SAFE
Bolting up columns and beams at 200 feet, signaling cranes, plumbing steel with turnbuckles, and burning and welding connections in wind, rain, and mud is the least robot-ready work in the economy — the environment changes every hour and every piece is a one-off fit. AI touches the paperwork around the job (shop drawing takeoffs, sequencing plans, erection schedules, safety documentation), not the raise. The main threats are construction cycles and prefabrication moving work into controlled shops, not language models.
This fall is concentrated in 2020 and has not recovered since.
Median pay $55,040 → $62,780 -8.8% 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
+4.4% 65,700 → 68,600 on the projections basis
Hard to automate, and growing
The work resists current AI and the BLS projects +4.4% 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.
~5,500 openings a year on average, including replacing people who leave.
BolterFitterErectorAssemblerBridgemanHousesmithIronworkerMetalsmithTower HandIron SetterIron WorkerSign HangerTank SetterIron ErectorJoist SetterSign BuilderSign ErectorSteel FitterSteel HangerSteel RiggerSteel WorkerLayout WorkerSteel ErectorSign Installer
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation An 18 reflects that the actual raise — spudding a wrench into a misaligned hole, tapping a beam into place with a sledge while it hangs on the choker, plumbing a bay with turnbuckles and a transit until it reads true — has no digitizable step, and the only reason it isn't 20 is that detailing, bolt takeoffs, and erection sequencing that used to be worked out on the deck now arrive as BIM output and shop-fabbed assemblies.
Hands-on in uncontrolled environments A 20 is the ceiling because the work happens on a 6-inch flange 200 feet up in whatever weather the day brings, with connection material, torch leads, and impact wrenches carried on the belt and every tie-in a one-off fit — there is no controlled-environment version of decking a floor or setting a column.
Certification preferred, not legally required An 8 recognizes that ironworkers hold real certifications with legal teeth — AWS/D1.1 welder qualification, crane signalperson and rigger cards under 1926 Subpart CC, OSHA 10/30 — but the structure is signed off by the engineer of record and the special inspector, so a failed weld comes back as a rejected joint and a requalification test, not personal liability to a third party.
Meaningful discretion A 12 covers the calls made without a supervisor at your elbow — whether a pick is rigged safely for the load and wind, whether a connection is stable enough to unhook, whether to reject bent or short-bolted material — but the erection sequence, crane plan, and connection design come to you already engineered, which keeps it below the 14+ band.
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 (18/20 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 (8/20) is whether the law requires a licensed human to sign. Trust premium (10/20) is whether buyers specifically pay for a person. Judgment and accountability (12/20) 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 30 of this occupation's 68 points (44%).
Embodiment (20/20) 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 82/100, still SAFE.
AWS D1.1/D1.8 and AISC erection practice increasingly tie critical connections to certified welders whose stamp travels with the weld; state and city rules that require a licensed/certified welder or a certified rigger-signalperson (already mandatory under OSHA 1926.1428 and NYC DOB rigger licensing) for each lift, with personal certificate exposure on failure, would push this up. Expansion of seismic-zone jurisdictions requiring a certified erector-of-record signature on connection completion reports would add most of the movement.
Two tiers exist: repetitive deck and joist work is what prefabrication and shop assembly pull away first, leaving field retrofit, connection re-fit on out-of-tolerance steel, and heavy rigging in occupied structures. If prefab keeps taking the routine tier, the residual job is the judgment tier and per-worker resistance rises even as headcount falls.
If stop-work authority is contractually vested in the connector/rigger — as in some Ironworkers Local project labor agreements and owner safety programs where the person hooking the load can refuse a pick without contest — the role formally owns the consequential call on wind, load charts, and unstable fit-up rather than the superintendent.
Owner or insurer specification of union-certified ironworkers with completed apprenticeship (common in PLAs and on public infrastructure under Davis-Bacon and federal project labor agreement requirements) is a buyer paying specifically for a credentialed human crew rather than lowest-cost erection.
The limit. Headcount, not displacement, is the real exposure: prefabrication and construction cycles can shrink the occupation sharply while every dimension score stays high. A high score here describes the surviving job, not the number of jobs.
| Phoenix-Mesa-Chandler, AZ | 3,470 | $59,540 -5% |
| Houston-Pasadena-The Woodlands, TX | 3,000 | $58,030 -8% |
| Dallas-Fort Worth-Arlington, TX | 2,850 | $53,600 -15% |
| New York-Newark-Jersey City, NY-NJ | 2,300 | $102,060 +63% |
| Los Angeles-Long Beach-Anaheim, CA | 2,150 | $78,040 +24% |
| Chicago-Naperville-Elgin, IL-IN | 2,120 | $102,350 +63% |
| Boston-Cambridge-Newton, MA-NH | 1,310 | $120,840 +92% |
| Riverside-San Bernardino-Ontario, CA | 1,120 | $72,790 +16% |
| Boston-Cambridge-Newton, MA-NH | 1,310 | $120,840 +92% |
| Trenton-Princeton, NJ | 70 | $116,360 +85% |
| Atlantic City-Hammonton, NJ | 30 | $114,570 +82% |
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 68. 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.