Wen/You Should Be Good
Source: Everyone is a product manager (ID: woshipm)
Nowadays, China is the fastest developing and the most promising country in the field of artificial intelligence in the world, and some of the indicators have already occupied the leading position in the world. However, there are still some vertical subdivisions of artificial intelligence in China that need to be developed urgently.
I. Purpose of investigation
This paper takes AI as the main research object, and deeply investigates the macro and micro background of the industry through industry background analysis (macro-economy, competition situation, industry technology forecast), market status analysis (market stage, market scale) and micro-analysis (application scenario, SWOT, competition strategy).
The purpose of this paper is to deeply analyze the profit prospects of AI market and provide support for finding possible development opportunities in the subdivision field.
II. Industry Background Analysis
1. Political factors
1.1 National Policy Analysis
The development of AI industry can not be separated from the policy and national strategic support of relevant national departments; since 2015, the top-level design of AI has gradually improved, and relevant policies have been issued intensively.
Figure 1. Artificial Intelligence Policy Stage (Deqing Jing slightly new, 2019)
In May 2015, "Made in China 2025" mentioned intelligent manufacturing for the first time. It proposed to accelerate the integration of new generation of information technology and manufacturing technology. Intelligent manufacturing should be the main direction of deep integration of the two modernizations. Intelligent equipment and intelligent products should be developed to promote the intelligent production process.
Planning identified six specific areas to support the development of AI, including funding, system standardization, intellectual property protection, human resources development, international cooperation and implementation arrangements.
The plan sets the goal of establishing infrastructure, innovation platform, industrial system, innovation service system and AI basic industry standardization by 2018.
At the same time, the primary goal is to develop a new generation of Internet technology and natural human-computer interaction technology.
Premier Li Keqiang mentioned in the government work report that we should speed up the cultivation and growth of emerging industries. We will comprehensively implement the strategic development plan for emerging industries, accelerate the research and development and transformation of artificial intelligence and other technologies, and expand and strengthen industrial clusters.
In October 2017, Artificial Intelligence entered the nineteenth National Congress report, which will promote the deep integration of Internet, big data, artificial intelligence and real economy.
In December 2017, the "Three-year Action Plan for Promoting the Development of the New Generation of Artificial Intelligence Industry (2018-2020)" was released. As a supplement to the "New Generation of Artificial Intelligence Development Plan" issued in July, it detailed planned the key development directions and objectives of Artificial Intelligence in the next three years. The objectives of each direction were quantified in great detail.
On the afternoon of January 18, 2018, the Artificial Intelligence Standardization Forum released the White Paper on Artificial Intelligence Standardization (2018 edition).
The National Standardization Management Committee announced the establishment of the National Group of Artificial Intelligence Standardization and Expert Advisory Group, which is responsible for overall planning and coordinating the management of AI standardization in China. It also interpreted the "Three-year Action Plan for Promoting the Development of the New Generation of AI Industry (2018-2020)" and "AI Standardization Helps Industrial Development" to comprehensively promote AI standards. Qualitative work.
1.2 Local Policy Analysis
Local government has become the main force to promote the development of AI industry in China.
In the provinces (municipalities directly under the Central Government) issued this level, there are 26 policies distributed in 16 provinces, including Beijing, Shanghai, Tianjin, Zhejiang, Anhui, Jilin, Guizhou and Liaoning provinces have issued two policies.
Strong Artificial Intelligence Policy
Weak Artificial Intelligence Policy
Based on the above analysis, the hotspot areas of AI development and investment in the future are comprehensively analyzed from five quantifiable dimensions: policy intensity, expected industrial scale, policy investment, talent support and enterprise support.
From the policy point of view: Jiangsu, Shanghai, Guangdong, Beijing and Zhejiang are the hot areas for future AI development and investment.
In terms of cities, Suzhou, Shanghai, Guangzhou, Beijing and Hangzhou in Jiangsu are hot cities for AI development and investment.
Figure 2. Artificial Intelligence Policy Scores of Provinces and Municipalities (Deqing Jing slightly new, 2019)
1.3 Legal Analysis
At present, there is no legislation on AI in China. Although AI still has a long way to go to change the law, this kind of change that has been initiated has attracted the attention of academia and industry.
In the future, the legal construction of AI will involve personality rights, intellectual property rights, property rights, tort liability identification, legal subject status and so on.
2. Economic factors
According to IDC's White Paper on Global Artificial Intelligence, global AI expenditure will reach 275.8 billion yuan by 2020.
The Chinese government and capital market attach great importance to AI and make continuous investment, which will promote the rapid development of AI in China. By 2020, the expenditure on AI technology in China will reach 32.5 billion yuan, with a compound growth rate of 32.8% in five years, accounting for about 12% of the global total expenditure.
Figure 3. Global Investment and Financing Amount/Number (Global Artificial Intelligence White Paper, 2019)
Figure 4. Regional distribution of AI investment and financing in China
3. Social factors
In recent years, artificial intelligence has attracted widespread attention of the society. The rapid development of this industry has penetrated into all aspects of national life.
Nowadays, from top-level national design to the rapid landing of industrial applications, AI is gradually becoming a new engine leading the industrial 4.0 subversive change.
According to Baidu Artificial Intelligence Search Index: Since 2016, the search volume center has gradually increased, indicating that the netizens'attention to artificial intelligence has continued to rise.
Figure 5. Baidu Artificial Intelligence Search Index (Baidu, 2019)
According to a survey of 3088 people in the "China Artificial Intelligence Development Report 2018" of Tsinghua University, more than half of the people support the development of artificial intelligence, and only 2.4% oppose it.
The top four areas of national concern for the application of artificial intelligence are as follows: finance, transportation, education and medical treatment.
Figure 6. National Prospects for the Future of Artificial Intelligence (Tsinghua University, 2018)
Figure 7. Ranking of Artificial Intelligence Concerns (Tsinghua University, 2018)
Among the people who search for AI, Baidu Index shows that most of them come from Guangdong, Beijing, Shanghai, Jiangsu, Zhejiang and other developed eastern coastal provinces and cities.
This data is basically consistent with the current regional distribution of AI landing projects, which shows that AI has better application and scale effect in large-scale cities.
Figure 8. Baidu Artificial Intelligence Search Regional Distribution (Baidu, 2019)
In the past year, there has been a marked differentiation among people concerned with AI:
People aged 30-39 had the highest attention, followed by those aged 40-49. In terms of gender, male attention is significantly higher than female attention.
Figure 9. Baidu Artificial Intelligence Search User Portrait (Baidu, 2019)
4. Technical factors
Our country's AI technology level is in the first echelon in the world. Overall, China's technological level is slightly lower than that of the United States, but its catch-up speed is faster.
Our country is the world's leading technology in speech recognition, visual recognition, machine translation, Chinese information processing and other technologies.
Figure 1.0 Proportion of Artificial Intelligence Technology Applications in China (Euro 100 million think tank, 2018)
In terms of basic disciplines, China has achieved rapid growth in Artificial Intelligence papers in the past two decades: the number of published papers has increased from more than 1000 in 1997 to more than 37000 in 2017, and the proportion of papers in the field of Artificial Intelligence has also been rising.
Figure 11. Development Trends of AI Papers in China (Tsinghua University, 2018)
Over the past two decades, the total number of AI papers and highly cited papers in China ranked first in the world.
Since 2006, China has surpassed the United States in the number of papers published, and the global proportion of papers in the field of artificial intelligence has increased from 4.26% in 1997 to 27.68% in 2017. China is far ahead of other countries, and the leading edge has gradually expanded.
Figure 12. Global Artificial Intelligence Paper Output (Tsinghua University, 2018)
China has become the most authorized country for AI patents in the world, slightly ahead of the United States and Japan, while China, the United States and Japan account for 74% of the total number of patent publications in the world.
Global patent applications are mainly concentrated in the following areas: speech recognition, image recognition, robotics and machine learning.
Figure 13. Global Artificial Intelligence Patent Authorization (Tsinghua University, 2018)
The total amount of AI talents in China ranks second in the world, but the proportion of outstanding talents is low.
By 2017, China had 18 232 AI talents, accounting for 8.9% of the world total, second only to the United States (13.9%).
Figure 14. Global Distribution of Artificial Intelligence Talents (Tsinghua University, 2018)
As of June 2017, there were 2617 AI start-ups in the world. The United States ranked first in 1078 households, and China ranked second with 592 enterprises, followed by Britain, Israel, Canada and other countries.
Among them, there are about 78700 employees in 1078 AI start-ups in the United States, while there are about 39200 employees in 592 companies in China, only 50% in the United States.
American AI start-ups mainly consist of teams of 1-10 people and 10-50 people. There are 759 small teams, accounting for 70.41% of the whole country. They are the main force of American AI start-ups.
China's AI start-ups mainly consist of 10-50 team members, a total of 384, accounting for 64.86% of the country.
It can be said that the size of small entrepreneurial teams in the United States is smaller than that in China. When the same technology is needed, the average ability and value-creating of American team is higher than that of Chinese team.
Figure 15. Global AI Enterprise Distribution (Tencent Research Institute, 2017)
5. Analysis of Competition Situation
Artificial intelligence industry can be divided into basic layer, technical layer and application layer from the industrial chain.
At present, it is widely used in the fields of finance, medical treatment, education, security and so on.
Take the financial industry as an example: large state-owned banks have become pioneers in the application of artificial intelligence. They have a large customer base, so AI solutions significantly reduce operating costs.
Due to the homogenization of supply-side competition in the application layer, the bargaining power of buyers has increased.
On intelligent hardware devices, the purchasers of AI enterprises are natural consumers. At present, the market has emerged some subdivision leading enterprise products, such as: Xinjiang UAV and Xiao Ai speaker Xiaomi.
In these distinctive ecological chains, the bargaining power of buyers is weak, while in industries where competition is fierce and no monopoly pattern has been formed, such as smart cars, the bargaining power of buyers is strong.
Figure 16. Domestic AI Industry Chain (Yi Guan, 2018)
In the AI industry chain, the basic level is mainly in the hands of a few giants because of its first-mover advantage.
Competition in the field of technology level is gradually intensifying, which is due to the rapid iteration of AI industry, the difficulty of accumulating comparative advantages, and the small differences among the algorithms of each manufacturer.
At present, the application layer is still in its infancy, and there are many participating enterprises, but the competition is still concentrated in a few sub-areas.
Figure 17. Map of 77 AI enterprises in China (Euro 100 million think tank, 2019)
The development of AI gradually shows the trend of infrastructure. Along with information and data, AI will become one of the infrastructure of future economy.
As the underlying technology of a kind of infrastructure, AI contains a variety of different information technologies, which will be difficult to replace.
Figure 18. Distribution of Artificial Intelligence in China (Euro 100 million think tank, 2019)
5.4 Potential entrants
China's AI enterprises were mainly established in 2014-2016. After 2018, the competition pattern has been gradually determined. New entrants have small technological advantages and single business model, so the number of enterprises established has been greatly reduced.
Figure 19. Establishment time and industry distribution of AI enterprises in China (Euro 100 million think tank, 2019)
6. Industry Technology Forecast
6.1 Foundation Layer
Leading enterprises have accelerated the layout research and development of AI chips.
At present, the types of AI chips mainly include GPU, FPGA, ASIC, brain-like chips and so on.
With the development and application of deep neural network (DNN), its multi-level computing needs can not be satisfied by the traditional CPU, while GPU has the parallel computing ability for deep learning, and its attention is increasing.
In addition, TPU and FPGA chips have also become fast-developing AI chips. In the chip layout of manufacturers to Invida, Intel, Qualcomm, ARM, Apple, Huawei and other manufacturers.
In the era of mobile internet, Android system is loosely coupled with downstream cloud services through GMS, and tightly coupled with upstream chips and vendors through version control to realize the closed-loop ecosystem of mobile Internet with Android operating system as the core.
In the era of artificial intelligence, the development framework also has the core position comparable to the Android operating system. It has the core role of leading the pace of industrial progress, driving the hardware configuration, the coordinated development of terminal scenarios and cloud services, and occupies the key position of connecting the preceding and the following.
Take TensorFlow as an example:
TensorFlow is tightly bound up with Google Cloud, providing cloud machine learning services in cloud platform mode, and making customization optimization in close coupling with chip and hardware manufacturers downwards. Google TPU is dedicated to TensorFlow.
6.2 Technical Level
Figure 20. Development Trends of AI Technology Layer (gwwymx, 2018)
With the rise of the platform, resources of technology, hardware and content have been further integrated.
Artificial intelligence covers a large number of industries and scenarios, and a single enterprise can not involve all aspects of the artificial intelligence industry.
Based on their own advantages, manufacturers enter into the industrial chain and cooperate with other manufacturers, integrating technology, hardware and content resources, and jointly promoting the landing of artificial intelligence technology. With the development of technology, content and hardware, the platform has risen further and the ecological layout has become increasingly important.
Artificial intelligence technology continues to sink into the vertical industry.
General AI technology can not meet the needs of various industries. Different industries have different application emphases and data resources. Market practitioners need to design different industry solutions according to the characteristics of the industry.
Artificial intelligence technology will continue to achieve technology landing from the scene. In vertical industries, medical, financial, security, education, home and other industries have begun to take shape, and the future development prospects are huge.
6.3 Application Layer
The energy efficiency of every link of AI-enabled medical treatment has been shown.
Intelligent education accelerates the innovation of education and teaching. At present, the technology of artificial intelligence and big data is developing rapidly. Intelligent education has become the direction of development in the field of education.
Intelligent education is changing the existing teaching methods, liberating teachers'resources, and triggering profound changes in educational philosophy and educational ecology. At present, the major developed countries in the world are accelerating the innovation of education and teaching, actively exploring new models of education and developing new products of education.
Intelligent transportation improves the level of urban management.
With the rapid development of the global economy, the process of urbanization is accelerating, the number of motor vehicles is increasing, the volume of road traffic is increasing, and various traffic problems are highlighted. Developing intelligent transportation can improve government management, improve user experience and promote urban development.
Artificial intelligence improves public security capability.
Artificial intelligence has been applied in public service fields such as social security, anti-violence and anti-terrorism, disaster early warning, post-disaster search and rescue, food safety and so on.
Artificial intelligence can accurately perceive and predict the major trend of social security operation, improve the precision level of public services, and ensure the safety of people's lives and property.
From the perspective of depth and breadth of application, global artificial intelligence is still in the exploratory stage in the field of public services.
III. Analysis of the Current Market Situation
1. Market Phase
The market stage is divided into: introduction period, development period, maturity period and decline period.
According to Gartner's 2008 emerging technology cycle, AI-related technology is still on the rise of hype; that is, in the early to middle stages of development, most of AI emerging technology will be applied in two years.
Figure 21. Gartner 2018 Technical Cycle (Garter, 2018)
2. Market size
The global AI market is expected to reach $1.2 trillion in 2018 and $3.9 trillion by 2022.
Among them, the calculator vision market with biometrics, image recognition and other technologies as the core is huge, accounting for 34.9%.
Figure 22. Size of AI market in China (China Industrial Information Network, 2018)
Figure 23. Market Structure of Artificial Intelligence in China (Tsinghua University, 2018)
1. Application scenarios
Artificial intelligence has brought changes and reconstruction to all walks of life.
On the one hand, new technologies are applied to existing products, innovative products and new application scenarios.
On the one hand, the development of technology has also subverted traditional industries. The replacement of artificial intelligence to artificial intelligence has become an irreversible trend of development, especially in simple repetitive and procedural links such as industry, finance and agriculture. In defense, medical, driving and other industries, AI provides more accurate and efficient professional services that can adapt to complex environments, thus replacing or strengthening traditional manual services. The service forms will tend to be personalized and systematic in the future.
For the application of artificial intelligence, technology platform, industrial application environment, market, users and other factors have a great impact on the industrial application market of artificial intelligence.
At present, the main application scenarios of AI technology include, but are not limited to, security, manufacturing, service industry, finance, education, media, law, medical, home, agriculture, automobile, etc.
2. Competitive strategy
Under the trend of AI platform, AI will present a competitive pattern in the future: several dominant platforms and a wide range of scenarios, among which ecological constructors will become the most important one.
With Internet companies as the main part, long-term investment in infrastructure and technology, and scenario applications as traffic portals, accumulate applications, and become the leading application platform, will become artificial intelligence eco-builders (such as Google, Amazon, Facebook, Aliyun, etc.).
Key success factors: a large amount of investment in computing power, the accumulation of large quantities of high-quality multi-dimensional data, the establishment of algorithm platform, general technology platform and application platform, with scenario applications as the entry point, the accumulation of users.
With software companies as the main platform, deep-plough algorithm platform and general technology platform, and scenario application as the traffic entry, application platforms (such as Microsoft, IBM Watson, etc.) are gradually established.
Key success factors: deep tillage algorithm and general technology, establish technological advantages, and take scenario application as the entry point to accumulate users.
Based on scenarios or industry data, a large number of scenario segmentation applications are developed, mainly by entrepreneurs and traditional industry companies.
Key success factors: grasp market segmentation data, select appropriate scenarios to build applications, build a large number of multi-dimensional scenarios, seize users; at the same time, cooperate with Internet companies, effectively combine traditional business models and artificial intelligence.
In the vertical field, the pioneers mainly rely on killer applications (such as travel scene applications, face recognition applications) to accumulate a large number of users and data, and to further cultivate the general technology and algorithms in this field, becoming subversive in the vertical field (such as drip trips, ignorance of technology, etc.).
Key Success Factor: In the scenario with extensive application and massive data, killer applications can be introduced first, thus accumulating users and becoming the leader of the vertical industry; by accumulating massive data, it will gradually expand to application platforms, general technology and basic algorithms.
With chips or hardware as the main infrastructure companies, we should start from infrastructure, improve technological capabilities, and expand to the upstream of industrial chains such as data, algorithms and so on.
Key Successful Factors: Develop new chips with intelligent computing capabilities, such as image, speech recognition chips and other extended chips application scenarios, widely integrated in mobile intelligent devices, large servers, UAVs, robots and other equipment and facilities, to provide more efficient, low-cost computing capabilities, services, and deep integration with related industries. (Boston Consulting, 2019)
At present, Internet companies and software giants are in the industrial chain technology layer and application layer to start layout.
The AI industry is still in a period of rapid growth.
China is the fastest growing and most promising country in the field of artificial intelligence in the world, and has led the world in some indicators. In the next five years, the AI industry will continue to grow at a high speed.
At present, there are still some subdivisions of vertical areas which need to be developed urgently. We can start from the following aspects:
The basic and technical layers of AI are gradually controlled by the giants, and the landing of the application layer is still in its infancy. Therefore, if the start-up company has no comparative advantage, it should concentrate on developing the virgin land of the application layer.
With the vigorous development of various emerging technologies, companies should make full use of existing and developing advanced technologies such as machine learning, artificial intelligence, speech recognition, image recognition and so on, enabling products.
According to the capital preference, in the current market stage of the field, we should make full use of the advantages of the pioneers, layout ahead of time to quickly improve product visibility and seize the market ahead of time.
Monitor and deal with possible risks at any time according to the competition situation.
With the gradual improvement of AI laws at home and abroad, legal risks should be avoided in advance.
All the contents of this material come from the open information on the internet. After sorting out and summarizing, this document is formed for study and use only.
National Policy Analysis is cited from Guowei, AI Industry Technology Innovation Alliance http://www.qianjia.com/html/2018-01/22_283111.html.
Deqing Jing slightly new. (2019). Policy Interpretation I Artificial Intelligence Policy Summary. http://www.sohu.com/a/292560106_100179411 (Local Policy Analysis Text Part is also cited from this link)
The legal analysis part of this article is quoted from the Legal Daily. What position should the law give to AI? Realize the overlapping of human intelligence. Http://tech.qq.com/a/20180123/007400.htm
Tsinghua University. (2018). Global Artificial Intelligence White Paper. 2019.
Euro 100 million think tank. (2019). China Artificial Intelligence Investment Market Research Report 2018.
Tencent Research Institute. (2017). White Paper on Global Artificial Intelligence Talents, 2017.
Analytical Consulting. (2018). Artificial Intelligence Industry Chain.
Gwgwymx. (2018) CSDN. https://blog.csdn.net/gwgwymx/article/details/80805236
Part of the development trend comes from China Information and Communication Research Institute and China Artificial Intelligence Industry Development Alliance. 2018. White Paper on Artificial Intelligence Development.
Garter. (2018) Gatner hype cycle. https://www.gartner.com/smarter with gartner/5-trends-emerge-in-gartner-hype-cycle-for-emerging-technologies-2018/
China Industrial Information Network. (2018). Forecasting the scale of China's artificial intelligence market in 2018 and the latest policy analysis.
Boston Consulting. (2019). 5 types of AI competition models will emerge. Big data is a strategic competitive advantage.