Meta is quietly testing robots inside its data centers, and at least one employee believes physical workers are no longer safe from automation. The experiment hints at a much larger shift in how AI spending could reshape labor costs across…
Artificial intelligence was supposed to make workers more productive. Instead, companies are increasingly discovering that the most productive worker can be one they don’t have to employ. U.S. technology companies have eliminated nearly 140,000 jobs in 2026, according to a Financial Times analysis, even as the industry’s AI spending reaches unprecedented levels.
Not every layoff can be blamed on AI — companies are also correcting pandemic-era excesses and cutting costs. But the direction of travel is difficult to miss. Now automation is moving beyond computer screens and into the physical world.
Meta’s Robots Are Coming for Data Center Jobs #
Meta Platforms (NASDAQ:META | META Price Prediction) is already spending billions to build the infrastructure needed for its AI ambitions. Now it is exploring whether robots can help operate that infrastructure with fewer people.
According to a report from WIRED, Meta is testing robots that can swap networking cables, power-cycle servers, reseat hardware, and perform other tasks traditionally handled by data center technicians. The company is testing equipment from Watney Robotics, Kinova, and ABB at facilities including Altoona, Iowa, and New Albany, Ohio.
One Meta worker told WIRED that a successful cable-swapping robot could eventually replace up to 80% of some technicians’ workloads, according to one data center worker:
“We thought those of us performing the physical tasks were safe for a while, but not anymore. It’s coming for us all, unfortunately.”
The robots remain slower than humans in some applications, require supervision and battery charging, and struggle with complicated cabling. Still, the experiment matters because it attacks a cost that is becoming enormous as Meta expands its data center footprint.
Meta’s second-quarter 2026 results show just how much money is moving into that infrastructure. Revenue rose 28% year over year to $60.8 billion, while capital expenditures reached $31.1 billion in the quarter. The company now expects $130 billion to $145 billion of capital expenditures in 2026.
AI Is Already Changing the Workforce. #
The Financial Times found that U.S. technology companies had gutted 140,000 jobs during 2026. Amazon (NASDAQ:AMZN), Oracle (NYSE:ORCL), Meta, and Microsoft (NASDAQ:MSFT) accounted for almost 50,000 of those reductions. But the FT also cautioned that AI isn’t responsible for every job eliminated. Companies are restructuring, correcting previous overhiring, and redirecting spending toward AI infrastructure.
That’s an important distinction for investors. The bullish case for AI has always been that productivity gains will create new industries and new jobs, just as previous technological revolutions did. That could still happen. AI-focused companies such as Anthropic and OpenAI are hiring, while Meta itself shifted thousands of employees toward AI-related work even as it reduced its overall workforce.
But AI has one unusual characteristic: speed and scalability. Software can replace tasks performed by thousands of people almost instantly once it works. Robotics could eventually do the same thing in the physical economy.
The $1.7 Trillion Opportunity #
Citizens Bank estimates that Tesla‘s (NASDAQ:TSLA) Optimus humanoid robot could eventually target roughly $1.7 trillion of U.S. wages. That’s not a forecast for Tesla’s revenue, and much of that potential market remains years away. But it illustrates why investors are paying attention to physical AI.
The economic incentive is straightforward. If a robot can work around the clock, doesn’t require benefits, and performs repetitive tasks consistently, its value isn’t merely that it replaces a worker. It changes the economics of the entire operation.
For Meta, that could eventually mean fewer technicians supporting vastly larger data centers. For Tesla, it could mean a new manufacturing and robotics business. For investors, it creates another way to measure AI’s potential: not just how much revenue it generates, but how much labor it can eliminate per dollar of capital invested.
Key Takeaway #
In short, investors shouldn’t interpret Meta’s robot experiments as proof that data center technicians are about to disappear. The technology isn’t there yet.
The more important signal is that AI automation is moving from white-collar software work to physical labor. Meta’s 2026 capital-spending budget shows how aggressively the company is building AI infrastructure, and robots could eventually help it operate that infrastructure more efficiently.
That makes the long-term thesis for Meta more compelling — but also changes the way investors should think about AI. The biggest gains may not come from selling AI products. They may come from using AI to require fewer humans to produce the same output.
That’s a much bigger productivity story. And, potentially, a much bigger margin story.
Contact [email protected] for any questions or corrections.