Have Strategists Drunk the “AI Race” Kool-Aid?
Has global strategic competition become a race for dominance in artificial intelligence (AI) between the United States and China? Versions of this claim have become something of an axiom, offered by officialdom and the analytical community alike. That AI will be the primary axis of future strategic competition is contestable, however. Moreover, the notion of an AI race in and of itself will generate policy risk. Making policy based on those assumptions could lead to narrowing options, not only in the realm of competition between states but regarding human affairs in general.
When technological change is driven more by hubris and ideology than by scientific understanding, the institutions that traditionally moderate these forces, such as democratic oversight and the rule of law, can be eroded in pursuit of the next false dawn. We’ve seen this before. An AI race, specifically, could undermine the social fabric of the countries engaging in it, which makes it a problem of the state.
Is this a risk worth taking? Bringing these assumptions to light reveals a number of unresolved questions that make caution a better strategy than racing ahead. Given the geopolitical and societal costs of being mistaken, strategists should think again before putting the “AI race” at the center of their analysis without clearly demarcating its boundaries.
The United States and its allies need to think more holistically about the impact of AI competition on human affairs. They should be hyper-aware of the way in which myths about human progress find vehicles in ideology, and how ideology can masquerade as science. The narrative of an unbridled “AI race” focuses on AI’s military uses while ignoring the broader societal implications of rapid technological change. The digital age is already having unintended effects on nations and the people who make them up. Strategists should take care not to home in too narrowly on the strategic and military debate while overlooking the longer-term ways in which AI stands to disrupt society.
Why an AI Race?
During and shortly after World War II, the age of digital computation began in earnest. U.S. leadership in the digital age set the United States and its allies and partners (mostly Western with the notable exception of Japan) on a technological pathway that was intertwined with grand-strategic aims. The U.S. lead in the digital age is considered one of the primary competitive stratagems that overwhelmed the Soviet Union, and the United States attempted to double down on this lead after the Cold War.
The military-industrial complex of Eisenhower’s day gave way to what became, in Linda Weiss’ words, a “National Security State,” and, eventually, as Shane Harris has put it, the ascendant “military-internet complex.” The personal computer, the internet, the mobile device, and the “Age of Surveillance Capitalism” that these technologies spawned are all products of this doubling down. Now, these digital-age tools of disruption are available to almost anyone: state and nonstate actor; individual and corporate; malicious, altruistic, and benign.
Today, analysis of how technology could shape the future focuses on the Internet of Things, machine-to-machine communication, augmented reality, 5G, and predictive technologies. Artificial intelligence is at the vanguard of this matrix of technologies. Advanced AI chipsets and 5G promise much greater data speed, throughput, and lower latency. In theory, these technologies will unleash the next wave of digital transformation on society. Their acolytes in industry and government are certain that the next stage of the rapidly advancing digital age is the big one. It could be make-or-break now for the enormous strategic investment the United States and its allies have made since the digital age began in earnest in 1945.
Strategists rightly see an old and a new story here. It’s an old story about competition between states over critical resources and transit infrastructure, recast and updated for the digital age. The race for AI involves a multifaceted competitive struggle over input components such as intellectual property, investment and capital flows, semi-conductors and their composite materials, the physical infrastructure of cyberspace such as internet cables, satellites and ground stations, data centers, cloud servers, and supply chains. The AI race also involves a new vector of competitive struggle over the outputs of this burgeoning system, such as the strategic manipulation of socio-political-economic systems in service of corporate, financial, and bureaucratic ends, and new tactical-level military advances. This is where AI capabilities make a claim as something new, and states appear convinced of the need to compete here, while uncertainty about the boundaries between civil and military competition is pervasive.
China has sought to build and scale its own network of digital-age capabilities, eroding the massive lead the United States held in many of the input components such as infrastructure throughout the digital age. Russia is not a significant player here, so it has focused instead on pre-emptively disrupting the output space, the products of the hyper-connected age, before it has matured, hoping to reduce its status as a zone of strategic advantage for the West. Europe is extremely wary of where this is going, but has made significant investments at the input level around infrastructure.
Two AI Superpowers?
How has this strategic ground — digital technology leadership — slipped from under American feet with such apparently little resistance? A number of factors are influential, but the key lies in understanding that, for at least four decades, the United States assumed that China would eventually be folded into a U.S.-led international order that remained favorable to U.S. preferences. This assumption was driven by China’s reliance on the U.S. economy and the Chinese Communist Party’s reliance on economic growth for popular legitimacy — a weak axiom with a shallow grounding in truth. It now seems difficult to deny that China under Xi Jinping has no intention of accepting continued American stewardship — despite the economic risk that entails — of an international order it sees as China’s to shape. Beijing has long sought to re-assert Chinese civilizational power. China has recapitulated the surveillance capitalism model, invented by Silicon Valley, and charged ahead with its culturally specific version via its own digital giants Alibaba, Baidu, Ten Cent, and Huawei.
Western hopes that the rejection of Maoism after his death was a sign that China was on a path to Westernization were disappointed. To the contrary, the demise of Maoism marked a point of departure after which China’s leaders would never again attempt to follow the West’s path to development. Now, with a focused and determined AI strategy as a core component of President Xi’s “Chinese Dream,” China has deliberately scaled at an infrastructure level. Companies such as Huawei are global titans, which means disentangling the West from Beijing’s version of the digital age is no simple matter. In addition, China’s authoritarian system enables AI’s introduction to society with less friction (and less caution) than in the democratic systems of the West.
The West has belatedly begun reacting to this reality. The U.S. attempt to knock Huawei out of the 5G race is a notable move, as were moves to block various semiconductor-related acquisitions by China-backed companies recently. Huawei, however, is the world’s most vertically integrated 5G competitor. It has a significant presence across the digital stack from the smartphone, to the cellular network technology and submarine cables. Huawei also has its own chip-maker in HiSilicon, and some 66 countries have memorandums of understanding for research, testing, and development with the company.
A Broader View of Digital Competition
Shaking Huawei out of 5G completely will be impossible. It is also unnecessary. Strategists should look to network theory to understand where power lies in these new horizontal worlds and broaden our collective understanding of how to compete in the digital age. Primarily, the hub of power in this era is at the level of routing and switching, which should be a guide to where to focus competitive strategy and the regulatory response. Strategists should understand every layer of the digital stack, from the submarine cable to the mobile device, as a terrain through which power and influence flow in novel assemblages. It will be impossible to control these spaces completely.
In defense matters, AI’s military battlefield applications are potentially significant. But conventional military battlefield superiority has, at least for now, been superseded by society-centric conflict in which state-of-the-art military capabilities are not the deciding factor in continuous political hyper-competition between states. Human conflict will likely cycle back to traditional forms of violence, but for now, the digital age has given rise to a connectivity-enabled form of cognitive warfare to which open democratic systems are significantly vulnerable.
Populations, not soldiers, are now on the front lines. The political ends to which all warfare remains wedded can be accessed very differently, and digital saturation is the central factor in this access. Digital-age influence operations are not merely propaganda with new tools. Rather, they involve the manipulation of a transformed cognitive environment whose critical zone is the human-computer interface. Thus, the “AI revolution” involves a cluster of technologies whose implications cannot only be understood instrumentally. Technology changes people, too. AI’s most significant implications may be existential.
The West, for decades, did not recognize how the digital age could shape sociopolitical stability — something no algorithm can fix and something we continue to ignore at our own risk. By turning the human-computer interface into a site of cognitive manipulation, we opened up a Pandora’s box of uncertainty. This is the backdrop against which the claim about AI as the culminating point of 21st century strategic competition is being made, without acknowledging the profound destabilization of Western societies that the AI race is engendering. To be sure, AI development will not automatically undermine sociopolitical stability. But when we allow hubris and ideology about “winning” in AI to obfuscate the investment needed in existing sociopolitical institutions, we risk making it a zero-sum game. The practices of surveillance capitalism undermine society’s social fabric. To quote Mathias Döpfner, “You can also win yourself to death.”
It’s Always Human
All of this leads back to a more familiar story. Somewhere beneath the prevailing narrative of a high-tech struggle for 21st century dominance are human beings in social systems — to which all the usual caveats apply. We still have to live with scarcity, in groups among others who have similar needs and capacities but radically dissimilar beliefs, values, and goals. These unamendable conditions are the source of the intractable conflicts of human life, which are never merely instrumental but existential in nature, meaning they are also irreducibly complex.
Cracks and fissures in the digital age are already manifesting. People who talk and write about strategy should be wary. It still matters what you think the ends are before you can justify the means. Otherwise “winning” in AI has no meaning at all. When it comes to AI, the West needs to consider, first and foremost, what the consequences of competition are for society. As the technological nihilism which accompanies digital saturation increases, this question is up for grabs in the minds of populations everywhere. Strategists who simply assume distracted populations have already bought into this deal may be harboring the biggest blind spot of all.
The chief challenge for China and the United States will be in sheltering their societies from the unpredictable effects of the digital age. The flawed mindset of an AI race needs to be modified. Instead, countries should slow down AI development and democratize its underlying tenets so that populations get a sense of what they are buying into. It will be important to have a transparent public discussion about how widespread AI adoption will benefit society, and what scientifically verified evidence exists for these claims. A precautionary principle with clear understanding of the limits inherent in the lives of real human beings, not those of idealized cybernetic systems, is appropriate. Ultimately, the state is the backstop and the custodian of the social fabric. It’s the state’s job to encourage a democratic conversation if the market will not.
Dr. Zac Rogers, Ph.D. is a research lead at the newly established Jeff Bleich Centre for the US Alliance in Digital Technology, Security, and Governance at Flinders University of South Australia. Currently he is lead researcher in a three-year defense/academic collaborative project exploring the impact of digital transformation from infrastructure to the human/computer interface on Australia’s internal and external security, national interests, defense planning, and strategy.