EXPOSED
This is a catch-all bucket — food runners, dining attendants in institutional settings, commissary and food-line helpers, sample demonstrators — and the modal worker spends the shift on their feet handling trays, stocking lines, portioning, wiping surfaces and answering customer questions. Almost none of that is text or screen work, so language AI barely touches it; the real threat is kiosk ordering, conveyor and robotic drink/fry stations, and simple restructuring that deletes the position rather than automating each task. There is no license, no signature, and little decision authority, so the only moat is a body in a messy, variable physical space.
Dipped in 2020, then grew past where it started.
Median pay $24,970 → $35,840 +14.8% 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
+6.4% 90,500 → 96,300 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +6.4% 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.
~14,600 openings a year on average, including replacing people who leave.
Food AideGalley BoyLine RunnerGalley WorkerKitchen StewardDietitian HelperFood Mobile DriverMini Bar AttendantFood Order ExpediterAutomat Car AttendantFood and Beverage CheckerVending Machine Host/Hostess
The BLS uses Food Preparation and Serving Related Workers, All Other 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 13 the physical core — carrying loaded trays across a crowded dining room, refilling a hot line, portioning trays for patient or student meal service, bussing and sanitizing between covers — still needs a person, but the score stops short of 14+ because the position itself is often deleted rather than automated: self-order kiosks, tray-return conveyors, automated beverage and fry stations and simple staffing cuts absorb the work without any machine replicating your hands.
Hands-on in uncontrolled environments 15 reflects a full shift standing and walking on wet tile with hot pans, sheet trays, steam tables, dish machines and unpredictable spills and customers — uncontrolled enough that no fixed robot arm can be pointed at it, though it stays below the high teens because the room is indoors, mapped, and climate-controlled rather than a construction site or a roof.
No licence, no signature requirement A 2 is because the only credential in play is a county food handler card — a few hours of training, renewable, transferable to any worker in a day — and when a foodborne illness complaint lands it is the establishment's permit and the certified food protection manager who answer to the health inspector, never you.
Executes defined procedures on defined inputs 4 fits because the calls you make are bounded by written specs: portion sizes on the recipe card, holding temperatures on the HACCP log, allergen questions escalated to the cook or supervisor, and a manager voids the check — real-time judgment about pacing and priority, but nothing where you own an ambiguous outcome.
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 (2/20) is whether the law requires a licensed human to sign. Trust premium (6/20) 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 12 of this occupation's 40 points (30%).
Embodiment (15/20) 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 54/100, still EXPOSED.
Sustained failure of robotic fry/drink/bussing deployments in cluttered, high-variance dining rooms — the visible pattern where Flippy and Servi-type units get pulled back to narrow single-station tasks and humans keep the tray-running, spill, allergen-cross-contact and reset-the-line work; if robot bussing stays confined to flat-floor chain layouts, the modal institutional/school/hospital tray role stays hands-only
State food-code adoption requiring a Certified Food Protection Manager or trained food handler physically present and personally documented for allergen-declared and texture-modified plates — FDA Food Code 2022 already pushes the Person In Charge and food allergen awareness requirement; a CMS or state long-term-care rule making the certified diet aide the named signer on tray-line allergen and consistency checks would attach a human name to each service
Task-mix shift within the bucket: kiosks and conveyor stations delete order-taking and simple portioning, leaving allergen verification, texture-modified/pureed diet assembly for dysphagia patients, and choking-risk observation in hospital and eldercare dining — which are the tiers current systems cannot do. Two genuine tiers exist here (line-helper vs. diet-aide) but the automated tier is where most of the 84,630 sit, so the shift is partly a headcount cut rather than an upgrade
Written escalation authority in institutional dining: rules or facility policy giving the tray-line attendant explicit power to reject a plate that fails a diet order or allergen check, with the refusal logged — analogous to the Person In Charge exclusion authority already in the Food Code
Union contracts that fix human staffing ratios per dining room or per resident count rather than per meal — UNITE HERE and SEIU have already bargained tech-displacement and staffing-floor language in hotel and healthcare food service; a contract that prices a human server into the meal rate creates a buyer-side floor even where a kiosk would suffice
The limit. Realistic ceiling around 55-60. The binding problem is not that machines do this work better — it is restructuring that deletes the position entirely, and no license or trust argument protects a role that simply stops existing when ordering moves to a screen. The liability and judgment levers only reach the minority of this bucket working in healthcare, school and eldercare dining under food-code and CMS oversight; food runners, samplers and commissary helpers in commercial settings have no plausible route on any institutional dimension.
| Los Angeles-Long Beach-Anaheim, CA | 14,270 | $36,810 +3% |
| San Diego-Chula Vista-Carlsbad, CA | 3,340 | $36,330 +1% |
| Riverside-San Bernardino-Ontario, CA | 3,180 | $35,260 -2% |
| New York-Newark-Jersey City, NY-NJ | 2,770 | $35,060 -2% |
| San Francisco-Oakland-Fremont, CA | 2,740 | $41,970 +17% |
| Chicago-Naperville-Elgin, IL-IN | 2,300 | $39,930 +11% |
| Baltimore-Columbia-Towson, MD | 1,980 | $37,390 +4% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 1,780 | $31,080 -13% |
| Salem, OR | 40 | $49,900 +39% |
| Kahului-Wailuku, HI | 40 | $49,400 +38% |
| Bend, OR | 30 | $45,810 +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 40. 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.