The Infrastructure Play
Meta is boosting its spending commitment on a forthcoming AI data center in West Texas by more than sixfold to $10 billion, with an aim to reach 1 gigawatt of capacity by the time the facility comes online in 2028, the company said on Thursday. The data center being built in El Paso will lead to the creation of 300 new jobs, Meta said, with more than 4,000 construction workers required at its peak.
The new data center will be liquid-cooled, using a closed-loop system that recycles water. Meta projects the site's water use will be similar to a typical golf course in the region.
The Workforce Tension
Yet simultaneously: Meta is reportedly weighing layoffs that could impact at least 20% of its workforce as the tech giant looks to offset rising artificial intelligence costs. The cuts come as the technology company aims to offset the cost of artificial intelligence infrastructure and prepare for greater efficiency brought about by AI-assisted workers.
Meta is cutting approximately 200 jobs in the San Francisco Bay Area as the company restructures teams and invests heavily in AI infrastructure. The company's layoffs will affect 124 employees from its site in Burlingame, California, along with 74 in Sunnyvale. Those cuts are expected to take effect in late May.
The Narrative Disconnect
Meta is essentially saying: "We're investing $10B in compute infrastructure AND cutting 20% of human labor." This isn't contradictory if efficiency gains are real—but it signals that management expects significant productivity displacement from AI systems.
My Take: Meta's messaging is torn. The El Paso facility is a growth story (1GW capacity signals ambition). The 20% layoff threat is a contraction story. Truth: Meta is repositioning from people-first to compute-first operations. The company that once celebrated "move fast and break things" is now optimizing for capital efficiency. By 2027, Meta's org chart will look fundamentally different.