Ai is the core driving force of the new round of industrial transformation, which will drive the transformation and upgrading of trillions of digital economy industries. The history of the three industrial revolutions shows that no matter mechanical technology, electric power technology and information technology, they can greatly promote the standardization, automation and modularization of production and have strong universality. Artificial intelligence technology also has similar characteristics and has great application potential. According to the New Generation OF ARTIFICIAL Intelligence Development Plan issued by The State Council, the scale of China’s artificial intelligence core industry will exceed 400 billion yuan by 2025, driving the scale of related industries to exceed 5 trillion yuan.
Artificial intelligence is the commanding heights of a new technology race, vital to economic growth and national security. In this a global competition, China’s strength is in ali platform type companies such as baidu, huawei, accumulated a solid technical foundation, rich application scenarios and huge amounts of data, under the new strategy for infrastructure, will create new competitive advantage, for national development into new growth momentum, is expected to become a leading force in artificial intelligence, new infrastructure. Of course, China still has shortcomings in basic scientific research, basic algorithms, core chips and high-end talents. The scientific and technological strength of a major country is the core of national strength, and whether China can seize the opportunities of the smart era is the key to building a modern great power.
Welcome the new era of intelligence
1.1 Artificial intelligence is “new electric energy” in the era of digital Economy
Ai is an important part of the fourth industrial Revolution and will drive the industrial transformation and upgrading of the digital economy. Since the 18th century, there have been three large-scale technological revolutions in human society, namely, the steam engine revolution, the electric power revolution and the information Internet revolution. Each technological revolution is accompanied by the development of relevant disciplines, and theoretical knowledge is improved in practical application. A virtuous cycle of “technological breakthrough – knowledge and discipline progress” forms, and becomes the support of subsequent technological development, and its influence on society will be enhanced accordingly. Artificial intelligence will be an important technology in the fourth technological revolution, thanks to the data accumulation in the Internet information age, the improvement of design and manufacturing process in the semiconductor industry and the improvement of chip computing capability, and the more accurate application of computer vision, voice technology and natural language processing technology brought by deep learning combined with reinforcement learning. As Andrew Ng, an international authority on artificial intelligence and machine learning, said, “ARTIFICIAL intelligence is a new electric energy that is changing major industries such as medical care, transportation, entertainment and manufacturing, enriching the lives of countless people.”
Artificial Intelligence (AI) has been in development for more than 60 years since it was first proposed at the Dartmouth Conference in 1956. It is generally believed that a computer needs to master many human skills such as drawing, singing, reading and designing through continuous self-learning and knowledge expansion, which is the manifestation of “intelligence”. In the White Paper on the Development of Artificial Intelligence (2018), The China Academy of Information and Communication mentioned that artificial intelligence can be understood as the use of machines to continuously perceive and simulate the thinking process of human beings, so that machines can reach or even surpass human intelligence, that is, artificial intelligence should have human-like perception, thinking and decision-making capabilities.
The basic, technical and application layers of ARTIFICIAL intelligence have developed rapidly, and many applications have penetrated into daily life. The basic layer includes hardware, algorithm and mass data. The core of the hardware is the chip with high computing capability, such as CPU, GPU, ASIC, FPGA, etc. The core of the algorithm is machine learning, including deep learning, shallow learning and reinforcement learning. The technology layer includes computer vision, speech, natural language processing and other technologies. The application layer is artificial intelligence products, services and solutions for home appliances, finance, robotics, automotive, medical and other fields. In the past 10 years, artificial intelligence has developed rapidly. Faced with the growing demand, some enterprises with long-term r&d experience, such as Baidu, Huawei and Ali, have launched ai development platforms or AI systems one after another, and are expected to become the leading force in the new AI infrastructure. Although with science fiction and film has a big gap about the concept of the artificial intelligence, artificial intelligence products and services has been widely exists in our real life, small to multilingual translation software, intelligent speakers, big to the autopilot system and city security system, the brain, the development of artificial intelligence has been far beyond the early vision, Governments, businesses and nonprofits are all embracing the technology.
1.2 From “+ AI” to “AI +”
Ai has been applied in many vertical fields, including home furnishing, finance, transportation, medical care and so on. By combining with many vertical fields, AI technology can empower industries in two ways: on the one hand, it can improve production efficiency, reduce cost and increase efficiency, i.e., “+ AI”; Second, we need to create new demand and growth points, namely “AI plus”.
1. “+ ARTIFICIAL intelligence” : fast and efficient data processing, taking into account ordinary and long-tail users at the same time, improve production efficiency, achieve cost reduction and efficiency increase. Take the financial industry as an example. At present, AI is mainly used in risk control, payment, claims settlement, investment and care, among which intelligent investment and care is the most mature application. Intelligent investment management was born in the United States in 2008. Due to professional quality and the nature of manual service, the investment threshold of major Financial institutions in the United States is relatively high, with an average investment threshold of about $50,000. The average management cost is 1.35% of the assets under management, and the service objects are mainly middle and high bourgeoisie. But as millennials grow and traditional customers become saturated, financial institutions are increasingly demanding ways to reach this long tail. Through massive data learning, precise algorithm analysis, artificial intelligence combined with the risk tolerance level provided by users, revenue targets, market dynamics, personalized customization services. Compared with manual services, intelligent investment service has a minimum investment threshold of $500 and a management fee of about 0.02%-1%. At present, mainstream domestic financial institutions such as China Merchants Bank and INDUSTRIAL and Commercial Bank of China have also launched intelligent investment products, while other institutions are also strengthening research and development of products and services with similar functions.
Ai has also played a huge role in COVID-19 prevention and control, mainly covering epidemic monitoring, temperature detection, virus detection, and resumption of work and production. The outbreak of COVID-19 during the Spring Festival has brought great challenges to virus detection, tracking, isolation and prevention and control. The application of ARTIFICIAL intelligence, supported by data, mainly helps temporal tracking and epidemic analysis. Take Baidu solution as an example: 1) Temperature monitoring and epidemic map to strengthen epidemic monitoring. Compared with the SARS period, the Spring Festival holiday of novel Coronavirus outbreak coincides with the period of national high speed movement, and the high infection characteristics of the virus increase the difficulty of manual screening in the early stage. The application of artificial intelligence computer vision, on the one hand, can meet the temperature monitoring in public areas such as airports and high-speed trains, and on the other hand, can meet the identification and record of suspected cases and people carrying the virus, so as to enhance the intensity and efficiency of epidemic investigation. 2) Online consultation and virus testing ease the pressure on medical services. China’s medical resources are insufficient and unevenly distributed, especially the multiple cases of cross-infection in hospital outpatient clinics caused by panic in the early stage of COVID-19. On the one hand, the development of online consultation tools reduces the risk of infection for medical staff, on the other hand, gathers medical resources, reduces the burden of clinicians, improves diagnostic efficiency and service quality, and makes up for the shortage of manpower. In addition, artificial intelligence has greatly improved the detection speed of novel viruses. Baidu has developed the linear time algorithm LinearFold, which reduces the detection speed of novel coronavirusRNA structure from 55 minutes to 27 seconds, increasing the speed by 120 times. 3) Telecommuting and online education help resume work and education. Baidu Rulu and Baidu Intelligent Cloud provide corporate communication, voice and video conferencing, collaborative office, online teaching and other services to speed up the resumption of office and teaching while ensuring the health of employees and students.
2. “AI +” : Create new demand, new business model and new economic growth point. Take automobile for example, intelligent network connection is the most concerned field of artificial intelligence application in automobile industry. On the one hand, intelligent network can improve the intelligence of the car, including automatic driving, intelligent voice, intelligent cockpit, etc. On the other hand, it can be combined with 5G to improve vehicle information communication ability and realize network connectivity, including personnel and vehicle safety management and urban road traffic planning. 1) Cars will become a gateway to all kinds of services and applications, spawnnew business models: The intelligent connected vehicle can be continuously updated with OTA over-the-air upgrades during its life cycle. The interface interaction will give the vehicle more application scenarios — in the case of driverless vehicles, the driver will have more free time, and the Internet of vehicles technology enables the vehicle to be connected to the office, home and public facilities at any time for remote control. Similar to the development of smart phone industry, with the maturity of intelligent connected vehicles, the importance and industrial value of data value-added (including shared travel, car insurance, financial services), entertainment and leisure, intelligent planning and other application links will surpass the simple car production and manufacturing links. 2) Increasing demand for automotive electronics and software: Automotive electronics and software are becoming more important to automobiles, and cutting-edge technologies such as autonomous driving, computing platform and on-board operating system have become new value growth points.
On April 19, 2020, Baidu Robotaxi launched Baidu Map and Dutaxi, an intelligent small program of Baidu APP, and fully opened its trial ride service to changsha citizens. This means that under the guidance of relevant laws and regulations, Baidu takes the lead in promoting Robotaxi into the normal test and trial ride stage in Xiangjiang New Area, Hunan province. At the end of the scene, ApolloRobotaxi covers an area of 130 square kilometers, and its driving routes cover the local residential areas, commercial and leisure areas, industrial parks and other multi-dimensional practical life scenes in Changsha. At the product end, the visual interface of Apollo Robotaxi can restore obstacles and dynamic prediction within the range of 360 degrees of vision, and present the road conditions of passing vehicles, lanes, intersection, traffic lights and other traffic conditions, accompanied by speed limit prompt and lane change reminder. Users can pay real-time attention to driving information such as speed and remaining mileage through the screen. Companies such as baidu in automatic driving, car road, collaborative, intelligent car platform technology research and development such as accumulation, is expected to further copied to smart credit control, intelligent public transport, smart parking, intelligent freight application scenarios, such as not only drive the sensor, chips, automatic car driving algorithm, smart cockpit, cloud service industry development, such as a trip and can improve the efficiency, reduce the cost, It is expected to become an important growth point for smart travel.
The battle for the commanding heights of ARTIFICIAL intelligence technology
Competition in the AI industry is a competition among countries in policies, basic research, technology, capital and other comprehensive strengths. At present, governments of various countries attach great importance to it and offer support and encouragement in infrastructure construction, basic scientific research, personnel training, funding for RESEARCH and development, cooperation and exchanges. Capital and enterprises are also actively seeking commercial landing scenarios to assist technology transformation. Technology lands in the vertical domain, which in turn generates new data, which facilitates iterative algorithms that serve the vertical domain further, and so on and so on. In this global race, China’s advantage lies in its vast amount of data and practical experience, but there are still weak links in basic scientific research, basic technologies and frontier expansion.
2.1 Policy: Major countries and regions in the world attach great importance to it
With the AlphaGo incident as the watershed, artificial intelligence has received unprecedented attention, and major countries and regions have joined in this competition concerning the scientific and technological strength of the future great powers. Due to the lack of universal infrastructure, advanced technology and numerous theoretical branches, the development of ARTIFICIAL intelligence has experienced three ups and downs. It was not until 2016 when DeepMind’s AlphaGo challenged the world’s top Go player Lee Sedol and won the final victory that the world felt the charm of artificial intelligence again. AlphaGo’s observation, thinking and decision-making abilities similar to or even better than human beings in the man-machine competition attracted countries and regions around the world to start and strengthen the research and development in the field of artificial intelligence. According to incomplete statistics, nearly 30 countries and regions, including the United States, China, the European Union, Japan, South Korea, India, Denmark and Russia, have released strategic plans and policy deployments related to ARTIFICIAL intelligence. Among them, about 80% of countries have intensively released relevant policies and official plans since 2016, such as the National ARTIFICIAL Intelligence Research and Development Strategic Plan of the United States, Robotics and Artificial Intelligence of the United Kingdom, and the Three-year Implementation Plan of “Internet Plus” ARTIFICIAL Intelligence Initiative in China.
According to the released policy plans, countries and regions recognize the importance of AI for future talent, industrial upgrading, social well-being and global influence, and promote it as a national strategy. The emphasis varies from country to country, depending on factors such as scientific research strength, talent pool, infrastructure and national conditions.
Ai is at the heart of the US’s technological landscape as it strives to maintain global dominance. From the Obama to Trump administrations, the US has actively supported AI research and shifted its policy attitude from “guide and support” to “must lead”. In 2019, the United States successively issued three important policies, namely, Maintaining America’s Leadership in ARTIFICIAL Intelligence, the National ARTIFICIAL Intelligence RESEARCH and Development Strategic Plan, and The Era of American ARTIFICIAL Intelligence: A Blueprint for Action, which show that the United States government attaches great importance to ARTIFICIAL intelligence technology and is determined to maintain its leading position. The main measures include: 1) Strengthen federal funding. The United States considers government funding to be an important part of participating in promoting scientific progress, but official funding has been declining. From 1976 to 2018, federal r&d spending as a share of GDP fell from about 1.2% to about 0.7%. Tax breaks to encourage companies to spend more on research and development; 2) Give full play to the innovation power of Silicon Valley, establish the technology and industrial ecological chain including computer vision, speech semantics, open source framework platform, etc.; 3) attaching importance to chip-based hardware, including promoting domestic semiconductor manufacturing industry, establishing multilateral export control, and protecting supply chains; 4) Attach importance to global talents, including the cultivation of domestic talents and the attraction of international talents, and consider it necessary to simplify the H-1B visa application procedures for relevant talents; 5) Strengthen cooperation, including organizing r&d centers or joint laboratories at home and abroad, holding innovation competitions, etc.; 6) Carry out cutting-edge technology research. The EU focuses on industry, manufacturing, healthcare, energy and other sectors, highlighting innovation and creativity and applying ARTIFICIAL intelligence to upgrade manufacturing and related sectors. Similar to the US, the EU started AI research and development earlier, and supported THE development of AI technology and industry by issuing policies, supporting funding, launching national programs and establishing key research laboratories, such as the declaration on COOPERATION in AI issued in 2018. Billions of euros will also be invested in ai-related projects as part of the digital Europe initiative and Horizon 2020. Compared with the United States, 1) of the European Union pay more attention to the moral and ethical research of artificial intelligence, and in many documents show that artificial intelligence development needs in line with the human ethics, for example, in March 2020, issued the excellence and trust – the eu regulation of artificial intelligence, new path clearly put forward, in order to solve the unequal and the information ability is not transparent, protect the rights of the people related, It is necessary to establish a regulatory framework of artificial supervision and attach importance to data security and privacy protection. 2) The EU focuses on the application of ARTIFICIAL intelligence in a more detailed way. Different from the US leading in all aspects, the EU hopes to make use of its advantages in manufacturing, industry, automobile and other fields to strengthen and upgrade the industry with ARTIFICIAL intelligence technology, such as the EU 2030 Autonomous Driving Strategy.
Due to the severe aging problem of fewer children, Japan focuses on the application of artificial intelligence in robotics, medical treatment, automobile transportation and other fields. Japan’s fertility rate has been low for a long time, and its aging level has been the highest in the world for a long time. The proportion of working-age population in Japan peaked in 1992, and the total population peaked in 2008, which has exerted profound negative impacts on Japan’s economic and social development, including challenges in old-age care and health. Released in 2016, “Japan’s next generation of artificial intelligence to promote strategy” as the starting point, Japan has introduced relevant policy planning, around the “industrialization of basic research and application research -” three aspects, including the Japanese ministry of information and communication technology institute and the science of artificial intelligence theory and technology research and development, the application scenario for the ministry to solve the problem, The ARTIFICIAL Intelligence Research Center (AIRC) established by METI promotes industry-university-research cooperation and mainly undertakes the transformation and promotion of achievements.
China’s AI is advancing gradually in three stages, focusing on integration with manufacturing and services. Since 2015, China’s AI-related policies have evolved from the intelligent manufacturing period, the “Internet plus” period (represented by the “Internet Plus” AI Three-year Action Plan) to the “Intelligence Plus” national strategy period (represented by the “New Generation AI Development Plan”). The policy focus has shifted from core technologies to practical scenarios, from specific industries to cross-border integration, from single technologies to human-machine collaboration. Similar to the United States and the European Union, China also emphasizes the establishment of relevant pilot projects, including technology demonstration pilot, policy experiment, social experiment.
2.2 Basic scientific research: America is the strongest, and China is catching up fast
The number of papers in the field of artificial intelligence in China is growing fast, but there is still a gap in the quality of papers compared with the United States. In 2018, China and the US published 25,000 papers and 16,000 papers respectively, accounting for 46.5% of the world’s total. In terms of growth trend, the United States has maintained a constant growth rate, while China has grown rapidly since 2014. The proportion of Chinese papers in the global total has increased from 8.9% in 1998 to 28.2% in 2018. The FWCI index (Average Weighted Citation Impact Index), which represents the quality of papers, shows that the quality of Chinese papers has also been steadily improving, from 0.43 in 1998 to 1.39 in 2018. The US maintains the highest level in the world, at around 2 for years. In 2018, the FWCI reached 2.38.
FWCI index: FWCI is normalized to 1. When the FWCI index of a country or institution is 1, it indicates that the country or institution has the world average citing influence. A FWCI index of 1.2 indicates that a country or institution’s papers are cited 20% more than the world average. A FWCI index of 0.8 indicates that a country or institution’s papers are cited 20% less than the world average.
In terms of publication institutions, universities are the core scientific research forces in China, the United States and the 27 EU countries. In 2018, the paper output of the three universities accounted for 92.1%, 84.6% and 90.7% of their total output. Apart from universities, China and the US have different main research bodies. In 2018, the output of Chinese research institutions was about three times that of Chinese enterprises, while the output of American enterprises was about 1.6 times that of American institutions.
2.3 Data volume: The “raw material” in the era of ARTIFICIAL intelligence, China has the advantage of scale
The popularity of computers and smart phones, as well as the explosion of data accumulated by the Internet and mobile Internet, is one of the important reasons promoting breakthroughs in AI technology and applications. Artificial intelligence needs to “feel, think, and make decisions”. First of all, it needs enough raw data, good enough, to train computers like fresh grass. “Enough” means a large amount of data. The invention of computers, which simplified calculations and allowed information to be stored electronically, and the proliferation of smartphones, which dramatically increased the penetration of Internet users around the world, have both allowed vast amounts of data to be stored. “Good enough” on behalf of the quality of the data to better, the birth of the Internet greatly shorten the physical distance, improve the transmission speed of communication, all kinds of the birth of the Internet application service, its data type is also more diversified, including web browsing preferences, outside the selling point of single frequency, travel records, multiple rich data to deal with all kinds of training requirements of artificial intelligence.
Data growth and utilization depend on information and physical infrastructure construction, China will become the world’s largest data center. Thanks to population, Internet penetration, smartphone penetration and Internet speed, China had 7.6ZB of data in 2018, accounting for 23.4 percent of the global total. With 5 g, the Internet of things, such as development, communications equipment access number and improve the bearing capacity, the terminal consumer increases, the amount of data that will be up to 48.6 in 2025 in China the ZB, accounted for 27.8% of total global data, become the world’s largest data centrally, will greatly promote and rich artificial intelligence training, relevant model structure and the result is more accurate.
2.4 Technology: Deep learning is driving the ai boom
There is enough data, good enough data, to support the “perception” stage of AI, and AI algorithms give computers the mind to “understand and make decisions,” a process to which deep learning makes a huge contribution. Deep learning is a general term for a kind of pattern analysis methods. Computers master internal logic and rules by learning sample data, so as to acquire analytical ability. This research can be traced back to the Perceptron invented by Frank Rosenblat in 1958. Using the perceptron, image-discrimination training can be carried out, such as selecting “apple” or “banana” from a pile of fruits. However, due to the lack of sufficient data at that time, the research was stuck in a bottleneck and Overfitting occurred. For example, students hoped to master a question type by practicing similar questions, but the amount of training was not enough and they did not understand the knowledge points behind the question type, so they could not solve the answer once there were some changes in the exam. Later, by studying the human brain, scientists tried to imitate the mechanism of human brain neural network to classify images and sounds, which gradually evolved into today’s deep learning.
The development of deep learning has promoted breakthroughs in basic application technologies of artificial intelligence. Since 2010, the number of patent applications of basic application technologies including computer vision and speech semantics has increased rapidly worldwide.
Computer vision technology is mainly to let the computer have human eyes, learn to “see” pictures, text, video and so on, often used for image recognition, face recognition and so on, suitable for automatic driving, security, face payment and other fields. In terms of applications for computer vision and image recognition technologies, as of December 31, 2018, a total of 143,000 patents of the same family were filed worldwide, with China, the United States and South Korea becoming the top three countries in the number of applications, with 53,000, 24,000 and 23,000 respectively. In terms of technology licenses, the United States has the highest number of technology licenses in the world with 13,000, followed by Japan and China with 10,400 and 10,000 respectively. In terms of applicants, Canon, Toshiba and Samsung were the top three with 2,900 applications, 2,700 and 2,300, respectively.
Speech semantic technology is mainly to let the computer learn to “listen and read” text, paragraph, article, etc., often used in text recognition, speech emotion analysis, human-machine dialogue, voice positioning, etc., suitable for translation software, vehicle operating system, intelligent speaker, voice assistant and other fields. As of December 20, 2019, a total of 43,000 patent families were filed worldwide. China and the United States remain the leading patent applicants in this field, accounting for more than 75% of the total. From the perspective of applicants, most of the applicants in the field of speech semantics are enterprises, among which IBM, Samsung and Microsoft are the top three applicants, with 1,741, 890 and 821 patents applied respectively. From the perspective of patent licensors, Microsoft, IBM and Nuance are the top three, with 672, 468 and 440 patents respectively. From the perspective of domestic enterprises, Baidu has become the only one in the world’s top 10 enterprises in both applications and authorization of voice semantic technology.
The number of patent applications in the field of artificial intelligence in China is on the rise year by year. According to data from the National Industrial Information Security Development Research Center, the number of domestic patent applications in 2018 reached 94,539, 10 times the number of applications in 2010. As of October 2019, Baidu, Tencent, Microsoft, Inspur and Huawei ranked the top five in the number of domestic AI patent applications with 5,712, 4,115, 3,978, 3,755 and 3,656 patent applications respectively.
The emergence of artificial intelligence chips significantly improves the speed of data processing and supports increasingly complex algorithms to deal with complex data, which is an important basis for the development of artificial intelligence. With the increase of the amount of data to be processed, from general scenarios to various specific scenarios, the framework and layers of algorithm model design become more and more complex, which puts forward higher computing requirements for basic hardware. In terms of patent applications, China and the United States are big applicants. As of October 2019, China and the United States had applied 16,000 and 11,000 patents for AI chips, respectively. From the perspective of relevant applicants, traditional chip and semiconductor enterprises have more advantages, among which Samsung, Hitachi and IBM are the top three patent applicants in this field. From the application trend in recent years, Samsung and Intel are more active. From the point of view of practical application products, the current representative products include Intel EyeQ series, Nvidia Xavier series, Huawei Ceng 310, Cambrian Cambricon 1M series, Baidu Kunlun chip, etc.
China and the United States are the clusters of ai enterprises in the world. Chinese enterprises focus on the application layer, while American enterprises focus on the technology layer. As of February 2019, there were 3,438 AI companies in the world, with the United States ranking first with 1,446, accounting for 42.1% of the global total, and China second with 745, accounting for 21.7% of the global total. In terms of enterprise types, China is mainly application-layer enterprise, while the United States is mainly technology-layer enterprise. China has the highest proportion of ai companies at the application level, accounting for 75.2 percent. Technology ranks second, accounting for 22%; The ratio of basic companies is at least 2.8 percent. In the U.S., companies of the three categories take up 39.1%, 57.7% and 3.2%, respectively.
2.5 Capital: Global investment continues to rise, and Chinese and American AI enterprises are the most favored by capital
Breakthroughs in AI technology and policy support have attracted continued capital investment, with an average annual growth rate of about 50 percent over the past decade. According to Stanford University, global investment in AI startups rose from less than $1 billion in 2009 to nearly $40 billion in 2019, with investment accelerating in 2014. From 2014 to November 2019, there were 16,000 investments in AI startups globally, with an average investment value of about $8.6 million each.
By country and region, American and Chinese companies are the focus of global investment. Due to its technological leadership, the United States remains the world’s first in both the amount of investment and the number of invested enterprises. Although the number of invested enterprises in China is not as high as that in the US, due to the high amount of each investment, such as The SERIES C financing of $460 million for Megapolis in March 2018 and the SERIES C financing of $620 million for Sensetime in April 2018, the amount of investment in Chinese start-ups is only the second in the US, about $25 billion. In addition, the UK, Israel, Canada, France, Japan, Singapore, Germany and India are the most frequently followed countries and regions.
3 Challenges and Suggestions
In the digital economy, 5G is like an “information superhighway”, providing high-speed transmission channels for the transmission of huge amounts of data and information, and completing the shortcomings that restrict artificial intelligence, big data and industrial Internet in information transmission, connection scale and communication quality. Artificial intelligence, like a cloud brain, relies on the learning and evolution of information from the “highway” to complete the process of machine intelligence. Industrial Internet is like a “bridge”, relying on the “highway” to connect people, machines and things, and promote manufacturing to intelligent manufacturing. Ai has obvious spillover effects and will, together with 5G and data centers, promote industrial transformation and upgrading in the era of digital economy. It is the commanding point of the current and future science and technology competition among countries. The scientific and technological strength of a major country is the core of national strength, and whether China can seize the opportunities of the smart era is the key to building a modern great power. In general, China’s AI industry is still in the early stage of development, facing problems such as lack of basic research and development, the integration of technologies and scenarios, and the inability of traditional infrastructure to keep up with technological development. Advice:
1) To provide “soft” support for the development of ARTIFICIAL intelligence, and do a good job in personnel training, cutting-edge technology research and liaison and cooperation. Strengthen domestic universities to develop relevant courses and cultivate local talents. We will actively attract overseas researchers and talents from around the world. Against U.S. attract measures for scientific research personnel, China should seize this opportunity, in research funding, personal tax, visa, registered permanent residence, children education launched a series of fields such as the introduction of overseas high-level talents package of policies, the scientific research personnel should be resolved trouble back at home, and to provide the scientific research, business providing greater support. We will accelerate the reform of science and education systems and establish a market-based, multi-tiered system of industry-university-research collaboration. The state should lead to increase the investment in basic research, and enterprises should lead to increase the investment in experimental development, so as to form a reasonable division of scientific research among various subjects.
2) To ensure the “hard” development of ARTIFICIAL intelligence, speed up the construction of information infrastructure, and carry out intelligent upgrading of traditional physical infrastructure. Form the industrial age with railway, highway, airport infrastructure, cloud computing, big data, such as artificial intelligence, 5 g, block chain will be the future focus, covered by the new infrastructure includes two aspects, category is the digital center, such as base station information equipment, another kind is roads, railways and other traditional infrastructure equipment. In order to deal with future digital challenges, we need to start from these two aspects. On the one hand, we need to accelerate the construction of broadband network and 5G network, and on the other hand, we need to strengthen the intelligent equipment of sensors, control platforms, cloud platforms and other public scenes such as traditional railways and airports. Hardware foundation for data collection, transmission, communication and analysis for subsequent technical development.
3) Attach importance to the ethical issues brought by artificial intelligence technology, and keep up with technological innovation from the perspectives of legislation and supervision. The development of artificial intelligence cannot be separated from data. Since most of the data are open, transparent and freely circulating virtual products, rights and responsibilities arising from the ownership of the data will arise, which also involves data security, intellectual property protection and privacy issues. For example, an enterprise can analyze consumers’ preferences and preferences through their online browsing information and make precise push. When the enterprise reduces marketing costs, consumers can better obtain information or products. However, it is also worth considering whether this behavior is consenting to by consumers and involves infringement of personal privacy. Since the production and use of data involves consumers, platforms, operators, service providers and other links, data is processed and integrated in each link, and it is difficult to use traditional product standards for unified management, which also hinders relevant legislation and supervision. Therefore, it is necessary to pay attention to ai ethics, technical standards, the relationship between AI and human society and other issues, people-oriented, attach importance to data security.