China will soon become the global center of artificial intelligence technology

What role will China's artificial intelligence play in the world? Recently, venture capital firm Vertex published an ecological research report, predicting the future development of China's AI field from the perspectives of industry, technology, policy and investment. The report believes that China will soon become the center of global artificial intelligence technology.

[[198973]]

1. Five factors contributing to the rise of AI in China

Five factors have contributed to China’s development into a global AI center: (1) multiple industries hope to use AI to achieve digital transformation; (2) a large number of high-end AI talents; (3) a promising mobile Internet market; (4) high-performance computing technology; and (5) government policy support. Among them, the first two factors are particularly important and are China’s unique advantages in developing into a global AI center.

1. Many industries hope to use AI to achieve digital transformation

Catalyst for enterprise growth: When non-AI technology companies grow to a certain scale, they often want to explore how to use AI to upgrade. However, there are still many traditional Chinese companies that lag behind the United States in the use of advanced technologies. Companies with data and economic strength are more willing to hire AI experts and use AI technology to discover business development opportunities or save a lot of costs.

Multi-industry application: It is expected that there will be a large number of companies that need AI technology, including state-owned enterprises in traditional industries such as finance, telecommunications, and agriculture. Other fields (such as the education industry) may also benefit greatly from AI technology.

Raising AI to a strategic level: China’s tech giants are engaged in an AI “arms race.” For example, Baidu is developing an autonomous driving system and has an AI team of more than 1,300 experts; Alibaba Cloud’s machine learning platform PAI provides AI services for enterprise data analysis; Tencent has established a dedicated team of 250 scientists and engineers and invested in iCarbonX, a Chinese AI startup in the health field.

Economic growth: According to McKinsey, China has benefited greatly from its demographic dividend over the past decade, but this advantage will no longer exist as the population ages. However, automation technologies such as AI will inject energy into the Chinese economy, and AI technology may contribute 0.8 to 1.4 percentage points to China's GDP, depending on the speed of technology utilization.

2. A large number of high-end AI talents

Applied research: Chinese scholars are actively and pragmatically engaged in AI research and are committed to discovering new areas of AI application.

Experience and expertise: There are fewer experienced top AI talents in China. Of the 250,000 AI experts on LinkedIn, about half of them with ten years of experience are in the United States, while less than 25% are in China. China should cultivate more experts in the field of AI, and corporate leaders and middle-level managers also need to learn technology and develop their ability to understand and apply data. However, China has a large number of high-end engineering talents, which is very important for the development of AI applications.

Talent costs and migration: Chinese companies such as Baidu hire top AI experts from the United States. Under the influence of Trump's immigration policy, Chinese companies have encountered a good opportunity to recruit more AI talents from Silicon Valley. About 1/4 of high-tech companies in the United States were founded by immigrants. Some practitioners in the AI ​​industry have very high salaries, and many experts leave large companies to start AI businesses.

AI specialization: At the application level, China's algorithm development is comparable to that of other countries; in particular, most Chinese AI scientists are dedicated to specialized fields such as computer vision and speech recognition. However, China still lags behind other countries in basic research.

3. The mobile Internet market has broad prospects

Vast mobile Internet market: In January 2017, a report from the China Internet Network Information Center showed that China had 731 million Internet users.

Data flood: China has a large number of Internet users, and massive amounts of user data provide a large amount of raw materials for Chinese technology companies.

APP-based AI applications: Once AI applications are ready, APPs such as Baidu, Alipay, WeChat, and Sogou Input Method will apply AI technology in their services.

Barriers to entry: Although Chinese Internet companies face far fewer restrictions than foreign companies such as Google and Facebook, Chinese technology companies still face market entry barriers.

4. High Performance Computing Technology

Hardware technology: Chinese technology companies continue to make breakthroughs in this area, such as Shenzhen, which has built an ecosystem to support AI hardware technology.

Rapidly evolving HPC technology: HPC technology is evolving rapidly.

Chip manufacturing capabilities: China has increased its attention to chip manufacturing and promised to invest $150 billion to improve China's chip manufacturing capabilities over a decade.

Global value chains: China needs to ensure that the AI ​​industry is built on an open system that integrates global technologies.

5. Government policy support

Government policies and plans: The Chinese government has been promoting the development of the AI ​​field through policy support since 2014.

R&D investment: China’s success in AI stems in part from government investment in university research. The China Development Research Council (NDRC) has established a national engineering laboratory for deep learning research and application.

China’s market entry barriers: China sets market entry barriers for foreign companies; China’s open government data sets promote domestic innovation, but the data is not open to the outside world. Restrictions on cross-border data flows are not conducive to international collaboration. However, China’s AI policy is more relaxed for experiments and solutions.

Promoting the development of the integrated circuit industry: The Chinese government has taken a variety of measures to promote the development of the integrated circuit industry.

2. China’s AI market will grow rapidly, affecting multiple industries

Traditional technology giants Baidu, Alibaba and Tencent are currently leading the field of artificial intelligence. After them, there are hundreds of domestic startups exploring new technologies in various directions of artificial intelligence. At present, voice and computer vision are the two hottest directions in the domestic artificial intelligence market, accounting for 60% and 12% of the market respectively (as shown below). On the other hand, companies in traditional industries are also actively introducing artificial intelligence to reduce their operating costs. iResearch Consulting said in a survey report: China's artificial intelligence market will reach a scale of US$9.1 billion in 2020, which means that the annual growth rate will reach 50% in the next few years.

Under the influence of artificial intelligence, new sub-sectors will emerge (such as drones and intelligent robots), and traditional industries (such as home appliances, automobiles and toys) will also undergo profound changes. As the world's largest automobile market (a total of 28.03 million cars were sold in China in 2016), the largest home appliance producer and the largest drone producer (DJI occupies 70% of the global commercial drone market), China is forming the world's most attractive artificial intelligence ecosystem.

Among them, the main industries that will undergo profound changes under the influence of artificial intelligence are:

• Automotive: The advanced driver assistance systems (ADAS) market is expected to reach $37 billion in 2020, while the autonomous vehicle market is expected to be worth $87 billion around 2030.
• Manufacturing: China's manufacturing market is worth $210 billion. IHS predicts that by 2020, China will export 700 million smart home appliances.
• Robotics: In 2014, home service robots accounted for 16.7% of the $35 billion global robotics market.
• Drones: The global drone market is expected to reach $12 billion by 2020.
• Toys: In 2015, the global toy market will be worth approximately US$80 billion, of which 30% ($26 billion) will be occupied by smart toys.
• AR/VR: Goldman Sachs predicts that in the future, the global AR/VR market will reach 80 billion US dollars, of which hardware products will account for 45 billion US dollars.

1. Research strength

China's AI research institutions are developing rapidly. A 2016 White House study showed that China has published more scientific research papers on deep learning than the United States. China's pure AI research has ranked first in citations, and the impact of published research has reached third in the world. Recently, a survey by the Nikkei Asian Review showed that China filed 8,410 patent applications in the field of AI between 2010 and 2014, an increase of 186% compared to 2005-2009.

Many new advances in the field of artificial intelligence have emerged from China and then spread to other parts of the world. In terms of deep learning, China's research level is in the first echelon. With the continuous investment of technology giants, this trend will continue in the future.

2. Adequate funding

In recent years, the Chinese market has witnessed a boom in AI investment. According to NetEase, a total of 202 AI startups received a total of $1 billion in financing in 2016. KPMG's research also shows that investment institutions have favored domestic AI companies in recent years and expects this trend to continue in the future. According to a report released at the Wuzhen Summit, domestic companies invested $2.6 billion in AI last year, while the US figure for the same period was $17.9 billion.

Although the current gap is still obvious, some reports show that the amount of venture capital in North America and Europe is shrinking, while the momentum in China is still strong. Some reports show that some research teams have received six times the investment in China than in Europe/North America, which allows them to build complete AI laboratories and recruit professional teams composed of PhDs.

In China, local governments in many regions are also actively supporting the development of the artificial intelligence industry. For example, Xiangtan City in Hunan Province has pledged to invest $2 billion to support the development of the robotics and artificial intelligence industries. In Suzhou, artificial intelligence companies can receive $800,000 in subsidies to establish branches; while Shenzhen offers a $1 million reward for companies that come to establish research institutions.

Currently, AI has attracted the vast majority of angel/A round investments, with the figure exceeding 90% in 2016. The current pace of development has even given rise to concerns that the industry is developing too fast.

3. Distribution and Patent Application of Chinese AI Startups

1. Distribution of AI startups

China’s AI startups are mainly concentrated in Beijing, Guangdong, and the Yangtze River Delta economic region.

AI startups in these regions account for approximately 85% of all AI startups in China.

2. Patent Application of Artificial Intelligence

Robotics, neural networks, speech and image recognition account for the majority of patent applications.

Since 2011, China has made rapid progress in the application of patents in robotics and computer vision.

IV. China’s AI Ecosystem

China's AI ecosystem includes large Internet companies and emerging vertical companies. The following analysis is from the application layer, technology layer and infrastructure layer.

Chinese AI companies at the application layer are divided into the following fields:

• Robots: Geek+, Rokid, Turing Robotics, UBTECH
• Autonomous driving: Baidu, Tiantong Weishi, Horizon Robotics, UISEE Technology
• Drones: DJI, Ehang, Hover Camera, Zerotech
• Voice assistant: Baidu, Mobvoi
• Business Intelligence: Yonghong Technology, Data KM
• Consumer Services: AiKF
• Industry applications: iCarbonX, Maxent, Toutiao, Xuebajun

Chinese artificial intelligence companies at the technical level are divided into the following fields:

• Speech recognition & natural language processing: AISpeech, Baidu, iFlytek, Mobvoi, JCT Watson, Tencent, Triangle, Unisound
• Machine Learning & Deep Learning: DeePhi Technology, China Science Vision
• Artificial intelligence platform: CloudMinds Technology, 4Paradigm
• Computer Vision: Yitu Technology, DeepGlint, Megvii Technology, SenseTime

Chinese AI companies at the basic level are divided into the following fields:

• Sensors: ICE DRINK, LeiShen, SLAMTEC, Benewake Photonics
• AI chips: Cambrian Technologies, Horizon Robotics
• Data: DataDouDou, Data Hall
• Computing power: Alibaba, Baidu

1. Application layer: The participation of experts in various vertical fields has raised the entry threshold of AI

Nowadays, all fields of the industry are embracing AI, and its application prospects are very broad. As the commercial value of AI is constantly gathering towards deep learning, there are several specific fields that are difficult to enter without capital and vertical field experts.

1) Education

• Process-based boring, basic activities such as upgrading exams
• Personalized learning through customized courses
• Identify subjects where curriculum needs improvement

2) Finance

• Cleaning and processing satellite image data for financial/sales forecasting
• Identify credit risks and close or restrict accounts that are likely to default

3) Logistics

• Artificial intelligence can be found in the use of robots and sensors that can transport, store and retrieve systems, collect orders, and check inventory.
• Algorithms also allow AI to automatically process events in the supply chain in real time.

4) Agriculture

• Optimize seed sowing, irrigation, spraying, and harvesting
• Sorting vegetables and fruits to save labor costs
• Identifying sick livestock based on changes in audio data

5) Retail

• Implementing image-based product search
• Improve recommendations with large data sets on sales, listings, and consumer preferences
• Improve online search and services
• Forecast product demand and optimize pricing

2. Technology: 26% of all AI companies in China are mainly engaged in computer vision, machine learning, and NLP

3. Basic layer: The rise of mobile edge devices drives the demand for AI embedded chips

The next wave of computing, including 5G, artificial intelligence, self-driving cars, cloud computing, and federated learning, is expected to be the main driving force behind the transformation of computing architecture.

We have seen widespread changes in computing architecture, including a shift from homogeneous to heterogeneous computing and from centralized to edge computing.

In China, edge computing has more traction. In October 2016, Huawei, Shenyang Automation Research Institute, China Academy of Information and Communications Technology, Intel, ARM, and iSoftStone jointly established the Edge Computing Industry Alliance.

The government, state-owned enterprises and locally controlled industries are the largest consumers of technology in China.

Most technological advances are supported by the government. With government involvement, the development of edge computing is expected to have great growth momentum.

Vertex believes that with strong government support and massive investment from technology companies, China's artificial intelligence will enter a golden age in the next few years. China will play a pivotal role in the field of artificial intelligence. For traditional and technology companies, artificial intelligence technologies such as deep learning will bring higher and higher barriers to entry in various vertical fields, which will benefit technology-leading companies. In the next decade, new opportunities will emerge in companies that build "intelligent systems."