to help employees get self-improvement so that they can complement each other in their work, life and family and inspire their potential and creativity. In this respect, Microsoft has many productivity tools. In short, Microsoft is a platform company. We need to provide ammunition behind the digital transformation of every company, so that they can dig gold in the era of AI. Our internal data show that Microsoft earns $16.45 for every $1 earned by Microsoft. Talk about planning: Microsoft Asia Research Institute will open AI technology and talents. NetEase Intelligence: Based on the above adjustment of Microsoft, what kind of business plan does Microsoft Asia Institute have this year? Hong Xiaowen: this year is Chinas reform and opening up 40th anniversary, the Microsoft Asia Research Institute was set up 20th anniversary, 20 years ago, the Microsoft Asia Research Institute was set up when the goal is to do the related research of artificial intelligence, we were born at the time. At present, the technology competition in the field of artificial intelligence, everyone is more than four fields, voice recognition, computer vision, Machine Translation, machine reading comprehension. In these four competitions, Microsoft took the lead in surpassing the level of human contrast, of which three were led by the Microsoft Asian Institute, and the other we were involved. So, first of all, we need to remain creative in scientific research. Secondly, the speed of technological progress is quickening, and the speed of technology and product landing has accelerated. The new plan of the Microsoft Asia Research Institute this year is to open up technology and talent and cooperate with domestic companies to help them to make digital transformation. We will launch DTaaS services recently to provide consulting and technical support for domestic companies to do digital transformation. The Microsoft Asia Research Institute set up innovation sinks in November last year. This is a platform for converging innovation, experience and technology. We hope to establish a wider and deeper cooperation link with large enterprises, investment institutions and government departments in the Chinese market, build a cross line communication platform and achieve win-win results. At the moment, it can be revealed that the Microsoft Asia Research Institute has reached strategic cooperation with companies such as the China fund, Pearson group and OOCL, developing innovative products and industry solutions, which will become a very new probe into the win-win model. NetEase Intelligence: how does Microsoft Asia Research Institute choose AI research direction? What are the technical directions that will focus on the future? Hong Xiaowen: we will strive for every AI technology, such as vision, semantics, translation, QA and so on. There are still many things to do. Talk about the market: the AI industry needs to cool down and precipitate the AI in the manufacturing, medical and agricultural fields. NetEase Intelligence: Recently, there is a saying that the heat of AI will slowly cool down in the coming year. What do you think of the industry situation in the next one or two years? Hong Xiaowen: I agree with that. History has proved that things that are too much to fry will cool down. AI is sure to be overheated now, so there is bound to be cooling time. The process of cooling is like gold rush and precipitation. After precipitation, there will be something of high value. So I think the proper cooling in the AI field is not necessarily a bad thing. For the landing application of AI, first of all, I am very optimistic about ToB business, because every company needs such services in digital transformation. In addition, if you ask me what kind of application can be done in the next one or two years, Im hard to say, but I think manufacturing, medical industry, agriculture especially need artificial intelligence technology. I also look at the application of AI in these fields. Worrying: data management must be strengthened, and more companies are expected to join the digital Geneva contract. NetEase Intelligence: the development of AI has also made people worry about data privacy, network security and even moral problems. What do you think should be done? Hong Xiaowen: the malicious use of AI technology will bring some worries to people, but the technology itself is innocent. The first thing to worry about is the disclosure of data privacy in the AI era. I think the management of data needs to be strengthened. When you have relevant data, there are some things that you cant do. Just like the staff of the securities regulatory department can not share the stock, and can not disclose the news, this is a good example. Specifically, most Internet applications are free at present. Neither WeChat nor Facebook charges consumers. What do they earn? Its actually advertising. The commercial mode of advertising is to let advertising companies find the target audience. There are two kinds of advertisements: one is to sell you products, the other is to make brand effects. So, my view is that if consumers do not want Internet Co to use their own data to push ads, then pay a penny, which is a possible future direction. From the government level, it may be necessary to manage the behavior of network companies by legislation. European GDPR is a good example. It first proposes the right of forgetting. That is to say, if one day users do not want to use the service of this company, they can apply for data to be completely deleted. However, the foundation of this rule is to establish trust between the company and the company, between the company and the users. Second, on the issue of network security, many people think that data is vulnerable to network attacks on the cloud, but the cloud computing is more secure and the public cloud is safer than the private cloud. Its as if you have a lot of money and put it in a big bank, which is safer than staying at home, because banks will spend more energy and resources to ensure safety. At present, most companies in the United States and even the government use public cloud, and in the field of public cloud, Microsofts trust ranks first. Third, on the question of AI ethics, last year, Microsoft President BradSmith proposed the digital Geneva contract, inspired by the Geneva contract in 1949, which suggested that even in the war era, the lives and property of the civilians were also guaranteed, and the contract was signed by many countries later. The digital Geneva contract is the desire of Companies in all countries to sign a contract, that is, in the era of peace, each company must be neutral, not to help any country attack civilians, and to help users prevent network attacks by 100%. In addition, Microsoft also established the AI ethics committee, and published a book on the importance of data management, ways to prevent data discrimination, how AI poverty, the relationship between AI and morality, and the impact of AI on work. For the topic of AI replacement work, I think the best solution is education; for the question of AI morality, we can follow the Hippocratic oath of the medical community to regulate the social responsibility and professional ethics of the AI practitioners. Talk about consciousness: at present all intelligence is AI+HI, it is very difficult to make conscious AI. NetEase Intelligence: do we want a conscious AI robot? Can we make this AI system in the future? Hong Xiaowen: I personally feel that it is very difficult to make conscious AI because there are very few conscious creatures. Once a scientist has done a MirrorTest, it makes many animals look at mirrors and see if they know that the creatures in the mirror are themselves, and that most animals are not passed, including dogs and cats, and only a few animals, such as dolphins, orangutans, and trained monkeys, have been passed. If this test is conducted, only a small number of people will be conscious. Consciousness is a very special thing. Some people even think that something with a soul is conscious. Therefore, machines will not be conscious, and machines may only simulate consciousness. What we can control is absolutely not consciousness. Now that all the intelligence is AI+HI (artificial intelligence + human wisdom), I cant say that in the future we cant do absolutely no conscious AI. I can only say its hard to say that no clues can prove that we can make a conscious AI, and it doesnt make much sense from the Perspective of the application. (finish) Note: exclusive interview with AI hero belongs to NetEase intelligent studio. It is an insight into technology trends with industry people to capture industry opportunities. Reprint please indicate the source. Click to see the topic of AI hero.