The full picture of China's artificial intelligenc

  • Detail

The overall picture of China's AI industry is shown in this AI sphere of influence map by Ma guanxia: the rise of AI in 2017 was like a rainbow. Entrepreneurs reveled in the capital boom and rushed to call for the revolutionary changes that AI might bring to various industries. However, in the past 2018, the wave of blockchain and the ups and downs of shared bicycles once eclipsed AI

this is not a bad thing. Away from the noisy foam, the AI industry is more down-to-earth after the arrival of the capital winter. No matter the technology giants or start-ups, finding application scenarios based on their own technical capabilities and characteristics has become the top priority, "landing" is the theme of this year

in the past year, the Matthew effect of the high resilience of China's artificial intelligence industry has been increasingly strengthened. When small companies are facing common financing difficulties, leading star enterprises are breaking the financing records of this industry one after another; Bat, which has already completed the layout of AI industry, is the first to enter the era of full application and landing, while the start-up companies in the hot track are facing more fierce landing disputes; In 2019, when the investment boom recedes, small AI start-ups facing a shortage of funds and limited application scenarios can only win more living space by finding application scenarios in vertical fields

industrial chain distribution

according to the data of the China Institute of information and communications, as of September 2018, there were 1122 artificial intelligence enterprises in China (excluding Hong Kong, Macao and Taiwan), which were distributed in all links of the AI industrial chain

ai industrial chain can be divided into three categories according to the basic layer, technology layer and application layer. The basic layer provides computing power for the whole industry. The hardware part includes chips, sensors and middleware. The software provides a computing power platform for the whole AI industry chain based on cloud computing

in the field of basic platform, it is mainly provided by giants such as Tencent, Baidu and Alibaba. It is worth noting that mining machine manufacturers such as bitland are transforming to artificial intelligence chip research and development

ai industrial chain distribution (Tencent deep mapping)

the technical layer provides the overall industrial chain with the ability to start and compile the directory of major short board equipment innovation and development for the general AI technology. The perception layer includes two machine perception tasks, computer vision and speech recognition, which are relatively mature at present; The cognitive layer is positioned as "machine brain", including two core areas: knowledge map/semantic analysis and intelligent question answering/virtual assistant; The top platform layer provides in-depth learning, pattern recognition and other technical application services in the form of a general technology application platform to connect with the application layer

at present, the domestic start-ups mainly facing the to C market are basically concentrated in the technology layer, such as iFLYTEK, which provides voice services, Shangtang technology and Kuangshi technology, which serve intelligent enterprises, and so on

the application layer is specific consumer terminal applications and industry scenario applications. Consumer terminal applications include intelligent robots, intelligent UAVs and intelligent hardware; Industry scenario applications interface with AI application scenarios of various external industries, such as autonomous driving, smart finance, smart medicine, smart education, smart retail, smart security, smart marketing, smart city, etc

except for some consumer applications, this is a huge to B market. The development direction of "industrial interconnection" proposed by Tencent is closely related to artificial intelligence

bat accelerated layout

as the most important participant in the domestic AI industry, the Internet giant bat has completed a comprehensive layout in AI technology, platform, application scenarios and foreign investment. Bat's focus on business areas also reflects its layout in the AI industry. Generally speaking, Baidu focuses on the platform and automatic driving; Alibaba focuses on the application and underlying technology of artificial intelligence in the field of data services; Tencent focuses on platform and technology opening, with balanced external layout

bat artificial intelligence layout (one group of Tencent jigs for crank test and one group of toothed disc test jigs for deep mapping)

at the world Internet Conference at the end of last year, Robin Lee said that Baidu hopes to build a basic platform to open some basic capabilities of artificial intelligence, such as the framework of deep learning, voice recognition technology and image recognition technology, to the outside world. At the application level, Robin Lee is optimistic about two scenarios: intelligent home and intelligent transportation

Ma Huateng believes that in the future, the interconnection of the entire future external network between man-machine and human brain will be meaningful. There will be many changes in the new man-machine interface in the next 10 years. If 5g plus VR foundation, endurance capability and computing power are solved, to a certain extent, more practical VR computing software and AR computing software will appear, which is also an epoch-making breakthrough

At present, AI Unicorn start-up companies are concentrated in the track with mature technology or wide application scenarios, such as computer vision, speech recognition, AI chip, intelligent finance, intelligent medical treatment, etc

most of these companies were founded around 2012. Highly educated entrepreneurs with technical background felt the development tide of AI technology, set up teams and started their entrepreneurial journey from their own fields

these companies started their businesses before the market knew about AI, which won them a substantial first mover advantage. In 2016, the hottest year in the venture capital market, these companies that have completed technology research and development, team running in and preliminary market testing received extremely high valuation financing. With the support of funds and the accumulation of public praise, the position of the head start-up companies has been consolidated and become an important pole of China's AI industry

Tencent has sorted out according to public data.

in the most mature field of computer vision, four star Unicorn companies, Shangtang, Yuncong, Kuangshi and Yitu, are widely deployed in the two most easy to land and broad market areas, finance and security

in the financial field, Shangtang technology's customers include many connected financial institutions and commercial banks; Yuncong technology, headquartered in Chongqing, is the largest supplier of face recognition in China's banking industry, although it does not have an advantage in terms of publicity force and region; Backed by the investor Alibaba, Kuangshi technology provides face recognition login function support for Alipay clients, and also provides face recognition services for many Internet financial companies and commercial banks; Yitu technology has a complete real name authentication solution. Relying on face comparison and live detection technology, it provides financial enterprises with omni-channel solutions from counters to self-service devices and then to mobile terminals

in the field of security and smart city, the main landing scenarios of Shangtang technology are to solve anti-terrorism, criminal arrest and public security through intelligent video; Cloud technology is mainly based on the combination of core technologies such as large database retrieval, dynamic control deployment, track tracking and algorithm items of various business lines, and is continuously applied in the inland of multi local public security systems; Based on intelligent camera hardware, Kuangshi technology covers multiple application directions, such as face recognition access control, sky eye monitoring system, dynamic face recognition monitoring, and integration of human and ID cards; The "Dragonfly eye" portrait platform developed by Yitu technology has also been applied in the field of public security. At the same time, it also provides a portrait comparison system for the General Administration of customs and China border inspection

in addition to finance and security, Shangtang, Yuncong, Kuangshi and Yitu also have different degrees of competition in medical, transportation, retail and other fields. When there is little difference between the technology and the landing scene, fierce fighting is no longer inevitable

however, for AI unicorn, which is still "not bad for money" in the cold winter of capital, in addition to expanding application scenarios in different fields, accelerating its own commercial landing through equity investment can also become another option

on August 7, 2018, Shangtang technology led projective spectrum technology in the round D financing of 1.36 billion yuan. On June 21 earlier, Shangtang technology also led the investment in the round B financing of medical Internet company Helian health of 75Million US dollars. The latter was valued at nearly 1billion US dollars after investment - unicorns began to build small unicorns. At present, the investment cases that Shangtang has made public include 7 projects, including 51vr, Helian health, Suning sports, and Yingpu technology

on November 2, Kuangshi technology appeared in the list of investors for a new round of financing of tens of millions of dollars by "good neighbors" of chain convenience stores. It is generally believed that the current continuous investment in store AI equipment and digital transformation by the chain convenience store industry are the basis for their cooperation. The phenomenon that Shang Tang and Kuang Shi are investing while financing is also seen as the epitome of the intense competition for the commercial landing of the current head AI enterprises

layout of investment institutions: the artificial intelligence track of capital inflows into the head company covers major mainstream investment institutions in China, including Zhenge fund, Sequoia Capital China, IDG, Innovation workshop and other funds, as well as Tencent, Alibaba, Baidu, Lenovo and other industrial capital

the active capital environment obviously has a great role in promoting the cultivation of the AI industry, and the mutual promotion between capital and enterprises has also enhanced the overall strength of China's AI industry

(Tencent deeply collates and ranks in no order)

in 2018, it is a general trend in AI venture capital to concentrate funds on leading companies. Even in the face of the cold winter of capital, star projects on popular racetracks such as Shangtang technology, youbixuan and Kuangshi technology are still popular. Taking Shangtang technology as an example, this Unicorn company in the field of computer vision technology completed the financing of round C, round c+ and round D totaling US $2.22 billion from April to September this year, and the valuation climbed to US $6billion

it orange statistics show that in 2018, there were 410 investment events in the AI field, with a total investment of 107.8 billion yuan. Compared with 2017, the number of investment events decreased by 1/3, but the total investment increased by 1/4. Among the invested projects, the investment of early projects decreased, and the proportion of projects in round C and later increased. On the one hand, it means that it is more difficult for start-up AI companies to finance. On the other hand, it also confirms the reality that funds are flowing to the head companies

from the technical point of view, computer vision is still the most popular artificial intelligence technology in the venture capital circle, which is related to its high technical maturity and wide application scenarios in security, finance and other industries. In addition to Shangtang technology, Kuangshi technology, Yitu technology and Yuncong technology, both Unicorn companies in the CV field, have also completed hundreds of millions of dollars of financing in the past year

however, voice recognition, another relatively mature AI technology, rarely sees large-scale financing. The valuation of Companies in the first echelon of the computer vision track has reached billions of dollars, while the valuation of Companies in the first echelon of the voice recognition track is generally only about 3billion yuan

like the Internet industry, the business models of AI companies can be basically divided into to C, to B and to g. in particular, the security market with an annual budget of more than trillion yuan in the to g business has already become a red sea for many AI start-ups to compete. However, due to the congenital lack of to G application scenarios in the voice recognition track, its revenue and valuation scale are not the same as those of the companies on the visual recognition track. According to the data previously released by EU billion think tank, the revenue of computer vision Head Companies in 2018 can reach about 2billion yuan, while the revenue of voice recognition head companies is mainly around 200million yuan

from the perspective of industrial applications, intelligent robots, intelligent medicine and intelligent finance are still favored by various investment institutions. In the smart robot track, Ubisoft technology ranks second only to

Copyright © 2011 JIN SHI