Are China’s ambitions to “lead the world” in artificial intelligence (AI) by 2030 credible? China’s rapid emergence as an AI powerhouse is often hyped and sensationalized, variously provoking alarm and enthusiasm that can sometimes overshadow the reality of real progress. At the same time, critical challenges remain in China’s quest to become “the world’s premier AI innovation center” and build up an AI industry of 1 trillion RMB (about $150 billion) in the process.
In China’s “rise” in AI, the active efforts of private enterprises have predated more recent policy support. However, since the State Council released the New Generation AI Development Plan in July 2017, there have been a number of indicators that its implementation is advancing throughout all levels of government. Although the future trajectory of its AI revolution remains to be seen, China is rapidly building momentum to harness the power of state support and the dynamism of private enterprises in a new model of innovation.
AI has only recently become a clear priority for Chinese leaders under the aegis of an agenda to transform China into a “nation of innovation.” For the 13th Five-Year Plan timeframe (2016-2020), China’s ambitions to transform itself into a superpower in science and technology are clear. In August 2016, the 13th Five-Year National Science and Technology Innovation Plan called for China to seize the “high ground” in international scientific development, launching a series of 15 “Science and Technology Innovation 2030 Megaprojects” that included both big data and intelligent manufacturing and robotics. At that point, AI was not explicitly included as a priority at that level, despite being mentioned in that and included in prior plans, such as the “Internet Plus” Artificial Intelligence Three-Year Action Implementation Plan, released in May 2016. Not until May 2017 did the Ministry of Science and Technology announce the decision to add “AI 2.0” to that initial lineup as a 16th megaproject.
The initial impetus for the development of a national AI strategy may very well have come from the private sector, which has pioneered China’s AI revolution to date. Baidu, in particular, has actively pursued an “AI first” agenda since launching the Institute for Deep Learning in 2013 and then establishing the Silicon Valley AI Lab in 2014. Perhaps of note, in 2015, Robin Li (Li Yanhong), Baidu’s CEO, in his capacity as a delegate to the Chinese People’s Political Consultative Conference, proposed the creation of a “China Brain” Plan that would devote extensive state investment to AI, even welcoming military funding for such an initiative. In particular, Li called for the government to “support capable companies in building an open platform offering AI-related basic resources and public services.” Coincidentally (or perhaps not), the plan that has since emerged does resemble his initial proposal, and major tech companies like Baidu, Alibaba, and Tencent (“BAT”) may have been quite actively involved in advising its formulation.
The decision to develop this AI plan appears to have been catalyzed in part by AlphaGo’s triumph over Go world champion Lee Sedol in March 2016, which has been characterized as a “Sputnik moment” for China. This feat occurred at least a decade earlier than experts had anticipated AI could master Go, given the game’s complexity. Such a notable advance highlighted the sophistication of U.S. and Western AI, whereas by contrast Chinese AI had achieved fewer cutting-edge advances at that. Against the backdrop of the U.S. AI plans and strategies released in mid- and late- 2016 under the Obama administration, AlphaGo was seen as another indication of the U.S. advancement disruptive technologies that could place China at a disadvantage. The cultural resonance of the game of Go may also account for the intense interest and attention that this event seems to have received from Chinese leaders. As of July 2016, central authorities had formally approved the drafting of a new AI plan, building upon prior research on AI strategy led by Chinese Academy of Engineering academician and AI expert Pan Yunhe.
The plan has acted as an impetus for new energy and motion across China’s science and technology bureaucracies over the six or so months since its release. For instance, in August 2017, the National Natural Science Foundation of China released Guidelines on AI Basic Research Urgent Management Projects, identifying a series of research priorities to receive millions in new funding, including new brain-inspired computing architectures and methods and man-machine cooperative hybrid intelligence. In October 2017, the National Development and Reform Commission (NDRC) has announced a parallel AI Innovation and Development Megaproject, highlighting priorities that included advances in deep learning AI chips and highly reliable intelligent unmanned systems and service robots. In particular, the NDRC will fund a series of new AI projects, with a focus on AI chips, cloud services, and open-source platforms, among others.
As of November 2017, the Ministry of Science and Technology (MoST) convened a high-level meeting that marked the official launch of the plan, standing up the New Generation AI Development Plan Promotion Office. This will be a whole-of-government endeavor involving no fewer than 15 different entities, with MoST, NDRC, and the Ministry of Industry and Information Technology (MIIT) taking the lead. In addition, the official involvement of the the Central Military-Civil Fusion Development Commission Office, the Central Military Commission (CMC) Science and Technology Commission, and the CMC Equipment Development Department confirms the inclusion of a focus on military applications of AI within this broader national agenda. In support of the plan, the New Generation AI Strategic Advisory Commission was also created at that time, convening senior academicians and experts from prominent private sector players, including Baidu, Alibaba, Tencent, iFlytek, and Horizon Robotics.
In December 2017, the MIIT released the Three-Year Action Plan to Promote the Development of New-Generation Artificial Intelligence Industry (2018-2020). This new plan calls for China to achieve “major breakthroughs in a series of landmark AI products” and “establish international competitive advantage” by 2020. In particular, it builds upon the prior plan to concentrate with greater specificity on objectives that will support the development of a world-leading AI industry. In the process, China intends to enhance such “core competencies” as the production of intelligent sensors and neural network chips. The new plan recognizes the importance of an AI industry “support system” to include a data resource base with standard test data sets, cloud-based training frameworks, and initial test and evaluation systems. (In this context, the availability of massive amounts of data, a natural feature of China’s information ecosystem, could be an advantage bolstered through policy, yet not necessarily decisive, particularly as more advanced algorithms become less dependent on big data.) In addition, MIIT’s plan reaffirms China’s commitment to accelerating the development of 5G networks that can enable China’s national “intelligentization.” In its entirety, this latest policy framework thus highlights the basic foundations of an ecosystem that could create favorable environment for AI development, bolstered through high levels of funding and a focus on cultivating an AI talent pool.
In parallel to these efforts at the national level, cities throughout China have started to develop and release their own plans and policies for AI, including Beijing, Shanghai, Hangzhou, Zhejiang, and Tianjin to date. Notably, Beijing plans to build a 13.8 billion RMB ($2.12 billion) AI development park that could host up to 400 AI enterprises. Zhongguancun Development Group, responsible for the project’s advancement, also intends to create a “national-level” AI laboratory within the new park. It is clear that Beijing is actively seeking to emerge as a national and global leader in AI. As of January 2018, the Beijing Frontier International AI Research Institute has been established, under the leadership of Kai-Fu Lee of Sinovation Ventures, a firm that has been actively investing in AI within the United States and China. The institute will include three centers, focused variously on basic research in AI, smart societal innovation, and AI patents. It will also construct a service platform for AI computing and data applications, which might enable economies of scale and lower barriers to entry. Lee has highlighted the particular importance of building Beijing up into a “talent center” and integrating the efforts of industry, academia, and researchers. At the same time, Shanghai plans to establish a special fund to invest in AI development, and Hangzhou has launched its own AI park, along with a fund that will invest 10 billion RMB ($1.59 billion) in it. In total, as many as 19 different cities and provinces may be engaged in the development of AI policy, so this is just the start.
As China throws the support and resources of the state behind AI development, major Chinese technology companies will remain integral players in this endeavor. Characterized as national champions in this domain, several leading Chinese AI companies, acting as the “national team,” will undertake the development of new “open innovation platforms” in AI, with Baidu responsible for autonomous vehicles, Alibaba Cloud (Aliyun) for smart cities, Tencent for medical imaging, and iFlytek for smart voice. Baidu is leading China’s National Engineering Laboratory for Deep Learning Technologies, established in March 2017, which will pursue next-generation research in deep learning. Baidu will also contribute to the National Engineering Laboratory for Brain-Inspired Intelligence Technology and Applications, established in May 2017, which aims to develop AI technologies that learn from the mechanisms of the human brain and to promote brain-inspired neural chips and brain-inspired intelligent robotics. Meanwhile, Alibaba’s Aliyun is working with the new National Engineering Laboratory of Big Data Systems and Software, led by Tsinghua University. This direct involvement of such private companies in national laboratories that may pursue dual-use technologies and applications reflects their deep entanglement with the overall agenda of the party-state, at a time when the “party-corporate complex” is also deepening.
It is perhaps too early to say whether this rapid ramping up of state support, in close partnership with key AI enterprises, will prove successful or result in inefficiencies in or even impediments to AI development. The high levels of investment and growing amounts of government funding could be advantageous in supporting long-term basic research and development, yet there are also concerns that an AI bubble might arise in the process. The creation of new parks and laboratories could create dynamic innovation ecosystems — but might suffer due to the shortages of talent that are the primary bottleneck to AI development in the short term. Indeed, China’s ability to educate and attract top AI talent from across the globe may be the key determinant of its ability to succeed in this domain. (For the time being, the active poaching and recruitment of top talent from Silicon Valley, along with the establishment of joint laboratories and research partnerships, may serve as a stopgap measure to advance indigenous innovation.)
Concurrently, as China places big bets on AI chip development, seeking to catch up and challenge NVIDIA, there are a number of early-stage efforts that appear promising, including those of Thinker, Deephi, Cambricon, and Horizon Robotics. However, NVIDIA’s dominance may be difficult to undermine, and China’s brute force attempts in semiconductors have, to date, demonstrated the limitations of such an approach. Increasingly, however, Chinese AI research is evidently becoming more prominent at the top international conferences, and China could achieve a human capital — and perhaps even hardware — advantage in the long term.
Going forward, it is clear that China will be at the forefront of the global AI revolution, though uncertainties remain about its future trajectory and prospects to realize its ambitions. Certainly, AI could transform society and the economy in China in positive ways, from education to healthcare. So far, there have tended to be higher levels of optimism and enthusiasm about AI in China, which may leapfrog the rest of the world in many of these applications. For instance, plans for hundreds of smart cities, such as Xiong’an New Area outside of Beijing, could result in futuristic metropolises in which 5G, AI, big data, the Internet of Things, and cloud computing are pervasively integrated into urban development to enhance energy, transport, and overall quality of life.
However, at the same time, it can be difficult to disentangle this expansive AI agenda from the Chinese Communist Party’s priorities and attempts to assure state security through bolstering its capacity for social control. Indeed, the creation of smart cities is linked to and will enhance social management. Unsurprisingly, the PLA, too, seeks to take advantage of rapid advances in AI to pursue a range of military applications that might enhance its future capabilities. For better and worse, China’s trajectory in AI will be transformative, within its borders and perhaps worldwide.
Elsa Kania is an Adjunct Fellow with the Technology and National Security Program at the Center for a New American Security, where her research focuses on Chinese defense innovation and emerging technologies, particularly artificial intelligence. Elsa is an analyst, a consultant, and a co-founder of the China Cyber and Intelligence Studies Institute.