Dr. Edoardo Mirabella

Europäischer Patentanwalt

It is generally believed that China will play and, indeed, is already playing an important role in the development of Artificial Intelligence (AI). Unfortunately, in many discussions, this belief reduces to nothing more than an unsubstantiated hearsay. In this regard, the latest event of the “What is AI?” meetup organised by Charlotte Han has the commendable merit of asking two Chinese AI experts, Yuxiang Chen and Jie Liu, to quantitatively analyze the present and future role of China in the field of AI, thereby answering Charlotte’s provocative question as to whether one should learn Chinese to advance in AI The main takeaway of the event is that China does indeed play an important role in both the theoretical and technological development of Artificial Intelligence. In Andrew NG’s words, “today, a significant amount of AI research and education is taking place in the United States and China. These countries are home to many of the best universities in the world, and their governments have provided funding and created thoughtful regulations that enable innovation. But the United States and China have also built incredible business ecosystems” (WIPO Technology Trends 2019 Artificial Intelligence, ISBN: 978-92-805-3007-0).

AI ECOSYSTEM IN CHINA

A recent analysis performed by the Allen Institute for Artificial Intelligence shows that China has already surpassed US in published AI papers. Projections shows that China may have more top-10% and top-1% papers than US by 2020 and 2025, respectively. One half of the top 20 universities and public research organisations by number of scientific publications is Chinese:

In October 2017, the Alibaba group has founded the DAMO Academy with the aim to promote technological collaboration and perform research in cutting-edge technology such as Machine Learning (ML), quantum computing, natural language processing and human machine interaction. The Alibaba group expects to invest in the Academy 15 billion US dollars in the triennium 2017 – 2020. In a recent study, a CB Insights team has selected the top 100 AI startups from a poll of about 3000 companies. Five out of eleven top AI unicorn startups are Chinese, although US companies still constitute the vast majority of the top 100 AI startups: The Chinese industrial ecosystem appears to have all the production resources to become the leading force in the fourth industrial revolution (4IR). Several Chinese companies already provide core technologies, which allow for collecting and transferring raw data from the physical world e.g. by means of an Internet of Things network. China has also know-how in the enabling technologies needed to store and process huge amount of raw data and to analyse them with unprecedented accuracy and computational power, thereby providing the raw material of the 4IR applications. Many Chinese industries provide such smart applications e.g. in the field of logistic, transport, safety control and education. An interesting exemplary 4IR application is the Alibaba ET city brain project which aims to optimise urban public resources by processing and analysing in real time multi-source city data: Artificial intelligence has become an important component of the toolkit of Chinese companies. A recent study of the Boston Consulting Group shows that 85 % of the Chinese companies either have already adopted or are running pilot programs to adopt AI into their business processes: A prominent example of a Chinese company, which successfully adopted AI to streamline its business is Alibaba, which, on the Singles‘ Day of 2018, has been able to successfully process more than one billion orders by heavily relying on AI and ML solutions:

CHINESE FACTORS FOR SUCCESS

The rising of the Chinese star in the AI firmament is the result of political, sociological and cultural factors. Chinese authorities consider AI of paramount importance for the economic growth of the country. AI was a major topic of the 19th National Congress of the Communist Party of China and the Chinese State Council released a plan for boosting AI development. China aims to become competitive in AI by 2020, a goal that seems to be within its reach, if not already achieved. In the next decade China intends to become a prominent player in AI and, ultimately, the AI world leader: It does not come as a surprise that the two most well-funded AI startups unicorns are Chinese companies, SenseTime and Face++, which benefit from government funds and have many government clients: The political agenda of the Chinese Government is motivated, and its success may be made easier, by distinctive sociological and cultural aspects of Chinese society:

INTELLECTUAL PROTECTION FOR AI IN CHINA

The importance of the Chinese contribution to AI developments may also be inferred by looking in the IP sector. A study published by the World Intellectual Property Organisation (WIPO) shows that about 17 % of the top 30 AI patent applicants by number of patent families is Chinese. The Chinese presence is even more significant if the ranking is restricted to Universities and the public research organisations. In this case, more than 100 Chinese institutions rank in the top 500 AI patent applicants, seventeen of them being in the top 20.

In the years between 2013 and 2017 the company enjoying the greatest growth of its patent portfolio in the AI field is the State Grid Corporation of China (SGCC), whose portfolio grew by an average annual rate of 70%. In the same time period, the AI patent portfolios of several Chinese Universities and public research organisations have grown by an average annual rate ranging from more than 20% up to about 80%:

Although the first and second biggest patent applicants in the AI field are US companies (IBM and Microsoft), the biggest patent portfolios worldwide in several specific AI techniques, functional applications, and application fields belong to Chinese companies or institutions:

Notably, the state-owned State Grid Corporation of China (SGCC) has the biggest patent portfolio worldwide in the AI applications to one of the hottest topics of the world political agenda, namely energy management.

The worth of the Chinese market in the AI sector may also be inferred by the amount of patent applications filed to its patent office, the CNIPA. Starting from the year 2009, the number of patent applications in the AI field filed to the CNIPA has grown by an average annual rate of 25% and, as shown in the chart below, the Chinese patent office ranks second behind the US one in the number of patent filings in the AI sector.

In general, applicants may seek protection for their invention by filing patent applications to several different patent offices. Typically, the choice of the office for the first filing unfolds interesting information concerning the geographical origin and the IP strategy of the applicant. The Chinese patent office is the top one in the number of first filings of AI patent applications:

Since the vast majority of first filing in China are made by Chinese applicants, the prominence of the CNIPA as first filing office reflects the importance of the Chinese AI ecosystem. Around the year 2007, the Chinese patent office shifted from being an office of second filing for foreigner applicants interested in the Chinese market, to an office of first filing for Chinese applicants.

It is worth noticing that only about 4% of Chinese first filings are further filed in other jurisdictions, showing that Chinese applicants seem to be more interested in the domestic rather than foreign market. For example, in the years between 2011 and 2016, only 6% of the European patent applicationsreferring to fourth industrial revolution technologies were submitted to the European Patent Office (EPO) by Chinese applicants:

In this regard, once published, the majority of the Chinese patent applications may be a gold mine of technological innovations, that can be produced, used and sold outside the Chinese market and jurisdiction. Judging from the IP sector, it is thus fair to say that the question as to whether one should learn Chinese to advance in AI should be affirmatively answered.

GALLERY