← Risk register SOC 51-4081 · reviewed 2026-08-11

Multiple Machine Tool Setters, Operators, and Tenders, Metal and Plastic

124,590 US workers · median $47,180/yr · Production

EXPOSED

This job is hands-on — loading blanks, changing tooling, listening for chatter, gauging parts, clearing chips across a bank of mills, lathes, or presses — so language AI barely touches it directly. The real pressure is conventional factory automation: bar feeders, robotic load/unload cells, pallet changers, in-machine probing and automated vision inspection are exactly the technologies that let one operator run twelve machines instead of four, and then zero on a night shift. The threat is headcount compression per unit of output, not the disappearance of the skill set, and it moves at the speed of capital spending, not model releases.

10-year outlook: Expect steady employment decline through the 2030s as machine-per-operator ratios rise and robotic tending cells spread to mid-size shops, with the remaining jobs concentrated in setup, programming, and short-run work.

Score — 35/100 resistance

Five dimensions, 0–20 each, summed. Higher means more protected. The arithmetic is shown so you can check it: 11 + 14 + 1 + 3 + 6 = 35.

Task resistance 11/20

Mixed — a routine tier and a judgment tier. Setting offsets, indexing a fresh insert, deburring a part, and reading a Bridgeport's sound at the cut still need hands and eyes at the machine, but the tending half of the job — cycle-start, load, unload, gauge-to-print — is already done by bar feeders, gantry loaders and in-process probing in shops that have bought them, which is why this sits at 11 rather than up with millwright work.

Embodiment 14/20

Hands-on in uncontrolled environments. You are inside the work envelope on your feet a full shift: hoisting fixtures, flooding coolant, augering chips, wearing safety glasses and sleeves around rotating stock and a press ram, in a shop where floor temperature, oil mist and material variation change hour to hour — short of the 17-20 band only because it happens on one guarded machine bank rather than a jobsite or overhead.

Liability shield 1/20

No licence, no signature requirement. No state licence gates running a lathe or an injection press; a shop can put an unbadged temp on second shift after a week of shadowing, and NIMS credentials or an apprenticeship card affect what you get paid, not whether you are legally permitted to cut metal.

Trust premium 3/20

Anonymous artifact production. The customer sees a first-article report and a dimension inside tolerance, never you; parts ship anonymously against a PO, and the only relationship that carries weight is the programmer or quality lead knowing your scrap rate — worth 3, not 0, because a shop will keep the operator whose setups run right the first time.

Judgment & accountability 6/20

Executes defined procedures on defined inputs. You work to the print, the router, and the setup sheet with tolerances and feeds specified, and the calls you own — pull the insert now or after ten more parts, scrap this one or send it to review — are real but bounded and reversible, with the engineer or QC signing off on anything that touches the process or a nonconformance.

Confidence: high · reviewed 2026-08-11 · how scoring works

Tasks already automatable

What survives

Active moats: embodiment, physical-presence

How to future-proof this job

Escape hatches — adjacent fields with better verdicts

Computed from U.S. Dept. of Labor O*NET skill profiles: high overlap with what you already do, materially higher resistance score.

Maintenance Workers, Machinery EXPOSED · 63/100 · you already have ~89% of the skill profile

Skills to close: Installation, Repairing, Troubleshooting, Equipment Maintenance

Machinists EXPOSED · 52/100 · you already have ~87% of the skill profile

Skills to close: Mathematics

Tool and Die Makers EXPOSED · 54/100 · you already have ~86% of the skill profile

Skills to close: Operations Analysis, Technology Design, Equipment Selection, Mathematics

Field report — do you do this job?

Has AI actually changed your work?

Self-reported and unverified — a sentiment signal, not a survey. One response per person per occupation; you can change your answer.

From people who do this job

Nobody has filed one yet. If you do this work, you know things the rubric can't see.

What has actually changed in your work?

Concrete beats general: a tool that arrived, a task that moved, a headcount decision you watched happen. Don't include anything that identifies you or your employer if that would put you at risk.