EXPOSED
The daily work — mounting wheels and abrasives, dialing in feeds and speeds, loading parts, checking dimensions with micrometers and gauges, deburring and finishing by feel — is hands-on and resists language AI almost entirely. The real threat here is not chatbots but the machinery itself: CNC grinders with in-process gauging, automatic wheel dressing, and robot part loading have been eating the tending tier for decades, and employment has trended down accordingly. Setters who can hold tight tolerances on unfamiliar geometries and troubleshoot chatter, burn, and taper keep their value; someone who loads a part and presses cycle-start does not.
This fall is concentrated in 2020 and has not recovered since.
Median pay $36,100 → $46,550 +3.2% 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
-12% 70,100 → 61,700 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -12% 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.
~5,500 openings a year on average, including replacing people who leave.
EdgerHonerBufferBurrerDingerLapperDresserGrinderSnaggerSnailerTrimmerBuffererDeburrerFinisherPolisherSmootherBurnisherKey FilerBurr FilerDry SanderJob SetterKey SanderPot SanderSaw Setter
Holding it up: task resistance . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Dressing a wheel to the right form, sneaking up on a 0.0002-inch bore size, and reading burn marks or chatter patterns on a part cannot be done from a keyboard, but repeat-run tending on a CNC grinder with in-process gauging is already automated in high-volume shops — that split between irreplaceable setup and replaceable cycle-start puts it at 13 rather than 17.
Hands-on in uncontrolled environments You are standing at the machine mounting wheels, balancing them, flooding coolant, loading fixtures, and hand-deburring with files and stones — but it is a shop floor with fixed equipment and known part flow, not a rooftop or a trench, so it lands at 13 rather than the high teens.
No licence, no signature requirement Nothing about this job requires a licence: no state credential to run a surface grinder, and when a part goes out of tolerance the quality inspector's signoff and the shop's certification (AS9100, ISO) absorb it, not your name.
Executes defined procedures on defined inputs Feeds, speeds, wheel grit, and dress amounts come off a process sheet or the setup sheet from engineering, and print tolerances define pass/fail — the discretion is real when you compensate for a taper or change dress frequency mid-run, but you are not deciding whether the part design is acceptable.
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 (13/20 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 (6/20) 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 10 of this occupation's 36 points (28%).
Embodiment (13/20) 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 50/100, still EXPOSED.
Task-mix shift is genuine here: this occupation has a clear two-tier structure. If the load/unload/cycle-start tier is fully absorbed by robot cells and in-process gauging, the surviving population is setters who dial in unfamiliar geometries, diagnose chatter/burn/taper, dress and select abrasives for exotic alloys, and hand-finish blend repairs on aero and medical parts. Watch for headcount falling while the remaining job titles shift to 'grind technician' or 'process setter' — the register score for the residual role rises even as employment shrinks.
If the scrap decision moves to the setter rather than QA — deciding whether a $40k forging with visible burn is reworkable, or whether taper is fixture-induced or spindle-induced before a shop stops the line — the role owns consequential calls. Watch for shops formally granting setters rework/scrap disposition authority, common in aerospace MRO cells.
If demand concentrates in low-volume, high-mix work — turbine blade tip and blend grinding, orthopedic implant polishing, weld-repair blending on castings — where fixturing is one-off and the surface is judged by feel and light, the physical component becomes harder to automate than pallet-fed production grinding. MRO and aerospace repair segments (e.g. blade repair shops) already work this way.
Nadcap/AMS2431-style special-process accreditation for grinding on flight-critical parts already requires a named, qualified operator on the process certification and burn/temper etch approvals. If primes extend operator-level qualification and personal signoff on abusive-grinding inspection to more part classes — or FDA implant surface-finish records require a named finisher — a thin shield forms. It stays thin: the signature is on process conformance, not personal professional liability.
The limit. No realistic route to a trust premium: buyers of ground parts specify Ra and tolerance, not who held the part. There is no consumer-facing 'hand-finished' market at scale outside niche firearms, knives, and restoration work. Even with every lever, the machinery threat is independent of AI capability and has been reducing headcount for thirty years; the score can rise while the occupation keeps shrinking.
| Los Angeles-Long Beach-Anaheim, CA | 3,920 | $45,440 -2% |
| Chicago-Naperville-Elgin, IL-IN | 3,110 | $46,420 +0% |
| Boston-Cambridge-Newton, MA-NH | 1,630 | $51,070 +10% |
| Cleveland, OH | 1,620 | $45,440 -2% |
| Houston-Pasadena-The Woodlands, TX | 1,290 | $43,280 -7% |
| Detroit-Warren-Dearborn, MI | 1,210 | $46,360 +0% |
| New York-Newark-Jersey City, NY-NJ | 1,180 | $48,690 +5% |
| Milwaukee-Waukesha, WI | 1,080 | $47,750 +3% |
| Canton-Massillon, OH | 350 | $67,210 +44% |
| Bend, OR | 40 | $60,180 +29% |
| Muskegon-Norton Shores, MI | 460 | $59,750 +28% |
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 36. 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.