In the later stages of life, Stephen Hawking, who is still keen to talk about AI and is deeply feared, is now sleeping with the masters of human development in Newton and Darwin, at Westminster Abbey on the Thames in London. Of course, people will no longer be able to hear any opinions from Stephen Hawking on the AI.
However, is there any object that is more exciting than AI or can lead humans to develop to a new civilization? Stephen Hawking did not leave an answer.
Concerns about AI are not what we are considering now, nor are we able to consider it. The level of human AI is likely to be in the “initial era” for a long time. During this period, we are more concerned about how to use AI technology to improve efficiency and reduce costs in areas such as life, city, transportation, finance, production, commerce, medical care, and security.
AI to China
At present, most of the countries in the world’s major economies have developed AI as a national strategy.
The United States continuously launched AI. In May 2016, the “Artificial Intelligence and Machine Learning Committee” was established to coordinate the actions of all walks of life in the field of artificial intelligence, and to develop policies and laws related to artificial intelligence. In October 2016, it continued to publish The Smart Future is Ready and the National Artificial Intelligence Research and Development Strategic Plan are two reports that raise artificial intelligence to the national strategic level.
Europe also identified AI as a priority development project. In June 2016, the European Commission proposed an artificial intelligence legislative motion; in April 2018, the European Commission submitted “European Artificial Intelligence”; in December 2018, the European Commission and its member states issued The Artificial Intelligence Coordination Plan with the theme of “Artificial Intelligence Made in Europe”.
Based on its global leading position in the field of intelligent robot research, Japan actively promotes the development of AI. In the “Social 5.0” strategy proposed in 2016, artificial intelligence is the core of the ultra-smart society, and the “Artificial Intelligence Strategy Conference” is established for the country. Integrated management at the level.
It is worth noting that the support of countries for the development of AI is not only reflected in the AI strategy itself, but also in laws, regulations, strategies and policies in other fields such as economy, society and industry.
The best news for AI in China is that in addition to the relevant national strategic guidelines and local policies, China’s development of AI has a natural advantage in terms of market and application basis. In February 2019, Microsoft and IT industry market research organization IDC released a research report in the Asia-Pacific region, “Watch the Future Opportunities: Artificial Intelligence Inspires Asia-Pacific Growth Potential”, showing that 7% of enterprises in the Chinese market have incorporated artificial intelligence into the core of the company’s future development. Strategy, 31% of companies are trying to integrate artificial intelligence into corporate development strategies.
In this regard, Ralph Haupter, president of Microsoft Asia Pacific, believes that “for today, data is a key indicator of AI quality, and China has more data than the United States.”
Li Yanhong confirmed the views of He Lefu. At the 4th World Internet Conference, Li Yanhong said: “This large group of people, such a large market, such a large data set (referring to the Chinese market), and constantly generating new data, it will further promote The constant innovation of the algorithm and the requirement for updating the computing power. The main driving force for the development of China’s digital economy in the future will be artificial intelligence, and artificial intelligence will move forward at a very fast speed.”
However, what is the direction of China’s AI development? Is there any relevant basis for action?
A book called “China Artificial Intelligence 2.0 Development Strategy Research” was published in early 2019. The content of this book focuses on the wisdom of the top elites in the domestic AI field. The project consultants are Academician Zhou Ji, Dean of the Chinese Academy of Engineering, and Academician Chen Zuoning, Vice President of the Chinese Academy of Engineering. The project leader is the former Executive Vice President of the Chinese Academy of Engineering, Pan Yunhe. The project leaders of the project have 17 academicians including Chen Chun, Li Wei, Gao Wen, Zheng Nanning, Wu Cheng, Li Bohu, etc. More than 200 students from universities, research institutes, enterprises and governments participated in various research projects. experts and scholars. These academicians and experts are the authoritative scholars of China’s new generation of artificial intelligence development research. They are the drafting team of the State Council’s “New Generation Artificial Intelligence Development Plan” and the drafting team of the “Technology Innovation 2030 – ‘New Generation Artificial Intelligence Major Project” Declaration Guide. They fully reflect China’s highest level of research on a new generation of artificial intelligence.
It is worth noting that Academician Pan Yunhe was the president of Zhejiang University. During his tenure, he led the founding of Zhejiang University, a listed company that mainly engaged in information technology. After the technology strategy route of Computer+, Internet+, Big Data+, etc. Today is the leading company in the domestic AI+ field. Later, Academician Chen Chun, a fellow of Pan Yunhe, took over as the new chairman of Zhejiang University.
Academician Pan Yunhe and Academician Chen Chun, as well as the current president of Zhejiang University, Wu Zhaohui, are both professors of He Zhijun, one of the founders of artificial intelligence in China. Professor He Zhijun is the founder of the Department of Radio and Computer Science of Zhejiang University.
According to the public information, Shi Li, the current chairman of Zhejiang University, is also under the command of Professor He Zhijun. He is also the director of the Computer Application Engineering Center of the School of Computer Science and Technology of Zhejiang University. It can be seen that as a listed company of the university department, Zhejiang University New Network and Tsinghua Tongfang and Peking University Founder have promoted the incubation process of technology and products from the laboratory to the enterprise through the integration model of production, education and research. Of course, all university companies have their own strategic positioning options. Obviously, Zhejiang University’s new development is a road that is rooted in the early AI field and then goes to today’s AI field.
Frankly speaking, there are a lot of self-proclaimed AI companies, but some of them may be pseudo-AI or self-hype. Now, companies that advertise artificial intelligence will attract a lot of attention.
In Europe, for example, London-based investment firm MMC Ventures studied 2,830 European companies that claimed to use AI, looked at each company’s data, products, websites and product files and found that 40% of them did not use any machine learning technology. More than two thousand companies cover start-ups in 13 EU countries, but did not disclose the list of companies participating in the study.
Does China have a situation like Europe? As there is no relevant report supporting evidence, it is still impossible to understand the situation in the Chinese market. Just as some of the hot spots just came out, there will be surface prosperity at the time, but the actual situation will be mixed. However, after the ebb tide, the nude swimmer will also show his body.
AI itself is a cutting-edge technology, but its real value lies in the combination of technology and industry, which is called AI+. This kind of thinking, like the previous Internet+, must serve the industrial economy through business models and business models.
However, it is necessary to be clear about the fact that it is easy to mention the Internet + and it is difficult to propose and implement AI+.
AI+ companies must have two capabilities at the same time: first, true AI and peripheral technical capabilities; second, their determined AI+, can find a practical direction of action. It seems that the new AI+ advantage of Zhejiang University with the integration of production, education and research is more obvious.
In the past Internet + period, using its own technology, “Smart City” as one of the company’s core business, Zhejiang University New Network, long-term with Huawei, ZTE, Ali and other companies, in China “Smart City Solution Providers Top 100” In the top position. Therefore, for some of the pain points in China’s “smart city”, there is a summary of its own experience, and in the era of more rapid AI, the new high-level of Zhejiang University hopes to find a more convenient way to solve the pain point.
According to the calculation of the new chief scientist Mao De Cao, “China has 160 million urban road cameras. Taking a city in East China as an example, only smart transportation projects involve more than 500,000 cameras, using traditional centralized AI processing solutions, according to 1.5. Mega code requires 750G of third-party bandwidth transmission. The transmission cost is only 100 million yuan per year, and the AI calculation cost is several hundred million yuan per year. However, if the distributed AI processing mode is adopted, the video of the camera is first extracted into information, and then The transmission of information to the cloud computing center can greatly increase the coverage of cloud computing and the efficiency of big data calculations, and can save the overall cost by one tenth.”
The distributed AI processing mode is the strategy chosen by Zhejiang University in the AI+ era. Zhejiang University New Network named this platform: IN-Edge distributed AI system.
Chairman Shi Li believes that the application of video surveillance is an important starting point for Zhejiang University to find the industrialization of AI. To this end, Shi Lie raised a topic: “In the past 10 years, China’s smart city construction has put a huge number of cameras on every corner of the city. The current generation of smart brain applications such as urban brains and snow projects are also largely It is based on the information collection of video surveillance images. However, the method of processing video information centrally requires a large amount of network transmission resources and GPU computing resources, which is difficult for general urban customers to bear. However, in the era of cloud-based intelligence, How to jump out of the framework of project integration thinking and re-architect smart city business with cloud and platform thinking?”
“Only when the cost is greatly reduced, the urban brain has the market space promoted throughout the country. So we have seen the opportunity of upgrading the smart city, and the new smart city model has transformed from project integration to distributed cloud computing. It is imperative,” Shi Li said.
Efficiency and cost control
Looking for the city camera means that Zhejiang University will compete with Hikvision, a security star listed company in Hangzhou, Zhejiang. So, do the two companies really form a short-term match?
In this regard, the head of Huatong Cloud Data Technology Co., Ltd. (hereinafter referred to as “Huatong Cloud Data”) of Zhejiang University’s new network said: “There is competition, but not directly. In a sense, the two companies can also work together. ”
Huatong Cloud Data is responsible for the transformation of Zhejiang University’s new “Smart City” business from Internet + mode to AI+ mode. It is also the core subsidiary of Zhejiang University’s new development and implementation of IN-Edge distributed AI system operation.
According to the public information, Huatong Cloud Data was established in July 2012. It is mainly engaged in the business development of “cloud data center, cloud computing, cloud application, cloud network” and is a close strategic partner of Alibaba Cloud. The company has relatively complete cloud computing infrastructure resources – 60,000 square meters, nearly 10,000 cabinets of green data centers, more than 100,000 virtual host computing power, covering the end-to-end network solution capabilities of major cities across the country, is There are few comprehensive service enterprises with relatively complete cloud computing resources in the country, and on this basis, three major types of main business and customer resources are formed:
First, the data center hosting business (IDC business). It mainly provides data center hosting services to Alibaba Cloud, etc. It is an important data center partner of Alibaba Cloud. At the same time, the company is the only data center hosting service provider of the Hangzhou Municipal Government and has become one of the main service providers of the data center custody business of the Zhejiang Provincial Government;
Second, the government-owned cloud service business. It is the first provider of government-owned cloud services in the country. It is the provider of government-owned private cloud services of the Zhejiang Provincial Government, the Hangzhou Municipal Government, and the Lishui Municipal Government. It provides a proprietary cloud computing platform for government informationization and smart city systems, providing computing and storage. , database and other comprehensive services;
Third, the Internet resource service business. Interoperability with more than 100 major websites (ICP) such as “Taobao, Tmall, WeChat, Baidu”, and interconnection with major cloud computing platforms such as “Alibaba Cloud, Tencent Cloud, Baidu Cloud, Jinshan Cloud” Achieve interconnection and convergence with millions of websites built on cloud computing platforms to build the nation’s largest Internet resource center.
In 2017, Huatong Cloud Data was acquired by Zhejiang University’s new capital, complementing Zhejiang University’s new advantages, and taking advantage of Zhejiang University’s new advantages, from cloud computing infrastructure service providers to professional cloud service providers and AI cloud. The service provider of the computing resource.
However, the same focus on the “city camera”, Zhejiang University and Hikvision, but they are very different.
Hikvision’s AI is mainly deployed on the camera’s terminal hardware, such as its product terminal 3D face camera, which has 3D face modeling capability, which can prevent photo, video and mask spoofing attacks and enhance local face and side face. The comparison, as well as “smart” series cameras, “fully structured” cameras, “AR” series products, “black light” series products, “environmental” traffic series products, “73 times super wide-angle star ball” and so on.
It is undeniable that in the field of photography, Hikvision’s AI technology has certain advanced capabilities, but the problem is that because its AI technology is bundled on the camera, it means that if you want to improve the AI capability of the city camera, replace it. For Hikvision’s AI-type camera, it is necessary to eliminate the country’s 160 million stock cameras. In this regard, all localities need to measure their financial capacity, on the one hand, they have to face huge cost pressures.
According to Mao Decao, the “IN-Edge Distributed AI” processing mode is actually a post-AI system, which is to set up a new set of Zhejiang University in the middle of Taiwan without replacing the existing 160 million stock cameras in the country. The distributed AI system developed by the company, through the coordinated distribution mechanism of data and tasks, achieves the same effect, on the one hand, it also has AI capability, on the one hand, it can save the overall cost to 1/10 of the original cost – this is the new network of Zhejiang University Conway also has different strategies and solutions for “city cameras”.
Quite counted, even in the initial upgrade of the strategy from the original Computer+, to the Internet +, and then the big data +, the new management of Zhejiang University has been low-key operating companies, there are few strategic conferences, but this May Senior executives such as Chairman Shi Lie and Chief Scientist Mao De Cao attended the strategy conference and held a grand ceremony to announce the company’s strategic adjustment to AI-Driven and the IN-Edge system. In the subsequent media communication, Shilie said that the IN-Edge system is not only for the “city camera”, but has already formed the actual operation of AI + media analysis, and then there will be more “AI+”. The form appears.
In Shi Lie and Mao Decao’s view, AI should not be high. AI not only needs to enhance the wisdom of the city, but also enhance the wisdom of life, business, etc., and must break through the high cost of use, and finally make AI easy to reach.