The „Tech War“ project continues apace. While we are busy coding policy documents (first Chinese, now American), we have compiled our theoretical approach, our research design and some initial results for a conference paper that we will present at the 2023 Swiss Network of International Studies Biennial Conference and the Open Conference of the IR Section of the German Association for Political Science next week. Here’s our conclusions so far.
The three „technopoles“ (China, the US, the EU) attach a great importance to AI and their own R&D capabilities. This is evidenced not just by a multitude of statements but by the sheer number of AI policy papers, regulations, investment programs etc. that these governments put out. Based on our research so far, we get the impression that while AI is definitively receiving a lot of attention, it is not fully securitized. In other words, the dreaded “AI Arms Race” so beloved of US political pundits is a misnomer. It is not at an arms race, neither in the classical IR definition of the term nor in a looser sense.
At the same time, AI is not a “normal” technology either – as Asaro sarcastically notes: “Just substituting the word ‘math’ for ‘AI’ in the recent headlines above helps to highlight why there may be hype around ‘AI arms races,’ but not ‘math arms races.’ The public clearly associates something more than the underlying mathematical techniques to AI” (Asaro 2019: 50). AI is attracting a lot of hype but even if most of it turns out to be smoke and mirrors, there is still a lot of potential for social, economic, and military innovation. In addition to this technological potential, AI is also imbued with status concerns. It seems that you simply cannot be a modern great power without having state-of-the-art AI capabilities.
These developments have a potential for conflict, in and of themselves. Even though AI development is in practice quite transnational and reliant on a free form of exchange among experts, political actors driven by narratives of geopolitical conflict and high-level economic competition may be pushed towards further hostile behavior. If conflict narratives become (even more) detached from reality, they might have serious consequences not just in the AI sector but for great power relations in general. At its most extreme, we can envision the three technopoles entrapping themselves in a security dilemma of their own making, which would be driven almost entirely by perceptions of threat with little material necessity.
It must said, however, that we consider such a dilemma situation to be very unlikely, and possibly even less likely than it was five years when the AI boom was beginning to unfold. On the one hand, it is our impression that while there are identifiable national approaches to AI, certain communities share similar concerns even across technopole boundaries. Given the tight coupling of discursive arenas and the close attention that each side pays to its “others”, we might see the emergence of transnational narrative coalitions. This is further supported by a gradual “normalization” of AI rhetoric. It is our impression that while tensions have mounted in US-Chinese relations, narrative constructions of AI have become slightly more down-to-earth as time has passed. In an optimistic reading, this “de-hyping” of AI may open spaces for more sober discussions about the global governance of this technology, about AI ethics and standards.
In short, we see interactions in the global AI field as subject to two countervailing trends. On the one hand, there is a normalization of AI, which is still considered a transformative technology but in somewhat less radical terms than it was only a few years ago (the current mini-hype about Generative AI notwithstanding). On the other, relations between these technopoles are fraught and still deterioriating. The US shows no signs of relenting on its hardline policy towards trade and scientific exchange with China. The Chinese government is busy tightening controls to solidify its increasingly authoritarian system of governance, including a “tech-lash” away from the formerly optimistic views on the potential AI. The EU, meanwhile, is trying to articulate a positive, more cooperative vision but is finding few takers among the other powers.