Jensen Huang Says Plumbing Will Unlock Six-Figure Jobs. The Data Says He's Already Right
Co-Founder and Chief Data Scientist
At the World Economic Forum in Davos this January, Nvidia CEO Jensen Huang, the man whose chips power the AI boom, told BlackRock's Larry Fink that the next wave of six-figure careers won't come from a coding bootcamp. Framing the AI buildout as the largest infrastructure project in human history, Huang argued it will demand plumbers, electricians, steelworkers, and construction crews by the hundreds of thousands, in roles paying into six figures with no degree required. "You don't need to have a PhD in computer science," as he put it, to make a great living in this economy.
It's a striking claim from a tech CEO. It's also one the labor data was already confirming before he said it.
Why AI needs plumbers
The logic is physical. Before a data center processes a single AI prompt, it is a construction site: megawatts of electrical work, industrial cooling loops (that's the plumbing: liquid cooling is now standard for AI-class hardware), structural steel, and years of skilled labor. Nvidia itself committed $100 billion toward OpenAI's data center expansion, and McKinsey projects global data center capital spending could reach $7 trillion by 2030. McKinsey has also sized the labor gap: the U.S. alone needs roughly 130,000 additional trained electricians this decade, plus hundreds of thousands of construction workers and supervisors. These shortages predate AI and that the buildout is now compounding.
The scale per project is easy to underestimate. A single large data center can employ on the order of 1,500 construction workers during its buildout, and each permanent operational job spurs several more in the surrounding economy.
The paychecks are already here
This is where "will unlock" undersells it. Bureau of Labor Statistics data puts the median electrician at $62,350, a quarter above the median for all occupations, with the top of the trade well past $100,000. But medians describe the whole country; the data-center corridors describe the future. Reporting from boom regions documents electricians on data-center projects clearing six figures with overtime as a matter of routine, with accounts of top hands in Texas reaching the mid-$200,000s and being poached repeatedly between sites. Plumbers and pipefitters working mechanical and cooling systems on the same projects ride the same wage curve, and mechanical trades broadly track electrician pay.
Stack the full financial picture and the comparison with the traditional path sharpens: a tradesperson starts earning at 18 in a paid apprenticeship, reaches journeyman scale around the age a student graduates, and carries no debt, while the average new bachelor's holder starts four years later, in the low-to-mid $50,000s, servicing loans.
The honest caveats
Huang's optimism deserves two asterisks. First, boom wages are boom wages: the $200,000 stories come with heavy overtime, travel, and project cycles that can end. The durable claim isn't that every plumber will make $250,000. It's that structural shortage plus unprecedented demand has moved the whole wage distribution up, and licensed trades offer six-figure ceilings without degree debt. Second, the work is physical and the pipeline is slow: an apprenticeship takes four to five years, which is exactly why the shortage persists and why the wage pressure won't resolve quickly. Other executives, including auto CEOs pointing to hundreds of thousands of unfilled factory and construction jobs, have been sounding the same alarm from the demand side.
The takeaway for anyone advising a teenager
For decades, "learn to code" was the default advice and the trades were the fallback. Huang's Davos remarks capture the inversion underway: the AI economy's most secure entry-level jobs may be the ones AI physically cannot do (running conduit, welding pipe, balancing a cooling loop), because someone has to build the machines that automate everything else. The data says that shift isn't a forecast. It's a paycheck that's already being cashed.
