On April 20, Meta announced the LevelUp Fiber Technician Pathway — a free, four-week training program to put thousands of Americans into skilled-trades careers building Meta's data centers. No experience required. Paid hourly during training. First cohorts begin this summer in Ohio and Indiana, operated in partnership with CBRE.
The same quarter Meta is announcing this program, the company is also cutting white-collar headcount.
That's the story. Same company. Same calendar quarter. Hiring with one hand for the construction site. Cutting with the other in the office.
The macro is unambiguous
Meta plans to spend $115B–$135B on infrastructure in 2026. They currently operate or are building 27 U.S. data centers. Their press release cites a 349,000-worker shortage in U.S. construction. The bottleneck on AI buildout is not chips — it's licensed humans who can pull fiber.
This is what makes the LevelUp program different from a standard workforce-development press release. Meta isn't training fiber technicians out of corporate goodwill. They're training fiber technicians because their $600B AI infrastructure roadmap through 2028 doesn't work without them.
What's at stake here
Meta is not the only one. Microsoft, Google, Amazon, and xAI are all in the same race. Year-to-date 2026 spending on U.S. data center construction has reached $36.9 billion — up from $1.4 billion in all of 2025. The labor pipeline is the binding constraint on the next decade of compute.
For thirty years, the cultural script told American kids the trades were a fallback. The largest tech companies in the world just spent the first quarter of 2026 telling them the opposite — with capital, with curriculum, and with paid hourly training the moment you walk in the door.
The signal underneath the news
Two truths now sit on top of each other:
The desk got dangerous. Stanford's Digital Economy Lab found a 16% relative employment decline for entry-level workers ages 22–25 in AI-exposed white-collar occupations since late 2022. Anthropic's CEO Dario Amodei has publicly warned that AI could eliminate 50% of entry-level white-collar jobs within five years. The data backs the warning.
The trades didn't. The BLS projects 9% growth for electricians through 2034 — three times the average for all U.S. occupations. Electricians earn a median $62,350 nationally, with the top 10% clearing $106,030. And the work cannot be offshored, automated, or replaced by a chatbot. In aviation maintenance, federal regulation makes that AI-resistance permanent — every aircraft in U.S. airspace must be signed off by a licensed human, by law.
Meta's announcement is not a feel-good story about workforce development. It is a publicly traded company telling the labor market — in $600B of capital expenditure — which careers are growing and which are being automated away.
The honest part
Fiber tech in a data center is not glamorous work. It is hot, repetitive, and physically demanding. You will spend stretches of your career inside the same square footage of concrete pulling cable through conduit. The romance of the job is in the paycheck and the durability of the demand, not the day-to-day.
That said: it is one of the few careers where the math is unambiguous. The work cannot be offshored. The infrastructure cannot be virtualized. The largest companies in the world have publicly committed to spending hundreds of billions of dollars over the next several years buying exactly this kind of labor.
If you are 22 and choosing a career, that is not noise. That is a signal pointed directly at you, paid for by the company that arguably created the white-collar economy you were told to chase.
The market just spoke. Read it before you sign your next student loan.