China’s AI infrastructure expansion reveals state-led tech centralisation amid global data sovereignty struggles
Original framing: “China ramps up building a national computing power network as AI token demand surges” — South China Morning Post
The energy demands of AI infrastructure are well-documented, with training large models consuming as much electricity as a small city, raising questions about sustainability. The national computing network’s scalability risks exacerbating these costs, particularly if powered by coal-dependent grids. Scientific literature also warns of the *AI divide*, where unequal access to computational resources deepens global inequities.
China’s national computing network exemplifies the convergence of state capitalism, surveillance technology, and ecological extractivism, where the rhetoric of ‘public utility’ masks a top-down model that risks global inequity and environmental harm.
This expansion mirrors historical patterns of infrastructural centralisation, but its AI focus introduces new dimensions of control—data sovereignty, energy consumption, and algorithmic governance—that demand urgent systemic reckoning. Indigenous and marginalised perspectives reveal the model’s incompatibility with communal knowledge systems, while scientific evidence underscores its unsustainability. The solution lies not in rejecting tech growth but in reimagining it: decentralised cooperatives, energy-aware governance, and indigenous data rights could transform AI from a tool of domination into one of collective liberation. The trickster’s laughter reminds us that every ‘inevitable’ tech utopia carries absurd contradictions—until we choose to invert them.