EXPOSED
Almost nothing this job does happens on a screen: mixing mud, taping and sanding drywall, masking and drop-clothing rooms, setting and moving scaffolding, hauling buckets, cleaning sprayers and brushes. Current robotics cannot navigate a half-finished house or feather a patch on an uneven wall, so the AI threat here is near zero — the real exposure is that helpers hold no license, own no decisions, and have no client relationship, which makes them replaceable by other people and squeezed by any tool that raises a journeyman's output per hour. Score reflects a role protected by physicality but structurally low-leverage.
Part 2020 shock, part continued decline in the years since.
Median pay $31,340 → $40,470 +3.3% 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
+2.3% 7,400 → 7,600 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +2.3% more of these jobs by 2034. Demand for the output is growing faster than the work is being automated away — the mechanism BLS gives for software developers, and the combination people most often misread as an error.
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.
~800 openings a year on average, including replacing people who leave.
HydroblasterPower WasherMason's TenderPainter HelperPlaster HelperPlaster TenderPressure WasherShipyard HelperPlasterer TenderPainter AssistantScaffolding HelperWallpaperer HelperShip Painter HelperDry Plasterer HelperExterior Work HelperHouse Painter HelperBridge Painter HelperPaperhanger AssistantHighway Painter HelperBillboard Poster HelperShipyard Painter HelperWallpaper Hanger HelperOrnamental Plasterer HelperSwimming Pool Plasterer Helper
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Taping an inside corner, feathering a skim coat on a wall that isn't plumb, cutting in a ceiling line freehand, and dragging scaffold planks through a stud-framed hallway are sequences no software touches — the 16 rather than 20 reflects that some of the day is genuinely repetitive volume work (spray priming flat new-construction walls, mass sanding) that drywall-finishing rigs and automated sprayers already do faster than a helper with a pole sander.
Hands-on in uncontrolled environments Every hour is spent on a ladder or scaffold in an unconditioned building, carrying 50-lb mud buckets, kneeling to mask baseboard, breathing joint-compound dust and solvent vapor, and working around whatever the framers and electricians left in the way — an 18 rather than 20 only because a fair amount of it is interior, out of weather, on a floor someone else already poured.
No licence, no signature requirement No state licenses a painter's helper; you work under someone else's contractor license, sign no permits, pull no inspections, and the only card you might carry is OSHA 10 or a scaffold-user briefing that any employer can hand out in a morning.
Executes defined procedures on defined inputs The journeyman specifies the coat schedule, the color, the grit, and where the scaffold goes; your calls are limited to how much water goes in the mud and when a patch is dry enough to sand, and a wrong one costs a re-sand, not a callback or a claim.
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 (16/20 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 (5/20) 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 10 of this occupation's 44 points (23%).
Embodiment (18/20) 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 56/100, still EXPOSED.
Registered apprenticeship indenture becoming a hard condition of bidding on public work. Several states (e.g. Washington, Oregon, Illinois) already attach apprentice-utilization percentages to public contracts; if such rules expand and specify ratios of indentured helpers to journeymen, the helper slot becomes contractually mandatory on funded jobs rather than discretionary headcount.
State-level lead paint and silica rules that name the individual worker, not just the contractor: EPA RRP already requires a certified renovator on-site for pre-1978 housing, and OSHA's respirable silica standard (1926.1153) requires trained personnel for drywall sanding controls. If states moved RRP-style certification from 'one certified person per firm' to 'each worker handling disturbance must hold a card' — as some states have done for asbestos abatement workers — helpers acquire a named credential that cannot be filled by an uncertified body.
Stop-work authority written into a collective agreement or site safety program, giving the helper a personally logged call on substrate moisture, scaffold condition, or ventilation before coating. IUPAT agreements and some GC safety programs already contain named stop-work language; where the helper's sign-off is the record, the role owns a consequential call.
Task-mix shift if spray-and-sand automation (drywall finishing rigs like Canvas, or robotic sprayers) takes the flat, open-wall repetitive tier. What remains for the helper is cut-in, closets, stairwells, patch feathering on out-of-plane substrate, and rig tending — all the geometry the machines are sold as not doing. This raises the resistance of the remaining hours even as total hours fall.
The limit. Trust premium has no realistic route: helpers have no client contact, are not named on the invoice, and no homeowner has ever paid extra for a specific taper's hands. The physicality scores are already near maximum, so the honest headroom here is roughly 10-12 points, all of it institutional, and it depends on credentialing and contract language that treats the helper as a named person rather than a warm body.
| Los Angeles-Long Beach-Anaheim, CA | 550 | $47,260 +17% |
| Houston-Pasadena-The Woodlands, TX | 470 | $36,960 -9% |
| Riverside-San Bernardino-Ontario, CA | 350 | $46,780 +16% |
| San Francisco-Oakland-Fremont, CA | 300 | $50,440 +25% |
| San Diego-Chula Vista-Carlsbad, CA | 260 | $42,210 +4% |
| Dallas-Fort Worth-Arlington, TX | 230 | $40,420 +0% |
| Austin-Round Rock-San Marcos, TX | 120 | $37,920 -6% |
| Portland-Vancouver-Hillsboro, OR-WA | 120 | $40,370 +0% |
| San Francisco-Oakland-Fremont, CA | 300 | $50,440 +25% |
| Los Angeles-Long Beach-Anaheim, CA | 550 | $47,260 +17% |
| Riverside-San Bernardino-Ontario, CA | 350 | $46,780 +16% |
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 44. 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.