The Cloud Covering Vancouver
While some celebrate, others fear it may already be casting a shadow
As published by The Wheel
In the language of politicians and executives, the arrival of data centres in Vancouver will be a step towards innovation, growth and the inevitable future of a digital society. Warehouses filled with servers are presented as monuments to progress, and symbols of our entry into the lucrative world of artificial intelligence and cloud computing!
However, below the polished rhetoric is a far more troubling reality, data centres consuming incredible amounts of energy, creating effectively no jobs and accelerating an economic transformation that increasingly sidelines actual workers and their labour.
They promise millions in capital expenditures and the prestige of participation in the AI economy, but unlike the factories of earlier industrial eras, modern data centres employ few people…
The proposals for new data infrastructure have been multiplying, with vast facilities being planned right in the heart of our city, drawn by access to inexpensive, clean energy and market access. These projects appear attractive to many, especially our current government which is particularly keen to find investment. They promise millions in capital expenditures and the prestige of participation in the AI economy, but unlike the factories of earlier industrial eras, modern data centres employ few people once construction is complete, with a facility costing hundreds of millions of dollars ultimately providing only a few dozen permanent jobs.
This reveals something fundamental about the political economy of surrounding AI. Marx observed that capitalism constantly seeks to reduce its reliance on labour in pursuit of greater efficiency and profit. AI represents perhaps the purest expression of this tendency yet devised. The enormous server farms now spreading across the continent are not just infrastructure, they’re machines designed to automate human cognition itself, and ultimately replace intellectual labour!
When it comes to the training of large language models and image generators, massive clusters of processors have to operate continuously, while cooling systems consume additional energy and water.
The environmental cost of this transformation is moreover immense. AI requires extraordinary computational power, and this power consumes electricity on a scale which is difficult to comprehend. When it comes to the training of large language models and image generators, massive clusters of processors have to operate continuously, while cooling systems consume additional energy and water. In places like BC where hydroelectric power is often framed as “clean,” the rapid expansion of data centres and AI have been placing new pressures on the electrical grid.
We’ll be needing more dams, transmission lines and industrial development, just to sustain an industry which aims for speculative financial value rather than present human needs, many of which are actively being undermined.
The industries most immediately threatened by AI are those traditionally associated with creativity and intellectual labour. Journalism, translation, illustration, film and music are all being reshaped by automation. Media corporations increasingly experiment with AI generated news summaries and synthetic content in order to reduce labour costs. The result isn’t just unemployment but the degradation of cultural production itself.
Journalism produced by an algorithm is optimized for speed and engagement rather than investigation, public accountability or actual insight. In other words, coverage that can’t be trusted.
The gains of such progress have on the other hand been largely privatized, while the social costs are generally redistributed downwards. The workers displaced by AI will not be granted freedom.
The irony is difficult to ignore. Technological optimists have promised for decades if not centuries that automation would liberate humanity from hardship, creating more space for creativity and leisure. The gains of such progress have on the other hand been largely privatized, while the social costs are generally redistributed downwards. The workers displaced by AI will not be granted freedom. They’ll quite simply be rendered economically precarious. The owners of such infrastructure will however consolidate unprecedented power over communication, labour and information.
The growing data centre sector then reflects a broader global transformation. Vancouver, like many similar cities, is becoming physically integrated into an economic system where immense quantities of electricity, land, and capital are devoted to replacing human work, while enriching a narrow elite.
Politicians celebrate this as modernization, but modernization for who? The question isn’t whether technology should exist but who controls it, who benefits from it and who bears the cost of it?
Without a fundamentally different economic approach, AI risks reproducing the oldest pattern of industrial capitalism, socialized sacrifice paired with a concentration of wealth. The cloud, it turns out, isn’t weightless after all.