SAFE
Construction foremen spend their days on active job sites — walking the work, sequencing crews against weather and late material deliveries, catching a bad pour or an unsafe scaffold before it becomes a claim. AI can absorb the paperwork tail (daily reports, schedule updates, RFI drafts, quantity takeoffs, timesheet reconciliation), but nobody is automating the call to stop work, retrain a green apprentice, or renegotiate the sequence when the crane shows up a day early. Licensure varies by trade and state, though OSHA competent-person designation and the personal accountability for crew safety function as a real shield.
Headcount grew steadily across the period.
Median pay $66,210 → $79,920 -3.4% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+5.3% 921,600 → 970,600 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +5.3% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~74,400 openings a year on average, including replacing people who leave.
ForemanFace BossMine BossRoad BossQuarry BossCrew ForemanMine CaptainMiners' BossPipe ForemanSurface BossCivil ForemanMining CaptainPit SupervisorRig SupervisorSuperintendentDike SupervisorGang SupervisorGeneral ForemanMine SupervisorRoad SupervisorSite SupervisorConcrete ForemanDrilling ForemanField Supervisor
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation The automatable share is real but thin — daily reports, look-ahead updates, takeoffs and timesheets — while the bulk of the shift is reading actual conditions (soil that's wetter than the geotech said, a subcontractor stacking trades in the same 10 feet) and redirecting bodies, which is why this lands at 15 rather than the 18 a pure field craftsman would get.
Hands-on in uncontrolled environments An 18 reflects that the job is physically inseparable from the site: climbing scaffold to check tie-offs, putting hands on a form before a pour, walking a trench box, standing in the excavation zone during a lift — all in weather, mud and live-equipment traffic, with only the trailer paperwork done sitting down.
Licensed human required and personally liable OSHA 29 CFR 1926 competent-person designation puts a named individual personally on the hook for inspecting excavations, scaffolds and fall protection, and citations follow that designation, but the underlying credential is a training card rather than a state license in most trades — so 11, not the 16+ a licensed electrical or plumbing contractor of record carries.
Exists to be accountable for ambiguous calls A 16 is warranted because the stop-work call, the decision to accept or reject a questionable weld or pour, and the choice to send a crew home on a lightning day are made in minutes with incomplete information and carry injury, fatality and schedule-damages consequences that the supervisor answers for by name.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (15/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (11/20) is whether the law requires a licensed human to sign. Trust premium (13/20) is whether buyers specifically pay for a person. Judgment and accountability (16/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 40 of this occupation's 73 points (55%).
Embodiment (18/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 83/100, still SAFE.
OSHA's long-pending crane operator/competent-person rulemakings and state analogs (e.g., California's Cal/OSHA excavation permit regime, NYC DOB Site Safety Manager and 40-hour SST card requirements after Local Law 196) extended so that a named, certified competent person must personally sign daily inspection records for scaffolds, excavations, and fall protection — with citations issued to the individual rather than only the employer. Also: general contractor insurers (Travelers, Zurich construction practices) conditioning wrap-up liability coverage on a named on-site superintendent countersigning AI-generated inspection and pre-task plans.
Task-mix shift: the paperwork tier (daily reports, RFI drafts, timesheet reconciliation, quantity takeoffs) is genuinely automatable and is being absorbed by Procore/Autodesk copilots, leaving a day composed almost entirely of the judgment tier — reading crew competence, catching a bad pour, resequencing around a late delivery. The score rises because the denominator shrinks, not because AI got worse.
If AI takes over scheduling, takeoffs, and reporting, the residual role is stop-work authority, crew discipline, and sequence renegotiation under ambiguity — and criminal-exposure precedent (post-Hard Rock New Orleans and Florida International University bridge cases, plus NY's construction-fatality prosecutions) keeps naming the supervisor on site as the accountable party.
Little headroom; already 18. Only route is jobsite conditions getting less structured (more retrofit/renovation and disaster-rebuild work versus greenfield tract building), which raises the share of unpredictable-environment judgment.
The limit. Trust premium is the one to leave alone: owners and GCs buy schedule and safety outcomes, not a human foreman as such, and no owner is specifying 'human-supervised' as a purchased attribute. Overall ceiling is roughly mid-80s — embodiment and judgment are near max already, and the realistic gain is concentrated in liability_shield if competent-person designation hardens into individual signature-and-citation exposure.
| New York-Newark-Jersey City, NY-NJ | 28,220 | $103,990 +30% |
| Houston-Pasadena-The Woodlands, TX | 24,570 | $76,020 -5% |
| Dallas-Fort Worth-Arlington, TX | 21,530 | $76,230 -5% |
| Los Angeles-Long Beach-Anaheim, CA | 19,690 | $97,600 +22% |
| Atlanta-Sandy Springs-Roswell, GA | 16,920 | $80,500 +1% |
| Phoenix-Mesa-Chandler, AZ | 16,770 | $79,840 +0% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 16,720 | $77,530 -3% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 14,660 | $83,540 +5% |
| Mount Vernon-Anacortes, WA | 560 | $130,790 +64% |
| San Jose-Sunnyvale-Santa Clara, CA | 4,770 | $129,700 +62% |
| Kennewick-Richland, WA | 1,020 | $126,630 +58% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 73. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.