1 Recent progress in AI falls within two distinct fields: (1) ‘narrow’ AI, and particularly, machine learning; (2) ‘general’ AI, which refers to AI with the scale and fluidity akin to the human brain. Most AI researchers anticipate that ‘general’ AI to be at least several decades away. Narrow AI is already utilized in the private sector, in particular, in data-rich research fields and applied sciences. Most experts generally agree that the development of ‘general’ AI is at least several decades away, if at all.

2 The line between core AI and ‘AI-related’ technology is a blurred one. For the purposes of this study, core AI technology includes: machine-learning (and deep-learning and deep networks sub-set), modelling, automated language and image recognition, voice assistants, and analysis support systems; whereas ‘AI-related’ (and AI-enabling) technology includes: autonomous vehicles, big data analytics, 5G networks, supercomputers, smart vehicles, smart wearable devices, robotics, and the internet of things, to name a few.

3 There is an important difference, however, between narrowing the gap in certain fields, and gaining an overall lead across all the categories (i.e., talent, research, development, hardware, data, and adoption) in the emerging AI race. In a recent report published by the Center for Data Innovation: “Overall, the United States currently leads in AI, with China rapidly catching up, and the European Union behind both. The United States leads in four of the six categories of metrics this report examines (talent, research, development, and hardware),…[while] … China leads in two (adoption and data),” (Castro et al., 2019 Castro, D., McLaughlin, M., & Chivot, E. (2019). Who is winning the AI race: China, the EU, or the United States? Washington, D.C.: Center for Data Innovation. [Google Scholar], p. 2.). Chinese open-source data also confirms the trend that the U.S. is ahead of China in these categories (China Institute for Science and Technology Policy at Tsinghua University, July 2018).

4 Polarity analysis focuses on whether the inter-state order is dominated by one (a unipolar order), two (a bipolar order) or three or more (a multipolar order) centers of power. ‘Multipolarity’ in this context implies that no single state is unambiguously in the lead (or polar) in the international order. In contrast, to ‘bipolarity’ that implies much less ambiguity in the stratification of power surrounding two poles. In addition to military power, economic capacity, demographics, ‘soft power,’ and the broader social dimensions of state influence have been associated with the shift towards a multipolar order. See, William C. Wohlforth, “Unipolarity, status competition, and great power war,” in International relations theory and the consequences of unipolarity ed. John Ikenberry, pp. 33-65 (Cambridge University Press, 2011). For critiques on this contested concept see, Harrison R. Wagner, War and the State: The Theory of International Politics (The University of Michigan Press, 2009); and Randall L. Schweller, “Entropy and the Trajectory of World Politics: Why Polarity has become Less Meaningful,” Cambridge Review of International Affairs, 23:1, 2010, pp. 145–63.

5 The principal forces driving this evolution include: (1) the exponential growth in computing performance; (2) expanded datasets; (3) advances in the implementation of machine learning techniques and algorithms (especially in the field of deep neural networks); and above all, (4) the rapid expansion of commercial interest and investment in AI.

6 Examples for technically advanced small-medium powers who have actively invested in AI-related technologies include: South Korea, Singapore, France, and the U.K. Today, however, the U.S. and China lead in most metrics (i.e., talent, research, development, adoption, data, and hardware) used to rank states in the emerging race for AI innovation leadership (Castro et al., 2019 Castro, D., McLaughlin, M., & Chivot, E. (2019). Who is winning the AI race: China, the EU, or the United States? Washington, D.C.: Center for Data Innovation. [Google Scholar]).

7 China’s recent five-year plan reportedly committed over USD$100 billion to AI. Moreover, as China moves forward with its One Belt One Road-related projects that extend to potentially more than eighty countries, AI would become an integral part of these international infrastructure projects.

8 In quantum computing, for example, China has made significant efforts to integrate its quantum computing and AI research for boosting computer AI power and achieve ‘quantum supremacy’ - or the point at which a quantum computer is capable of outperforming a traditional computer. Chinese researchers have claimed to be on track to achieve ‘quantum supremacy’ as soon as 2019.

9 The economic gains that China may make through commercial applications such as BRI are not dependent upon dual-use technology or geopolitics alone; gains are also based on geoeconomics.

10 A distinction exists between the erosion of U.S. advantages in ancillary applications based on dual-use AI technologies, and in military-specific AI applications. Where the U.S. retains an unassailable edge in military capacity and innovation, the actual ‘threat’ posed to U.S. in the military-technological sphere is less immediate than in general-use AI. This implies the ‘threat’ narrative is more centered on perceptions of Beijing’s future intentions as its military-use AI matures.