Insight from Project Flux’s James Garner
How the infrastructure race is rewriting the payslip for construction workers. The drywall contractor from Columbus, Ohio, probably didn’t expect to be managing 200 workers on a multimillion-pound construction site. But DeMond Chambliss isn’t alone in experiencing something akin to winning the career lottery. Across the country, construction workers are discovering that the infrastructure powering artificial intelligence pays far better than the work they left behind.
When tech giants race to build hundreds of new data centres to fuel their AI ambitions, someone has to do the physical work. And right now, that someone is being rewarded like never before in the construction trades.
The Numbers Don’t Lie
Workers transitioning from conventional construction to data centre builds are seeing pay increases of 25 to 30% compared to their previous roles. In some cases, the bump is far steeper. Chambliss now earns over £100,000 annually supervising construction at an Ohio data centre. In Oregon, electrical safety specialist Marc Benner pulls in roughly £220,000 a year. Electrician Andrew Mason, managing workers across six Northern Virginia data centres, exceeds £200,000 in annual earnings.
The baseline figures across the sector tell the story clearly. On typical construction projects tracked by Skillit, an AI-powered platform for trades hiring, the average salary sits around £62,000, or £29.80 per hour. On data centre projects using the same workforce, salaries jump to £81,800, or £39.33 per hour. That’s a near 32 % bump for the same workers doing more specialised work.
Yet these headline salaries barely scratch the surface. Companies building data centres are sweetening compensation packages with perks that traditional construction doesn’t offer. Heated break tents, free lunches, daily incentive bonuses, and in some cases, £100 per day in additional pay are becoming standard. Some sites even offer remote project management positions and stock options to retain the best workers.
The scale of investment backing these projects is staggering. According to research from the Pew Research Centre, data centres consumed 183 terawatt-hours of electricity in 2024, equivalent to more than 4 % of total US electricity consumption. That figure is projected to grow by 133 per cent to 426 terawatt-hours by 2030, demanding construction at a velocity few sectors have ever witnessed.
Why the Premium?
The intensity and precision demanded by data centre construction differ fundamentally from standard commercial work. These facilities demand specialised cooling systems, redundant power supplies, and earthquake-resistant designs that keep AI models running 24/7. The margin for error is thin. The timeline is ruthless.
Tech companies like Amazon, Google, and Microsoft aren’t price-conscious when delivering infrastructure that fuels their AI operations. The stakes are too high. As Raul Martynek, CEO of DataBank, told Fortune, data centre project budgets operate differently. “Because of the huge demand and the nature of this construction work, which is fueling the arms race of AI, the budgets are not as tight. I would say they’re a little more frothy.”
That frothy approach collides with a brutal labour shortage. The Associated Builders and Contractors trade group estimates the construction industry is short roughly 439,000 skilled workers nationwide. The shortage is particularly acute for the specialised trades required by data centre work: electrical specialists, HVAC technicians, and concrete experts who understand the unique demands of these facilities.
What’s often overlooked in data centre economics is the environmental context. Water consumption is emerging as a critical challenge. According to research by UC Riverside and the think tank Next 10, California’s data centres more than doubled their water use between 2019 and 2023, consuming 49.9 billion litres annually, equivalent to filling 20,000 Olympic-sized swimming pools. By 2028, that figure could rise to 116 billion litres. For construction professionals, this means data centre projects are increasingly complex infrastructure ventures that require environmental impact assessment, water system integration, and compliance with emerging regulatory frameworks. It’s not simply building boxes for servers; it’s engineering entire industrial ecosystems in water-stressed regions.
A Generational Shift
What’s perhaps most significant is what this boom signals about the future of work itself. For knowledge workers, AI has sparked widespread anxiety about automation and displacement. But for construction professionals, the opposite is true. Fraser Patterson, CEO of hiring platform Skillit, points to a notable trend: the data centre boom has convinced a growing number of Gen Z workers and recent college graduates to pursue careers in the construction trades.
“AI is creating a lot of job anxiety among knowledge workers,” Patterson observed. “Construction work is, by definition, very hard to automate.”
That confidence has begun to reshape career decisions at scale. We’ve been saying for some time now that the AI narrative is incomplete without acknowledging how infrastructure creates genuine economic opportunity outside knowledge work sectors. The data centre boom validates that thesis more convincingly than any theoretical argument could.
From our perspective, this dynamic challenges the prevailing narrative about AI and work. Rather than simplistically replacing human effort, AI’s infrastructure demands are creating unexpected winners among skilled trades, reshaping how young people think about career opportunities, and proving that genuine economic value still flows to those who build the physical systems that enable digital transformation. The professionals who understand both traditional construction expertise and the unique demands of data centre engineering will command premium compensation for years to come.
The Hidden Infrastructure Complexity
Beyond wages, the professionalisation of data centre construction is reshaping project delivery itself. These projects demand precision typical of semiconductor manufacturing, combined with the scale of heavy infrastructure. Electrical systems must handle the redundancy required by banking-grade operations. Cooling systems must function at efficiency levels that industrial refrigeration specialists typically work on. Structural engineering must account for weight loads and vibration that traditional office buildings never encounter.
This complexity means construction project managers are increasingly requiring technical literacy about power systems, thermal dynamics, and data infrastructure logic. It’s not enough to know how to manage concrete pours and workforce logistics. Understanding why a 1-degree variance in cooling efficiency might cost a technology company millions annually in operational expenses transforms how project teams approach problem-solving. The professionals who bridge construction expertise with this technical context become invaluable, and project teams that fail to understand these interdependencies often face cost overruns and schedule delays that traditional project management methodologies can’t explain.
The Long-Term Calculus
The obvious question: how long will this last? Google and Meta have committed to spending tens of billions on data centre infrastructure in the coming years. OpenAI is reportedly pursuing a network of data centres that could cost hundreds of billions. The infrastructure pipeline appears robust, but sustainability concerns are reshaping the timeline.
The environmental externalities of this boom, particularly water consumption in already water-stressed regions, are creating regulatory pressure that will reshape project timelines and budgets. Construction professionals working in states like California, Arizona, and Nevada face increasingly strict environmental compliance requirements, longer permitting processes, and potentially higher infrastructure costs as mitigation measures become mandatory. Contractors who build environmental resilience and water management expertise into their service offerings will have a competitive advantage.
Still, one thing is already clear. The notion that artificial intelligence creates only knowledge-work disruption is demonstrably false. It’s also creating opportunities in sectors that seemed mature, pushing wages upward in trades, and convincing a new generation that physical infrastructure work might be the better bet than software engineering after all.
For construction workers and small contractors willing to specialise, develop environmental and technical literacy, and navigate increasingly complex regulatory landscapes, the next decade might offer something rare: the chance to cash in on the most significant technology transformation of the century whilst building something tangible that millions of people ultimately depend on.
Stay updated on how AI is reshaping every layer of work. Subscribe to Project Flux for analysis built for project delivery professionals.
