AI's physical infrastructure as an anchor to curb youth flight in New Mexico

By Zendoric · June 28, 2026.

There is a paradox in the public conversation about artificial intelligence: it speaks almost exclusively of models, algorithms and startups, but rarely of the tons of steel, kilometers of cable and millions of cubic meters of water that make all of it work. An opinion piece published in the Las Cruces Sun-News by Darrell Deaguero, business manager of the Laborers' International Union of North America Local 16 (LIUNA Local 16) in Albuquerque, puts the focus exactly there: on the physical —and human— dimension of the technological revolution.

The text opens with a concrete story: a young man from Doña Ana County who announces his intention to leave New Mexico because he sees no future there. It is not an isolated case. According to Deaguero, New Mexico was one of only five U.S. states that recorded a net population loss last year, a statistic that turns talent retention into an economic and demographic urgency rather than an abstract debate.

The author's central argument is that projects like Project Jupiter —part of the AI infrastructure investment cycle that includes data centers, power-generation facilities and computing campuses— represent a window of opportunity comparable to the one once opened by highways or railroads. Not for software engineers, but for electricians, welders, machinery operators and construction workers who can build, maintain and expand that infrastructure without leaving their community.

The piece is not just a pro-labor plea: it articulates a concrete model of development. Registered apprenticeships allow a young person to earn a wage while learning a skilled trade under the supervision of experienced workers, obtaining recognized credentials without taking on student loans. It is a structural alternative to the four-year university which, paradoxically, remains invisible in much of the U.S. debate on social mobility.

From an analytical perspective, what Deaguero describes is not new in historical terms —major infrastructure projects have always acted as generators of skilled local employment— but it is pertinent in the current context. As sector context, the race to build computing capacity for large-scale AI models is driving an unprecedented wave of data-center investment: industry analysts estimate that demand for electrical power for these facilities will grow steadily for at least a decade. That demand is not met with chips and cooling alone: it requires labor specialized in construction, electrical installations and water management, sectors where the shortage of workers is already a bottleneck.

New Mexico, moreover, offers some non-trivial regulatory and environmental advantages that the article explicitly mentions. The state's Energy Transition Act charts a roadmap toward carbon-free energy sources by 2045, which means the electricity that powers these data centers will become progressively cleaner, a relevant fact for technology companies with public sustainability commitments. Water management —a critical resource in the arid American Southwest— is regulated with accountability requirements for large consumers, with environmental agencies overseeing quality standards that protect aquifers and rivers. In a sector where data centers' water consumption is a growing criticism, being able to demonstrate that operations are carried out within a rigorous regulatory environment is an asset.

The text does not dodge the political dimension: Deaguero explicitly calls on state and local leaders to demand solid labor standards, apprenticeship programs, environmental responsibility and tangible community benefits as a condition for welcoming the investment. It is the expected position of a union leader, but it also reflects a real tension that recurs across all regions competing to attract this kind of infrastructure: the difference between a project that extracts value from a community and one that distributes it depends largely on the conditions negotiated before the excavators arrive.

The underlying reading is that AI, as a general-purpose technology, is creating a massive demand for physical infrastructure that can become a lever for regional development if managed with judgment. New Mexico does not compete in the market for language-model engineers or machine-learning researchers; it can compete —and by this argument, should do so actively— in the market for infrastructure siting: a predictable regulatory climate, energy transitioning toward renewables, rigorously managed water and a construction workforce with decades of union history. LIUNA, with more than 123 years of existence nationally and more than a century of presence in New Mexico through Local 16, is part of that asset.

The young man from Doña Ana County who wants to leave does not need to become a developer to benefit from the AI boom. He needs someone to build the institutional scaffolding —apprenticeship programs, contracts with labor clauses, requirements for community benefits— that connects his talent with the investment that is already arriving. That is, in short, the bet Deaguero articulates from the local level, and it holds a validity that transcends New Mexico's borders: any region that aspires not to be left out of the AI economy should ask itself the same question.