EXPOSED
Cooking 400 lunches on a line — knife work, batch cooking, holding temps, plating, cleaning hot equipment — is physical work in a chaotic environment that language models cannot touch and current kitchen robotics handles only in narrow, high-volume niches (fryers, rice, bowl assembly). The real exposure isn't a robot cook; it's software and supply chain: AI-generated cycle menus, automated nutrition and allergen compliance, forecast-driven ordering, and the ongoing shift of school, hospital, and prison food toward centralized commissaries and pre-portioned heat-and-serve, which cuts scratch-cooking headcount per site. Food handler cards are cheap and near-universal, so the licensing moat is thin.
Dipped in 2020, then grew past where it started.
Median pay $27,750 → $37,450 +8.0% 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% 466,100 → 475,400 on the projections basis
Growing, and only partly exposed
The BLS expects +2% more of these jobs by 2034, and at 50/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
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.
~69,700 openings a year on average, including replacing people who leave.
CookCamp CookLine CookMess CookSous ChefRanch CookDinner CookGalley CookSchool CookShip's CookDietary AideDietary CookKitchen CookCafeteria CookSpecial Diet CookInstitutional CookBoarding House CookCulinary SpecialistFood Service WorkerSchool Cafeteria CookFood Service SpecialistNutrition Care SpecialistPrep Cook (Preparatory Cook)
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Breaking down cases of chicken, running a steam table through a 40-minute lunch rush, adjusting a batch that's too thin, and scrubbing a tilt skillet are tasks no software touches — the 14 rather than 18 reflects that the recipe scaling, cycle-menu writing, production sheets, and par ordering that fill your prep morning are already being generated by Nutrislice-type systems and district dietitians.
Hands-on in uncontrolled environments You are on your feet six to eight hours in 90-degree kitchen heat, lifting 50-pound stock pots and sheet pans, working around slicers, convection ovens, and 350-degree fryer oil in a space where the floor is wet and three other people are moving through it — an 18 not a 20 only because it's a fixed kitchen rather than an outdoor or unpredictable site.
Certification preferred, not legally required A ServSafe food handler card or a county health card is typically a few hours of online training and a $15 test, and when an inspector writes up a violation or a norovirus outbreak is traced to the kitchen, the citation lands on the facility's permit and the certified food protection manager — not on your personal license.
Meaningful discretion You make real calls on the fly — substituting when the delivery is short, deciding a holding pan that's been at 128°F gets dumped, catching that a tray headed to a peanut-allergy student has the wrong cookie — but menus, portion sizes, HACCP critical limits, and USDA meal-pattern requirements are handed to you as written procedure, which caps this at 7.
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 (14/20 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 (5/20) 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 (7/20) 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 18 of this occupation's 50 points (36%).
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 65/100, still EXPOSED.
Allergen-fatality litigation and state school allergen laws (following the Natasha's Law model in the UK for pre-packed-for-direct-sale labeling) placing the final substitution/cross-contact call on the named kitchen lead, with documented refusal authority to serve a tray. Also IDDSI dysphagia texture-level verification in hospitals and nursing homes, where a wrong-consistency tray is a reportable adverse event.
Procurement rules that pay for on-site scratch cooking as a line item: California's $720M school kitchen infrastructure and scratch-cooking grants, USDA farm-to-school local-purchase set-asides, and university dining RFPs scored on percentage of scratch-prepared meals. These buy labor hours specifically because the buyer wants food made by people on premises. Thin outside publicly funded and unionized food service.
Task-mix shift within the same job: as heat-and-serve absorbs the routine batch tier, what remains per site is medical and religious diet execution — renal, low-sodium, texture-modified, halal/kosher separation, allergen-isolated prep — which is low-volume, high-variance, and not what commissaries ship. This raises the score only in healthcare and corrections settings, not K-12 lines.
State/local adoption of the FDA Food Code provision requiring a Certified Food Protection Manager physically present during all hours of operation (not just one per facility), plus CMS long-term-care survey findings that name the on-site cook as responsible for HACCP deviation logs and therapeutic-diet accuracy. Would make a certified human's signature on cook-chill and holding-temp records non-delegable rather than a $15 food handler card.
Union contract language (UNITE HERE, SEIU, AFSCME school food locals) setting minimum on-site certified-cook staffing ratios per meals served, or bargaining limits on commissary substitution — the mechanism that has actually protected headcount in hospital and campus dining.
The limit. Embodiment is already near ceiling at 18 and cannot meaningfully rise. The dominant threat is not robotics but centralization of production off-site, and no liability or trust lever stops a district from closing kitchens and buying pre-portioned trays — those levers only raise the value of the cooks who remain. Realistic total headroom is roughly 60-65, and only in healthcare, corrections, and grant-funded school systems.
| New York-Newark-Jersey City, NY-NJ | 15,270 | $46,870 +25% |
| Chicago-Naperville-Elgin, IL-IN | 14,010 | $38,290 +2% |
| Los Angeles-Long Beach-Anaheim, CA | 9,860 | $45,400 +21% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 9,810 | $38,870 +4% |
| Dallas-Fort Worth-Arlington, TX | 6,620 | $37,370 +0% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 6,220 | $36,480 -3% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 5,670 | $44,880 +20% |
| St. Louis, MO-IL | 5,410 | $36,200 -3% |
| Kahului-Wailuku, HI | 250 | $55,400 +48% |
| San Jose-Sunnyvale-Santa Clara, CA | 2,640 | $51,880 +39% |
| Mount Vernon-Anacortes, WA | 240 | $51,620 +38% |
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 50. 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.