The rise of artificial intelligence has created an era where progress is measured in speed rather than reflection. Data centers, the cathedrals of raw compute, rise across continents, their hum a steady reminder of ambition and appetite. They demand staggering amounts of electricity, consume rivers of water for cooling, and strip minerals from fragile landscapes. Their presence is sold as inevitable, the infrastructure of progress. Yet behind the polished narratives lie questions that cannot be obfuscated. Who truly needs this scale of raw compute, and to what end? Who pays for it, and who inherits the consequences? How long can the Earth sustain this deliberate acceleration before the costs eclipse the gains? These are not engineering puzzles but moral dilemmas.
Automation, too, is framed as pragmatic. A system arrives, jobs are declared redundant, and balance sheets improve. But in reality, each such decision reorders lives, casting aside workers who discover that their value has been recalculated by machines. The usual refrain—“if I don’t do it, someone else will”—does not absolve responsibility. It only hides it. Responsibility cannot be outsourced, no matter how the paradigms of competition are invoked. The ethical truth remains: every act of replacement is also an act of judgment about whose work still matters.
The toll is not confined to work. Every teraflop of raw compute has a physical cost. Minerals are mined, grids are strained, landscapes are reshaped. The paradox grows clearer with each generation of chips: we design machines to predict floods, model droughts, and manage climate disasters, while simultaneously deepening the very conditions that make those disasters worse. To construct the future by consuming the foundations of its survival is not progress but contradiction.
