EXPOSED
Nothing a dishwasher does is text or screen work, so language AI touches almost none of the job: scraping plates, loading racks, hauling bus tubs, sanitizing pots, mopping floors, and stocking the line all happen in a hot, wet, cramped, constantly changing space that current robotics handles poorly and cheaply-priced labor makes uneconomical to automate. The exposure is not AI — it's conveyor machines, disposable serviceware, and restaurant closures shifting the work rather than replacing the worker. No license, no client relationship, and almost no discretion means there is no moat beyond the body doing the work.
Nearly all of this fall was the 2020 shock. It has been climbing back since.
Median pay $23,970 → $34,810 +16.2% 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
+0.2% 477,700 → 478,600 on the projections basis
Exposed, but growing
AI can already do a lot of these tasks, and the BLS still expects +0.2% 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.
~76,800 openings a year on average, including replacing people who leave.
BusserStewardScullionBus PersonDishwasherPan WasherPot WasherDish PersonDish RunnerPan CleanerDish StackerGlass WasherKitchen HelperSilver CleanerSilver WrapperUtility WorkerKitchen CleanerKitchen StewardBreakdown PersonDish Room WorkerTray Line WorkerTray Room WorkerSilverware CleanerRestaurant Dishwasher
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Scraping congealed food off irregular plates, reaching into a three-compartment sink, wrangling a warped sheet pan, and knowing a burnt stockpot needs a soak instead of a rack pass are grip-and-judgment-of-the-hands tasks no software output replaces — 16 rather than 19 because the machine cycle itself, rack conveyors, and dish return automation already absorb part of the volume in high-throughput kitchens.
Hands-on in uncontrolled environments You stand for the whole shift on wet quarry tile, lift 40-pound bus tubs and full trash bags, work near 180°F final rinse and caustic detergent, and move through a line where cooks, servers, and delivery pallets change the floor layout every ten minutes — a 17 rather than 20 because it is at least indoors within one fixed room, not a roof or a roadside.
No licence, no signature requirement No state licence exists for the dish pit; the food handler card some counties require is a short online quiz, and it is the certified food protection manager, not you, whose name is on the health inspection and who answers for a failed sanitizer test.
Executes defined procedures on defined inputs Sanitizer strip readings, water temperature, and rack sequence are set by the health code and a chart on the wall; your calls are what to soak and when to run a half-rack, decisions that get overridden by whoever is expediting.
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 (2/20) is whether buyers specifically pay for a person. Judgment and accountability (2/20) 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 5 of this occupation's 38 points (13%).
Embodiment (17/20) 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.
Cooks, Restaurant EXPOSED
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, still EXPOSED.
Cramped wet back-of-house geometry is the actual barrier: if commercial kitchen build-outs stay at current square footage and dish pits keep varied non-standard ware (cast iron, sheet pans, stemware, sous-vide gear), the manipulation problem stays unsolved even as flat-surface loading arms (Dishcraft Robotics' model, which shut down in 2023 after standardizing on its own proprietary plates) get better. Score rises only in the sense of staying pinned high — a shift the other way is standardized ware plus dish-pit-first kitchen redesign in new franchise prototypes.
Dish/porter roles are increasingly merged into 'kitchen steward' positions that own walk-in temp checks, chemical dilution, allergen-separation washing, and pest-log entries — Ecolab/Sysco-style stewardship programs and hotel union job classifications already define this tier. Where the classification exists with allergen cross-contact responsibility, the role owns a consequential call.
Health-code sanitation verification becoming a logged, named-person duty: FDA Food Code adoption of per-cycle temperature/chemical-concentration logging with an assigned employee signature (already the direction of Certified Food Protection Manager requirements spreading down from managers) would convert part of the shift into an attestable inspection task rather than throughput labor.
State or county food-handler certification becoming individually revocable and personally cited on inspection reports for sanitation failures (California and Texas already require individual food handler cards; the missing piece is personal citation rather than establishment-only fines). This is a weak shield even if it happens — certification is a two-hour course, not a scarce license.
The limit. Realistic ceiling is low-to-mid 40s. No route to a trust premium exists — diners never learn who washed the plate and there is no market willing to pay for a human at that station. The real displacement pressure here is not AI at all but disposable serviceware, off-premise/delivery formats with no dishware, conveyor machines, and restaurant closures; none of the five dimensions capture that, so a high score here would not mean job security.
| Los Angeles-Long Beach-Anaheim, CA | 27,520 | $37,580 +8% |
| New York-Newark-Jersey City, NY-NJ | 25,580 | $35,670 +2% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 12,240 | $31,480 -10% |
| Chicago-Naperville-Elgin, IL-IN | 11,910 | $36,850 +6% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 11,690 | $37,010 +6% |
| Boston-Cambridge-Newton, MA-NH | 10,540 | $37,750 +8% |
| Dallas-Fort Worth-Arlington, TX | 9,930 | $31,200 -10% |
| Las Vegas-Henderson-North Las Vegas, NV | 9,610 | $35,630 +2% |
| Kahului-Wailuku, HI | 800 | $45,930 +32% |
| San Jose-Sunnyvale-Santa Clara, CA | 3,940 | $43,960 +26% |
| Seattle-Tacoma-Bellevue, WA | 7,910 | $42,750 +23% |
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 38. 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.