COOKED verdict contested
The work — positioning parts, fastening, soldering, running presses, visual inspection against a spec sheet — is barely text work, so language AI touches it only at the edges (work instructions, quality documentation). The real exposure is industrial automation and offshoring: fixed automation, collaborative arms, and AI vision inspection already take the high-volume, low-variation stations, and the modal assembler works a repeatable bench task on a controlled line. What holds is short-run, high-mix, awkward-geometry assembly and rework, where fixturing a robot costs more than paying a person.
Roughly flat across the period, with year-to-year wobble.
Median pay $32,350 → $44,650 +10.4% 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
-0.1% 1,467,100 → 1,465,900 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -0.1% 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.
~156,300 openings a year on average, including replacing people who leave.
AgerBaserCanerFinerLinerBeaderFitterFramerJoinerLinkerPeggerRaiserReelerRollerStakerTankerWeaverWebberClamperCovererJointerLinemanMounterSneller
The BLS uses Miscellaneous Assemblers and Fabricators for work that doesn't fit any named occupation, so it covers roles that have little in common with each other. Two consequences worth knowing before you read anything below:
Holding it up: embodiment . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier At 12 rather than 6, the job survives because a huge share of assembly is short-run and high-mix — wire harnesses that must be dressed by feel, gaskets and O-rings that need tactile seating, subassemblies where the parts arrive with burrs, tolerance stack-up, or bent leads — and every model changeover means new fixtures, so the payback math keeps a human at the bench; it isn't higher than 12 because the other half of the work (press-and-fit, screw-driving, torque-to-spec, board loading, visual go/no-go against a print) is exactly what fixed automation and vision systems already do on the high-volume lines.
Hands-on in uncontrolled environments You are on your feet at a bench or line for the shift, handling parts by hand, running a soldering iron, arbor press, riveter, heat gun and torque driver, with lifting, reaching and repetitive-motion exposure that no software touches.
No licence, no signature requirement There is no licence to assemble; you may hold an IPC-A-610 or J-STD-001 solder cert or a forklift card, but those gate a workstation, not the occupation, and when a unit fails in the field the liability lands on the manufacturer's quality system and engineering signoff, not on the operator who built it — hence 1 rather than 5.
Executes defined procedures on defined inputs Most calls are bounded by the work instruction, the traveler and the torque spec, and the discretion you do have — pulling a suspect part, flagging a fit problem, stopping the line or calling the tech — is escalation rather than ownership, which puts this at 4 instead of 0 but nowhere near a role that decides in ambiguity.
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 (12/20 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 (2/20) is whether buyers specifically pay for a person. Judgment and accountability (4/20) 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 7 of this occupation's 33 points (21%).
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.
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 46/100 — EXPOSED.
Task-mix shift: as fixed automation and vision inspection absorb high-volume, low-variation stations, the surviving assembler population concentrates in short-run, high-mix, awkward-geometry build and rework — cable harnessing into tight enclosures, prototype and NPI builds, warranty teardown/repair. Watch for job postings shifting toward 'NPI technician' and 'rework/rebuild' titles, and for aerospace/defense harness shops where per-unit volumes never justify fixturing.
Certification-of-operator requirements tightening in aerospace/medical: FAA repair-station rules (14 CFR 145) and AS9100/NADCAP already require named, trained personnel for certain special processes; IPC J-STD-001 Space Addendum soldering requires individually certified operators with traceable stamps. If FDA QSR/ISO 13485 or defense primes extend named-operator traceability to more assembly steps — or if DFARS-driven supply-chain rules require a certified human sign-off on each build record rather than machine attestation — the person becomes contractually unsubstitutable for those stations.
Formal authority over disposition of nonconforming parts: if MRB (material review board) or stop-line authority is written into the assembler-technician role rather than reserved to quality engineers — as in some Toyota-derived andon systems and UAW/IAM contract language on stop-work rights — the role owns consequential accept/scrap/rework calls under ambiguity.
Reshoring of high-mix, low-volume work under CHIPS/IRA content rules and tariff regimes: if content requirements push assembly of awkward, cabled, low-volume products (medical devices, defense electronics, EV service parts) into US plants where robot fixturing cost per unit stays uneconomic, the surviving work is disproportionately the unstructured-environment kind. Watch tariff-driven onshoring announcements that specify manual final assembly rather than automated lines.
The limit. No realistic route to a trust premium: buyers of assembled goods do not pay extra for a human hand, and 'handmade' framing does not transfer to industrial subassemblies. Even with every lever above, gains concentrate in regulated aerospace/medical/defense niches — a minority of the 1.4M — and are structurally offset by continued offshoring and falling cobot fixturing cost. This occupation's realistic ceiling is roughly the mid-40s, and only for the certified-process subset.
| Detroit-Warren-Dearborn, MI | 55,100 | $49,240 +10% |
| Chicago-Naperville-Elgin, IL-IN | 47,970 | $43,650 -2% |
| Los Angeles-Long Beach-Anaheim, CA | 33,310 | $46,190 +3% |
| Dallas-Fort Worth-Arlington, TX | 31,310 | $38,340 -14% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 29,570 | $47,200 +6% |
| New York-Newark-Jersey City, NY-NJ | 26,880 | $39,640 -11% |
| Atlanta-Sandy Springs-Roswell, GA | 22,970 | $38,960 -13% |
| Houston-Pasadena-The Woodlands, TX | 21,170 | $37,890 -15% |
| Lafayette, LA | 530 | $64,950 +45% |
| Jackson, MS | 4,810 | $63,170 +41% |
| Lafayette-West Lafayette, IN | 6,280 | $63,140 +41% |
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 33. 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.