The AI Boom Just Hit a Labor Shortage. Field Service Has Been Living There for Years.
The most advanced industry in human history just ran into the oldest problem in business: not enough skilled people.
NBC News reported on August 6 that AI data center construction is being choked by a labor shortage. By some estimates, the U.S. is 58,000 people short of the workforce needed to install the fiber-optic cable that connects data centers to the internet. Think about that. Companies are spending hundreds of billions of dollars on GPUs, and the bottleneck is a person with a spool of fiber and two trained hands.
Jensen Huang saw this coming. “If you’re an electrician, you’re a plumber, a carpenter — we’re going to need hundreds of thousands of them to build all of these factories,” he told Channel 4 News in late 2025. At Davos in January, he went further and said the AI buildout will unlock “a lot” of six-figure jobs in plumbing and construction.
The tech industry is treating this like breaking news. It isn’t. Field service, manufacturing, and equipment support have been living inside this exact shortage for a decade. Welcome to our world.
The shortage is structural, not cyclical
The numbers are brutal. Associated Builders and Contractors estimates the construction industry needs to attract 349,000 net new workers in 2026 just to meet demand, rising to 456,000 in 2027. CNN reported that meeting proposed data center buildout deadlines would require adding 500,000 electricians, 300,000 welders, and 550,000 plumbers, per the American Edge Project.
You cannot hire your way out of that. Training an electrician takes four to five years of apprenticeship. A senior field technician who can diagnose a failing industrial compressor by sound? That takes a career. The pipeline is measured in years, and the demand curve is measured in quarters.
Here is the part Silicon Valley finds uncomfortable. AI can write code, draft contracts, and generate images. It cannot pull fiber through a conduit. It cannot torque a bolt or terminate a cable. The physical world still runs on human hands, and the humans with the right hands are scarce and getting scarcer.
What field service learned the hard way
Industries that depend on skilled technicians hit this wall long before AI data centers did. Their experienced people started retiring faster than replacements arrived. Every departure took twenty years of tribal knowledge out the door.
The companies that adapted did not adapt by waiting for the training pipeline. They adapted by changing the math on where an expert has to physically be.
Your best technician can be in one place at a time. That is the constraint everyone accepts. But most of what makes that technician valuable is judgment, not hands. Judgment travels at the speed of light. Hands do not.
So the smart operators split the two. The scarce expert stays put and looks through the eyes of whoever is already on site: a junior tech, a customer, a facilities manager with a smartphone. The expert sees the equipment live, points at the exact valve or connector, and walks the on-site person through the fix. One senior expert stops handling four jobs a day and starts influencing twenty.
I wrote before about the $13.8 billion field service market still built around the assumption that driving there is faster than seeing it. That assumption was always expensive. In a labor shortage, it becomes indefensible.
The leverage play nobody is talking about
Everyone in the AI conversation is focused on the supply side: train more electricians, fund more apprenticeships, raise wages. Fine. Do all of it. But those are five-year answers to a this-quarter problem.
The faster answer is multiplying the reach of the experts you already employ. If your constraint is expert hours, you have two options. Add experts, which takes years. Or waste fewer expert hours on things that do not require an expert on site, which takes weeks.
Most site visits by senior people fall into a few buckets: verifying what the problem actually is, supervising work a less experienced person could do with guidance, and confirming a fix worked. All three can happen over live video. The senior person only travels when the job genuinely requires their hands, not just their eyes.
Data center construction crews will figure this out under pressure, the same way field service organizations did. A fiber-splicing specialist guiding three crews remotely beats one specialist driving between three sites. The physics of the shortage forces it.
The uncomfortable conclusion for AI companies
There is real irony here. The industry promising to automate knowledge work is bottlenecked by manual work it cannot automate. And its response so far, throwing money at training programs, is the same slow response every trades-dependent industry tried first.
The pattern that actually worked was leverage, not headcount. Stop treating your scarce experts as people who go places. Treat them as knowledge that gets deployed anywhere, instantly, through the camera of whoever is standing in front of the problem. That is what we build at Viewabo: remote visual assistance that lets one expert see and solve problems through any smartphone, no app install, no truck roll.
The AI boom did not create the skilled trades shortage. It just made it impossible to ignore. The companies that win the next five years will not be the ones with the biggest training budgets. They will be the ones who figured out how to be in twenty places with the experts they already have.