Data centers need electricians, plumbers and carpenters before algorithms can operate.
SANTA CLARA, United States | June 2026
Nvidia Chief Executive Jensen Huang believes the artificial intelligence boom will create extraordinary demand for electricians, plumbers, carpenters and other skilled tradespeople. While public debate often concentrates on office jobs threatened by automation, Huang argues that the physical infrastructure supporting AI cannot be built through software alone. Data centers, chip plants and advanced manufacturing facilities require large construction crews and specialized technical workers. His message challenges the assumption that the most promising careers of the next decade will exist only in programming or engineering.
Huang has estimated that hundreds of thousands of skilled workers will be needed to construct the factories and computing centers powering the new technological economy. These facilities require electrical networks, cooling systems, water infrastructure, structural work and continuous maintenance. Every server rack depends on cables, pipes, ventilation and physical installation completed by people working on site. The cloud may appear intangible to users, but its foundations are made of concrete, steel, electricity and water.
The scale of planned investment explains why demand could rise so sharply. Consulting estimates suggest global spending on data centers may reach several trillion dollars before the end of the decade. Technology companies are expanding capacity to train and operate increasingly powerful AI models, while governments are supporting domestic semiconductor production and digital infrastructure. Each project creates construction work long before the first computer begins processing information.
A large data center covering approximately 23,000 square meters can require as many as 1,500 workers during construction. Electricians must install high-capacity power systems, while plumbers and pipefitters build the cooling infrastructure needed to prevent servers from overheating. Carpenters, welders, equipment operators and construction technicians complete the surrounding structures. Many of these positions can offer salaries above $100,000 in the United States when overtime and specialized certifications are included.
The permanent workforce inside an operating data center is much smaller than the construction crew, but the economic effect continues after completion. Facilities require technicians, security personnel, maintenance teams and energy specialists. Those direct positions also generate employment through transportation, food services, housing and local suppliers. The AI economy therefore produces jobs outside the technology companies that own the computing equipment.
Huang’s argument does not mean every manual occupation is protected from automation. Robotics, prefabrication and AI-assisted planning will also change construction and industrial work. However, many physical tasks remain difficult to automate because they occur in unpredictable environments and require judgment, dexterity and adaptation. Repairing a pipe in an existing building is fundamentally different from repeating a controlled movement on a factory line.
Skilled trades also involve responsibility that cannot easily be transferred to an algorithm. Electricians must interpret safety codes, identify hidden faults and respond to conditions that differ between projects. Plumbers work with water pressure, drainage, heating and cooling systems that can cause serious damage when installed incorrectly. Carpenters must translate plans into precise physical structures while adjusting to materials and site conditions.
The growing shortage of qualified workers has already become a concern for business leaders. BlackRock Chief Executive Larry Fink has warned that insufficient skilled labor is slowing the construction of data centers. Ford Chief Executive Jim Farley has raised similar questions about efforts to bring manufacturing back to the United States. Industrial investment cannot succeed when companies have financing and technology but lack the people required to build and operate facilities.
The problem reflects decades in which many educational systems presented university as the preferred path to professional success. Vocational schools and apprenticeships often received less prestige, funding and public attention. At the same time, experienced tradespeople began retiring faster than younger workers entered their occupations. The result is a widening gap between infrastructure ambitions and available labor.
Part of Generation Z is beginning to reconsider that hierarchy. Some young people are choosing technical institutes, apprenticeships and trade schools because they offer faster entry into employment and lower educational debt. These paths can provide stable earnings without requiring a four-year university degree. Rising demand connected to AI, energy and advanced manufacturing may strengthen that shift.
Training remains essential because these are not occupations that can be mastered through brief online instruction. Electricians, plumbers and carpenters require supervised practice, safety education, licensing and knowledge of local regulations. Modern data centers also demand familiarity with high-voltage systems, industrial cooling and specialized materials. The future tradesperson may work with traditional tools while operating inside one of the world’s most advanced technological environments.
Governments and companies will need to expand apprenticeships if Huang’s forecast proves accurate. Scholarships, employer-supported training and partnerships with technical colleges could help prepare workers for large infrastructure projects. Programs must also attract women and communities historically underrepresented in the trades. Increasing labor supply should not come at the expense of safety, wages or professional standards.
Huang’s optimistic position should be understood alongside Nvidia’s commercial interest in accelerating AI infrastructure investment. The company sells the processors and systems used inside many of the facilities creating this demand. A larger construction cycle strengthens both Nvidia’s business and the surrounding industrial economy. Even so, the physical labor requirement exists independently of the company’s promotional message.
The prediction also complicates fears that AI will simply eliminate human employment. Automation may reduce demand for some administrative and entry-level digital tasks while increasing employment in construction, energy, maintenance and manufacturing. The transition will not necessarily benefit the same workers whose jobs are disrupted. Education and labor policy must therefore help people move toward occupations where demand is expanding.
AI will not exist only on screens or inside software applications. Its growth requires power plants, transmission lines, cooling systems, semiconductor factories and enormous buildings filled with computing equipment. Those systems must be constructed and maintained by skilled people. The workers shaping the AI era may be carrying toolboxes as often as laptops.
The future becomes real when human hands build its foundations. / El futuro se vuelve realidad cuando las manos humanas construyen sus cimientos.