Union Carpenter Warns: AI in Construction Needs Journeyman Judgment

Greater Boston's union construction sector graduates more than 1,000 apprentices a year — and Lance Mazzariello argues that pipeline matters more, not less, as artificial intelligence spreads across job sites.

Mazzariello, a union carpenter and vice president of Specialty Services of New England, laid out the case in a recent opinion essay. He leads operations in union commercial construction around Boston, where firms are adopting technology to fix chronic problems: inefficient planning, poor communication, labor shortages, rework, scheduling and stagnant productivity.

AI is moving into estimating and project management. BIM catches conflicts before materials arrive. Reality capture documents entire jobsites. "We need these technologies," Mazzariello wrote. "But I worry we're overlooking a related problem: Who will have enough construction knowledge to know when the technology is wrong?"

When the drawing looks right and the wall says no

Mazzariello opens with a scene any construction veteran knows. A younger tradesperson studies a tablet drawing; the dimensions check out. An experienced journeyman looks at the same drawing, looks at the wall, and says, "That's not going to work." Most of the time, he's right.

Construction runs on tacit knowledge, Mazzariello argues — the carpenter who senses a problem before reaching for a tape measure, the foreman who sees a sequence will trip up three other trades, the installer who knows a detail works on paper but fails in the field. Apprenticeship has moved that judgment between generations for decades, combining classroom instruction with thousands of hours of supervised field experience.

In his framing, AI and apprenticeship are attacking the same problem from opposite ends. Both capture, transfer and apply knowledge. One distributes information instantly across an organization; the other moves judgment from one generation of workers to the next. The industry, he writes, will need both.

What does the labor shortage actually mean?

Mazzariello challenges the phrase itself. He meets plenty of people interested in the trades. What's changed is the economic proposition: a wage that looks good on paper looks different through a windshield at 4:30 a.m.

He tallies what the career demands: a dependable vehicle, fuel and insurance, drives to wherever the next project is, Boston traffic, outdoor work in heat, cold and rain, no remote work, plus overtime, layoffs and physical wear. "The hourly wage doesn't compensate for every hour the career actually consumes."

His diagnosis: the shortage partly reflects a gap between what employers can pay and what workers require. Labor can't simply get cheaper — skilled people have alternatives — and it can't simply get more expensive without pushing project costs past the point of penciling out. That leaves productivity as the variable the industry hasn't attacked aggressively enough.

Sharing the gains, not just boosting output

Massachusetts wants to register 100,000 new apprentices statewide by 2036 and has said the state needs both a bigger workforce and higher productivity. Research from the Illinois Economic Policy Institute found union apprentices receive substantial training and can reach career earnings comparable to four-year college graduates — without the same student debt.

Mazzariello's warning draws on Economic Policy Institute findings of a widening national gap between worker productivity and compensation. If construction lets technology raise output without workers sharing the gains, he writes, the industry risks reproducing the disconnect other sectors still struggle with.

The paradox, in his view: the more powerful the tools, the more valuable genuine expertise becomes beside them. Software can flag a conflict; someone still has to resolve it. A robotic layout system can put a point exactly where the model says; someone still needs the experience to recognize when the model is wrong.

There are signals the industry gets it. The Lowe's Foundation launched the Building Futures Skilled Trades Coalition, aiming to prepare one million people for skilled-trades careers by 2035. But Mazzariello notes that training more entrants doesn't fix a career people are calculating their way out of — wages, benefits and working conditions have to keep pace.

His bottom line for contractors: treat technology as a way to increase the output of increasingly valuable skilled workers and share those gains, not as a headcount cut. Because beneath the software, economics and workforce policy, as he puts it, lies a remarkably simple problem: someone still has to know how to build.