Historical Comparison Mapping: British Colonial Rule over India and China and AI Power Concentration

2026 02 23

Evolving historical comparison argument I have surrounding the emerging discussion on developing countries and middle powers

The closest historical allusion to a general and total technological advantage is arguably found in the British Industrial Revolution. The mechanization of textile production, the steam engine, and the expansion of rail and telegraph networks collectively restructured the material basis of global production. Britain not only outcompeted other states but also reorganized the terms on which economic activity was conducted. It is for this reason that the Industrial Revolution has become a common reference point in discussions of AI’s transformative potential: both represent a rupture of the productive order itself.

Many arguments employing this parallel, however, neglect what is most analytically valuable about it: the relationship between Britain and its colonial subjects. Britain’s industrial advantage was leveraged through a colonial institutional apparatus that restructured the political economy of dependent territories in Britain’s favor. Reflecting upon the mechanism of technological leverage, what sustained it structurally, and how it was ultimately undermined, offers a more rigorous basis for understanding AI-driven power concentration, as well as the limits of the analogy itself.

The mechanism of British technological leverage over India was not simply that British textiles were cheaper. It was that industrialization itself destroyed the material basis of a prior social formation. Clingingsmith and Williamson document that Indian textile exports collapsed from roughly a quarter of world output in 1750 to near-negligible by 1900—structurally analogous to the dissolution of comparative advantages in developing countries, namely labor-cost-driven services and data processing upon which many countries build their development strategies.

What sustained this leverage was not technological advantage alone but its co-constitution with hegemonic institutional power. The East India Company was simultaneously trade monopoly, military force, and proto-financial institution. Sterling, Lloyd’s, and the Bank of England formed the financial infrastructure through which exchange was mediated on British terms. Similarly, US dominance is inseparable from dollar-denominated cloud infrastructure and semiconductor supply chains concentrated among allied producers, arrangements that make sanctioning AI-adjacent transactions feasible. The British relationship with China further illustrates this dynamic: China lost functional economic sovereignty while retaining formal state sovereignty during the Opium Wars, and found in self-reliance an ideological solution to dependency.

What ultimately undermined this leverage is the more theoretically generative question. As Marx observed in his 1853 New York Tribune essays, British colonialism was simultaneously destructive of existing social formations and the unwilling agent of their supersession. The railways, built for efficient military movement and resource extraction, reduced the cost of pan-Indian communications by orders of magnitude, creating the organizational infrastructure for the Indian National Congress and, eventually, mass nationalism. The telegraph, initially a military instrument, became available to merchants, lawyers, and political organizers. Print capitalism in vernacular languages—enabled by diffused printing technology—created the communicative preconditions for national consciousness. The technology of domination contained within it the conditions for the colonial subjects’ eventual capacity to organize.

The limits of this analogy, however, deserve explicit acknowledgment. Railways required physical deployment in the colonized territory, which inherently transferred technology and created local employment. AI does not. A country may consume frontier AI services entirely through a browser without any corresponding capability building. Data extracted to train models, similarly, need not be returned. The diffusion mechanism is structurally weaker for AI than for industrial technology—precisely what makes AI concentration a qualitatively distinct governance problem. Whether AI will generate the same internal contradictions that eventually undermined British leverage, or whether the absence of diffusion mechanisms will sustain dependency, is, arguably, the central question motivating serious work on international AI governance.