EXPOSED
The tending tier — loading bar stock, cycling parts, gauging with calipers and micrometers, logging counts — is exactly what bar feeders, robotic loaders, in-process probing, and lights-out CNC cells already absorb, and CAM software plus AI now drafts G-code and tooling strategies that operators once wrote by hand. What holds is the setup and troubleshooting tier: fixturing an odd part, dialing tool offsets, reading chatter and chip color, chasing a tolerance drift on a live job, and swapping inserts on a machine that doesn't care what the program says. The modal worker here straddles both tiers, which is why the number sits near the border rather than in the safe zone.
This fall is concentrated in 2020 and has not recovered since.
Median pay $40,100 → $50,620 +1.0% 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
-13.6% 18,900 → 16,400 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -13.6% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
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.
~1,500 openings a year on average, including replacing people who leave.
ChaserHobberPlanerChuckerSpinnerConcaverDeburrerThreaderBack FacerBar TurnerLathe HandNut TapperAxle TurnerBar PointerGear CutterGear HobberGear SetterGear ShaperPlaner HandRoll TurnerTool SetterBell SpinnerGun ProfilerHand Spinner
Holding it up: embodiment . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Loading a bar feeder and pressing cycle-start is already lights-out work, but nobody has automated indicating a four-jaw chuck to a half-thou, hearing a boring bar start to chatter and dropping feed on the fly, or re-shimming a soft jaw because the casting came in oversize — the split between those two halves of your shift is what puts this at 10 and not 4 or 15.
Hands-on in uncontrolled environments You are inside the machine envelope with a dial indicator, changing inserts and toolholders, blowing out chips, dealing with coolant mist and hot swarf, and hauling bar stock — 14 rather than 18 because it's your own guarded machine on a shop floor, not a rooftop or a trench where conditions change on you.
No licence, no signature requirement No licence, no stamp, no certificate anyone checks: a shop can hire you Monday on a trial run, and when a shaft goes out of tolerance the liability sits with the QC sign-off and the company's contract, not your name — the 1 rather than 0 is only the OSHA lockout/tagout duty you personally carry at the machine.
Meaningful discretion You decide when to bump an offset, when an insert is done, and when to stop the run and call it scrap — real calls with real money in them, but bounded by the print, the tolerance block, and the setup sheet, which is why this is 7 and not the 14+ of someone deciding what the tolerance should be.
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 (10/20 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 (1/20) is whether the law requires a licensed human to sign. Trust premium (3/20) is whether buyers specifically pay for a person. Judgment and accountability (7/20) 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 11 of this occupation's 35 points (31%).
Embodiment (14/20) 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.
Riggers SAFE
Sheet Metal Workers SAFE
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 50/100, still EXPOSED.
Task-mix shift is genuine here: as bar feeders, robotic loaders and in-process probing absorb loading/gauging/counting, the surviving job is setup, fixturing odd geometries, tool offset dial-in and live troubleshooting (chatter, chip color, tolerance drift). A shop that moves to lights-out spindles and retains one setter per several machines has raised the per-worker judgment content even as headcount falls.
Growth of low-volume/high-mix work — aerospace spares, medical implant one-offs, defense reshoring under DoD Industrial Base programs — where setup time dominates cycle time and no fixture or program is reused, keeps the non-automatable fraction high.
AS9100/NADCAP and FDA 21 CFR 820 audit practice already requires a named individual to sign first-article inspection and setup approval sheets. If prime contractors extend the requirement to explicitly bar AI-generated or unverified programs from being run without a named setter's signed process verification — the way Boeing and Rolls-Royce supplier quality clauses have tightened on process control — the signature becomes a job-defining requirement rather than paperwork.
Where the setter owns the call to scrap, rework, or accept a drifting lot on a $40k forged blank, and that call is logged against their name in the MES for CAPA traceability, the accountability is real. Formalizing scrap-disposition authority at the setter level (rather than escalating to a quality engineer) raises this.
Shift toward large-diameter, long-bar, or exotic-alloy turning (titanium, Inconel) where workholding is manual, deflection and thermal growth are felt not measured, and robot cells are uneconomic at the batch size.
The limit. Trust premium has no realistic route — buyers of turned parts specify tolerance, material cert and PPAP, not human authorship, and no customer pays a premium for hand-cranked chips. Even with every lever pulled, this stays a shrinking-headcount occupation: the levers raise the value of the workers who remain, not the number of them.
| Los Angeles-Long Beach-Anaheim, CA | 1,280 | $60,110 +19% |
| Houston-Pasadena-The Woodlands, TX | 1,020 | $47,900 -5% |
| Detroit-Warren-Dearborn, MI | 740 | $49,210 -3% |
| Chicago-Naperville-Elgin, IL-IN | 590 | $49,980 -1% |
| Boston-Cambridge-Newton, MA-NH | 510 | $61,890 +22% |
| San Jose-Sunnyvale-Santa Clara, CA | 360 | $74,420 +47% |
| Cleveland, OH | 350 | $51,340 +1% |
| Waterbury-Shelton, CT | 280 | $36,770 -27% |
| San Jose-Sunnyvale-Santa Clara, CA | 360 | $74,420 +47% |
| Bridgeport-Stamford-Danbury, CT | 80 | $66,980 +32% |
| Pittsburgh, PA | 220 | $66,500 +31% |
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 35. 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.