It is generally recognized that the US, Europe and China are the three major forces in the global competition around Artificial Intelligence. As the US is considered the definite leader in terms of research and economic activities, China and Europe represent two different cases.
In particular, the European economy is developing AI based upon previous generations of digital development, while China is trying to leapfrog by leveraging AI. A comparison of the European Union and China provides a more comprehensive perspective of the increasingly ubiquitous technology.
As it is difficult to compare national statistics for AI, due to the fast-evolving nature of this technology and a sometimes baffling definition, this article aims to provide a qualitative analysis of the AI background and development in China and Europe, as well as the cross-border cooperation between the two economies.
To make sure the numbers are comparable, we used data from Crunchbase for both EU and Chinese analysis, and used data from IT Juzi, a Chinese business database, for adjustments.
Ambitious union-level and national strategies
The European Union Member States co-signed a Declaration of Cooperation on Artificial Intelligence in 2018, which discusses the cooperation on AI from multiple aspects, ranging from R&D and applications to related social, economic, ethical and legal questions.
After the cooperation announcement, the EU set a target (in a Communication on AI for Europe) to meet EUR 20 billion (USD 22.05 billion) per year over the next decades, combining the national and private sector forces of the EU, leading to greater investments.
Other supporting technologies are also linked to the AI agreement, including AI chips, 5G and high-performance computing – for which the bloc has budgeted EUR 9.2 billion (USD 10.14 billion), as indicated in the Digital Europe program.
AI policy seems to have come later than expected, but the EU launched a 10-year Human Brain Project as early as 2013, which is the most important human brain research project in the world.
China was the third economy to release a specific AI national policy, following the US and Canada. In China, the most marked AI policy is the 2017 "New Generation Artificial Intelligence Development Plan"(新一代人工智能发展规划) and the Action Plan following it, which formed the fundamentals of China's AI strategy.
The 2017 policy outlines three goals:
1) to build a CNY 150 billion (USD 22 billion) worth core AI industry and CNY 1 trillion (USD 150 billion) AI-related Industry, targeting at the level same as leading countries by 2020; 2) to build a CNY 400 billion (USD 60 billion) industry and CNY 5 trillion (USD 754 billion) AI-related industry achieving an important breakthrough in AI fundamental theory by 2025; 3) CNY 1 trillion (USD 150 billion) AI industry and CNY 10 trillion (USD 1.5 trillion) in AI-related industries becoming the global leading innovation power for AI by 2030.
Without a precise definition, it could be considered that the AI core technologies are the all-encompassing ones, including but not limited to neural science and quantum computing in future development, while AI-related industries refer to applications to specific industries.
The social and ethical concerns
In terms of AI, the Europeans are thinking beyond technology and industrial uptake. Two out of three objectives during the 2018 Communication are about more profound, long-term impacts. One is preparing for socio-economic changes impacted by AI in terms of the talent and labor market and the social protection system. Another is ensuring an appropriate ethical and legal framework to develop AI ethics guidelines.
Globalization: Data, talent and capital flows
Data – the blood of the AI economy
Given the relative scarcity of data due to the smaller populations of many countries, the EU is making efforts to share data to a larger group of AI players by building systems and sharing offline to online facilities. It has published policies to encourage better data sharing and guidelines in both business-to-business and business-to-government contexts.
For example, the statement on 'Building a European data economy' stated that the European Commission has implemented measures to make it easier for businesses and the public sector to access and re-use data. The regulation on the free flow of non-personal data is applicable this May. It ensures that: 1) every organization to be able to store and process data anywhere in the EU; 2) data is available for public authorities for regulatory control; 3) cloud service providers for professional users will be easier to switch; 4) cybersecurity will be supported.
According to the EU, the value of the European data market was EUR 6.16 million, accounting for 1.99% of the GDP, in 2016 and is predicted to reach EUR 10.43 million, 4% of GDP, by 2020.
Meanwhile, an abundance of data generated by the 1.4 billion population across various channels contributes to China's AI booming scene. According to the MIT Sloan Management Review, 78% of leading Chinese companies maintain their corporate data in centralized data lakes, compared with 37% of Europeans.
Because of certain internet firewalls, China has the opportunity to isolate part of data strictly for domestic use, and therefore creating its own substantial data pool. However, that move also limits cross-border cooperation in data sharing. Within the border, Chinese leading AI companies are selected to be the 'Next Generation National AI Platforms', required to provide and share AI services, including data.
In addition, it is commonly disputed that China's success in AI and connected areas is due to the low opportunity cost of acquiring private data. Fortunately, the problem is being solved, as a national standard on information privacy that even harder and more comprehensive than the European Union's General Data Protection Regulation has been released, indicated by CSIS, a Washington, D.C.-based is a think tank.
Competing for larger talent pools
As demand for AI grows, there is a continuing shortage of AI talent, both technical professionals who know mathematics, engineering or neuroscience and industrial experts who understand how to turn the impressive new tools into real products.
Both European countries and China are actively seizing oversea experts. Switzerland and Sweden have 50% and 49% of PhD talents from overseas respectively, according to a 2019 LinkedIn report. China is transferring talents mainly from the US and the UK. Chinese companies, institutions and state & local governments are attracting overseas researchers and expertise, by providing better research positions, funding and grants.
Collaboration is picking up
Straightforward technology partnerships from governments are uncommon, but there are efforts and progress coming from the private sector.
Collaborations between Chinese and overseas researchers are increasing, as indicated by an opinion article published in Nature. PhD candidates from cross-border labs co-developed algorithms and published them on open-source platform GitHub to global developers.
Besides, increasing AI-focused conferences provides channels for AI experts to communicate and exchange ideas, written by Sarah O'Meara on Nature.com. Tsinghua University AI researcher Jie Tang built a relationship with Vazirgiannis, who runs a data science group at the École Polytechnique in Paris.
China's tech companies are sharing data assets and injecting capital resources to Europe, Vazirgiannis said Tencent sends data to European institutions for research use without releasing the data. Huawei has set up over 20 R&D centers across EU. Ping An Insurance invested EUR 34 million to London-based fin-tech firm 10X Future Technologies.
Industry focus and the recent entrepreneurial milieu
Among the 28 EU members, from 2017 to October 2019, there were 1,140 investment activities among companies with the label of artificial intelligence. Excluding the 256 companies that did not disclose their funding amount, the 884 companies raised more than USD 7.7 billion, leading to an average total funding amount of USD 8.78 million for EU Artificial Intelligence companies.
16 out of 28 EU members implemented over 10 AI investment activities. The United Kingdom was the most active country, with total investment activities of 411, France with 141 and Germany with 136.
Chinese Artificial Intelligence startups participated in 345 investment activities, according to Crunchbase data, with the total amount more than USD 15.17 billion. Although the number of Chinese AI investment activities is significantly less than that of European countries combined, the total amount is significantly higher. This can be explained by the massively funded top AI companies.
Particularly, computer vision and facial recognition companies SenseTime, MEGVII and CloudWalk have raised significant amounts. Robotics specialists UBTech and Horizon are also two active entities that received a large sum.
On the contrary, European companies raised smaller amounts from venture capital and private equity. Babylon Health, a healthcare-focused company combining AI and diagnosis, got the largest equity funding of about USD 635 million. European AI uptake shows a preference for healthcare and auto, and more of them are enterprise services companies.