In the past two years, China's artificial intelligence (AI) industry has been booming. Both the government and enterprises are making efforts to contribute to the AI development in a coordinated way. The concept of AI is substantially popular in the China's society. There are even some students that have started to take up artificial intelligence courses since elementary school. This makes people feel like the country is undergoing a national AI campaign. Most foreign media also hold a positive view that China's AI has great potential to surpass that of the United States.
According to reports, Amy Webb, who is the founder of Future Today Institute (FTI) and also a professor at New York University, released her annual technology trend report at the South by Southwest® (SXSW) conference in Texas. Her report focused on the influence of the AI and highlighted that "in 2018, China will rely on the strong basis of AI to become the world's powerhouse". In her report, she also said that BAT (Baidu, Alibaba, Tencent) are three companies worthy of attention. With the release of the news, many people were heartened to hear about China's potential in AI. However, people should calm down and think carefully, whether China is leading the world in artificial intelligence? Recently, Bill Gates gave his opinion on the current development of AI in an interview with reporters. In brief, Bill Gates' core point is that although China's AI is developing rapidly and supported by the policy, he does not believe that China can surpass the United States in AI industry, and China will be the global second in the field of AI. This statement sent a wave of shock through the AI industry.
So, what are the facts about the AI industry? Who is the leader in the AI industry? In this regard, Anbound's industry adviser and semiconductor expert, Mo Dakang has made a detailed comparative analysis of the current AI industry. First of all, from the perspective of development dimensions, there are three AI development dimensions. They include foundation, technology, and application. The primary indicator to measure the AI competitiveness of an enterprise or even a country lies in its AI chip's development level in the foundation dimension of development (ex. Computational RAM). According to the data, in the current chip market, especially the main artificial intelligence chip (GPU) market, Intel and Nvidia accounted for 71% and 16% of the chip market respectively. That is to say, American enterprises currently occupy about 90% of the AI chip market share and have a natural monopoly advantage in AI chip industry.
According to the "China-US Artificial Intelligence Industry Development Report" released by Tencent Research Institute last year, there are 33 chip companies in the U.S. while there are only 14 chip companies in China. China's chip companies only accounted for 42% of that in the United States. In term of financing in the AI industry, the report shows that the U.S.'s chip/processor industry has a financing scale of USD 31.5 billion, accounting for 31% of the total U.S.'s AI industry financing. The proportion of financing scale also ranked first in the U.S.'s AI industry. In contrast, China's proportion of financing scale in chip/processor industry ranked fourth in China's AI industry, only accounting for 7.55%. From these two figures, it can be seen that the development of China's AI chip is difficult to surpass the United States in the short run.
The second indicator to measure the AI competitiveness of U.S. and China lies in the AI talent pool. According to the "Global AI Talent Pool Report" released by Linkedin, there are more than 1.9 million AI relevant talents in the worldwide (based on the LinkedIn platform), among which the total number of AI relevant talent in the United States exceeds 850,000, ranking first. Meanwhile, the number of AI relevant talent in China is only 50,000, ranking it seventh in the world. In addition, United States has a high concentration of AI talents at the foundational dimension of AI development, account for over 70%. On the other hand, China has a wide distribution of talent at the applicational dimensions of AI development, such as in the fields of robotics, image recognition, precision marketing, and autonomous driving. This means that Chinese AI talents are more focused on the application of AI, and there is still a big gap in the development of basic chips compared with the United States. Foreign media used to prove that the AI development in China has surpassed that of the United States with the top volume of China's papers published in the field of artificial intelligence since 2014. In fact, considering the number of weighted citations, the influence of China's paper ranks only 34th in the world, far behind the 4th that of the United States. Hence, it is obvious that China's pursues quantity over quality in the development of the AI industry.
To sum up, although the Chinese government, investment institutions, and enterprises have invested heavily in the AI industry, the input of enterprises, capital and talent in the foundation dimension of AI development are insufficient. This indirectly reflects the entirety of China's AI industrial chain to generally favor developing the technology and application dimension of AI development to bring short-term profitability rather than slower profitability and large investment required of the AI chip industry. It is unlikely for China's AI industry to surpass that of the United States without some advantages in the AI chip industry, and the constraint of IT development due to the deficiency of chips industry in the PC and mobile Internet era will continue into the AI era. Therefore, although Bill Gates' remarks are slightly scathing, but it does sober up China's AI development from its overconfidence and prompts China to reflect on the current inadequacy of its AI development.
Final analysis conclusion:
China's AI industrial chain generally avoided the slow profitability and large investment required of the AI chip industry. Without advantages in the chip industry, the constraint of China's IT development due to the deficiency of the chips industry will continue into the AI era.
