Across rural America, a new kind of giant has arrived – windowless, brick-like monoliths housing the world’s digital brains.  

Major companies like Meta, Amazon, and Tesla are acquiring large parcels of land, equivalent to the size of Manhattan, to develop massive AI data centers, each potentially costing more than $3 billion.

But their arrival brings serious challenges for families living in their shadows, as the centers guzzle massive amounts of water and electricity.

These unsightly buildings are also the poster children of a seismic shift in the economy that experts warn will lead to widespread job losses.  

The recent government data on private sector construction spending highlights the significant influence of AI’s rapid expansion on office employment.

When ChatGPT launched in November 2022, investment in office space to house workers was five times greater than spending on data centers. 

Today, as the graph below shows, the two are nearly neck and neck. 

This development is not accidental. Automation driven by AI is already beginning to replace jobs, as evidenced by recent layoffs at companies including Amazon, Microsoft, and Procter & Gamble, raising concerns that this trend may speed up.

On August 7, Open AI’s CEO Sam Altman posted this image of the Death Star from Star Wars ahead of the release of Chat GPT 5, the latest and most powerful chat bot from the Silicon Valley behemoth

For months, Wall Street executives have openly discussed plans to replace white-collar Americans with cost-cutting AI robots. 

Meanwhile, a report out last week showed layoffs in July had risen 140 percent from a year ago – with AI cited as a major factor.

The turn to AI is also sucking up resources from local communities. 

‘It feels like we’re fighting an unwinnable battle that we didn’t sign up for,’ Beverly Morris, 71, a resident in Newton County, Georgia, told The New York Times. 

In 2018, Facebook-owner Meta built a data center a thousand feet from her home. Since then, the Morris family has struggled to get enough water from their well to use the toilets, sinks, or dishwasher. 

When water does reach their home, it often leaves behind a brown residue. Morris received a $25,000 quote to fix the issue but can’t afford the repairs. 

‘I’m scared to drink our own water,’ she said. 

Meta officials claim their water use isn’t affecting Morris’s property, citing a local water analysis they conducted. 

Mark Zuckerberg compared his latest data center to the size of Manhattan in a recent Threads post

But local authorities are raising alarms: Newton County, with a population just over 100,000, is expected to run completely dry by 2030. 

The Georgia community is not alone. Thirsty data centers – which are training on data from American consumers – are sucking up municipal water, leaving a wake of problems for residents in drought-prone areas. 

For example, millions of Texans have been advised to take shorter showers as parts of the state face a shortage.

Meanwhile, multiple data centers, including a San-Antonio-based facility run by Microsoft and the US Army Corps churned through 463 million gallons of water in 2023 and 2024. 

Concerns are also growing that the energy grid is not prepared for AI’s gargantuan power needs. 

Studies by the International Energy Agency showed that the centers alone are expected to burn through 945 terawatt-hours of energy each year by 2030. 

That’s more energy than the country of Japan uses in a full year. 

‘AI is one of the biggest stories in the energy world today,’ Fatih Birol, the agency’s executive director said.  

Gigantic data centers, storing the well-invested AI computers, are popping up across the US

In cities that have already been experiencing water issues, like this water break in Texas, centers are exacerbating local droughts

Despite questions around resource dependence, major companies are pouring money into even bigger facilities with even more job-displacing capabilities.  

Meta, which plans to spend $72 billion on AI in 2025, is building two massive data centers in the US, named Prometheus and Hyperion. 

Mark Zuckerberg, Meta’s CEO, compared the scale of Hyperion to New York City’s 14,600-acre Manhattan borough in a post on Threads. 

Its nearly the same size. 

‘We’re building multiple more titan clusters as well,’ the social media boss said on his platform. 

‘Just one of these covers a significant part of the footprint of Manhattan.’