I recently flew into Mumbai for a closed-door summit on the future of generative artificial intelligence, eager to see how Prime Minister Narendra Modi’s administration is racing to close the compute gap. What I found just outside the city was nothing short of breathtaking: a sprawling, windowless, gigawatt-scale data center rising from the coastal plains like a beautiful monument to human obsolescence. It was a temple of progress.
Yet, the moment I stepped out of my air-conditioned luxury sedan, I was confronted by a group of local farmers protesting the facility. Their grievance? The megaproject was apparently draining their municipal reservoirs, dropping the water table, and leaving their taps completely dry.
Let’s be honest: we have to look at the bigger picture here. A single generative AI query takes roughly ten times the electricity of a standard Google search, and keeping thousands of NVIDIA GPUs from melting into slag requires millions of gallons of chilled water every single day. The locals argue that they need this water for "agriculture" and "basic human survival." But if we take a step back and examine the return on investment, human hydration is simply a grossly inefficient use of a scarce resource.
When a coastal villager drinks a liter of groundwater, that water is locked up in a biological system that produces zero shareholder value. But when that same liter of water is pumped through the cooling towers of a hyper-scale server farm, it allows a marketing executive in Chicago to auto-generate a six-paragraph email about synergies in under three seconds. We cannot let narrow-minded local interests stand in the way of that kind of global interconnectedness.
We are currently extracting three million gallons of groundwater a day to ensure our enterprise clients can generate high-resolution images of the Pope in a puffer jacket with absolutely zero latency.
The short-sightedness extends to the electrical grid, as well. Critics have been loudly complaining that the region is now suffering from rolling, fourteen-hour blackouts because the data center is pulling more wattage than the surrounding three districts combined.
I find this framing incredibly negative. We live in an era where wellness influencers pay thousands of dollars for "digital detox" retreats in the desert, voluntarily giving up their screens to reconnect with nature. The AI industry is providing this exact same mindfulness service to the working-class residents of coastal India, entirely free of charge. While the servers hum away, training the next generation of large language models, the locals are being given the invaluable gift of uninterrupted, un-electrified family time in the sweltering dark.
Then there is the loudest, most tiresome complaint of all: the lack of jobs. Activists love to point out that once the concrete is poured and the servers are racked, a massive AI data center only requires a handful of security guards and a few specialized technicians to operate. They claim the megaproject extracted their resources without providing any long-term employment for the displaced community.
Our entire municipality has no power for most of the day, our lentil crops are turning to dust, and the facility employs exactly six people who commute in from a different state.
I actually thanked Ms. Deshmukh for highlighting this, because it perfectly illustrates the triumph of the modern tech economy. The sheer efficiency of running a billion-dollar extraction operation with only six employees is a miracle of low-overhead management. Why does everything have to be a "job"? The residents should feel a deep sense of civic pride knowing that their ancestral lands are hosting the very algorithms that will soon automate away the rest of the global middle class.
India is moving fast to build the infrastructure of tomorrow. If a few coastal cities have to go thirsty so the rest of us can enjoy slightly better automated customer service chatbots, that is a sacrifice I am perfectly comfortable making on their behalf.