Cooks, Institution and Cafeteria
In short: today's best AI models could do about 20% of this job's working time (projected from METR's April 2026 measurement and Epoch's November 2026 index) if the work were set up for them, and about the same by the end of 2028 if progress keeps its long-run pace. The conversations, in-person and hands-on work, about 80% of the time, stays with people.
Prepare and cook large quantities of food for institutions, such as schools, hospitals, or cafeterias.
441,050 people do this job in the US. $37,450 a year.
Reviewed on September 30, 2026: what each task is made of, and what handing it over would ask.
A typical 40-hour week
- AI could take on0 h
- A person checks AI's work8 h
- Desk work AI can't do yet0 h
- Stays with people32 h
Cooks, Institution and Cafeteria, now: of a 40-hour week, AI could take on 0 hours, 8 hours become checking its work, 0 hours are desk work it can't do yet, and 32 hours stay with people.
The largest pieces: Plan balanced menus (about 4 h, a person checks); Watch menu costs (about 1 h, a person checks); Keep food use and cost records (about 1 h, a person checks).
Estimates for a typical week of this job, if progress keeps its long-run pace, counting work the best models finish four times in five. Every task is below.
of the working time. Now is projected from METR's April 2026 measurement and Epoch's November 2026 index.
- 14%
- 7%
- 0%
- 15%
- 64%
of the work: Anthropic's measure of what people already bring to Claude at work (2025). It counts help with part of a task, so it can be higher than the share within reach.
On the chosen pace and reliability. Estimate.
Share of the working time within reach, now to 2030
Change assumptionsHide assumptionsLong-run pace, four in five
What each choice means: , , , , . Every pace starts from .
At this date and pace, models finish tasks up to 2.4 work days long four times in five.
What handing it over would take, now
The 20% within reach splits by what a model would need. The dates, pace and reliability follow the controls above. Estimates.
of the working time. The material can be handed to a model and the result handed back.
of the working time. The model has to be connected to the system the work lives in.
is conversation short enough by length. This site leaves conversation with people, so it is not in the total.
How far this has come
With GPT-4, measured in March 2023, 0% of this job's working time was within reach at . With Claude 3.7 Sonnet, measured in February 2025, 0%. With Claude Mythos Preview (early), measured in April 2026, 14%. Today's estimate: 20%.
A backcast: today's task data with the best model METR had measured by each date. It shows how fast the models have moved, not what anyone forecast at the time.
Measured use over time
Of the Claude.ai conversations Anthropic sampled and tied to a work task, 0.021% went to this job's tasks in May 2026, the 300th largest of the 447 jobs that showed up at all; 1 of its 17 tasks showed up. Within each of Anthropic's methods, never across a change: 0.011% to 0.0038% (released February and March 2025); 0.0019% to 0.008% (August 2025 to February 2026); 0.023% to 0.021% (April to May 2026).
Each release
| Conversations | Share | Rank | Tasks seen |
|---|---|---|---|
| Released February 2025 | 0.011% | 528 of 708 | 3 of 17 |
| Released March 2025 | 0.0038% | 605 of 703 | 2 of 17 |
| The week of Aug 4, 2025 (released September 2025) | 0.0019% | 636 of 655 | 1 of 17 |
| The week of Nov 13, 2025 (released January 2026) | 0.0044% | 584 of 672 | 2 of 17 |
| The week of Feb 5, 2026 (released March 2026) | 0.008% | 525 of 677 | 2 of 17 |
| April 2026 (released June 2026) | 0.023% | 295 of 434 | 1 of 17 |
| May 2026 (released June 2026) | 0.021% | 300 of 447 | 1 of 17 |
Measured, not estimated: Anthropic's samples of Claude.ai conversations, each tied by a model to an O*NET task, from the Anthropic Economic Index. It says where Claude's use goes, not how much of this job's working time it covers, so it is not on the same scale as the reach above. The releases differ in their sample, their classifier and the smallest share they report (the June 2026 file rounds each task to a hundredth of a percent, so a smaller task reads as none), which is why a change of method starts a new line. At least 96% of each older release's task-tied use matched this site's tasks, and all of the newest. Claude.ai only, so use through other tools and other companies' models is not in it. Every job ranked.
Robots and the physical work
Anthropic rates 87% of this job's working time as physical work. For 87% of the working time, a robot has been shown doing most of the task somewhere: 87% in a purpose-built cell. 0.87 of 3, the 194th highest of the 923 jobs.
- Not physical 13%
- In a purpose-built cell 87%
Anthropic's rating, not this site's estimate: Claude Opus 5, searching the web, rated each physical O*NET task by the kind of setting in which a robot it could cite (deployed, sold or demonstrated) does most of the task, weighted by its own estimate of each task's share of the time (What work can robots do?, September 30, 2026, data CC BY 4.0; every job ranked). A robot shown somewhere is not a robot in your building: cost, setup and safety decide that. The share within reach on this page still counts no body work. The robot behind each rated task is under the task below.
What changes for the people in this job
From the task-by-task review, by the end of 2028 on the long-run pace at four in five. It says which parts of the work stay with people and which change, not whether anyone keeps a job: that depends on decisions no data set can see.
What stays with people
80% of the working time: the conversation, in-person and hands-on parts of the tasks.
- Cook to menus and diet orders, the hands-on work, 18%
- Clean and inspect kitchen equipment, the hands-on work, 9%
- Portion and serve meals, the hands-on work, 9%
- Wash pots and utensils, the hands-on work, 8%
- Prepare and cook meats, the hands-on work, 7%
- Train new kitchen staff, the in-person work, 6%
What becomes checking the machine
20% of the working time: desk work a model could do by then, in tasks where a mistake costs something or a licensed person signs. A person checks it.
- Plan balanced menus, costly if wrong, 9%
- Watch menu costs, costly if wrong, 3%
- Keep food use and cost records, costly if wrong, 3%
- Requisition food and equipment, costly if wrong, 2%
- Monitor government food commodities, costly if wrong, 1%
- Log food temperatures, costly if wrong, 1%
What to learn
- The hands-on craft. About 64% of the time is hands-on work, which no model here does; robots are rated separately on this page.
- Reviewing work you did not do yourself. The tasks under What becomes checking the machine, where a mistake costs something, stay a person's to check: spotting what a draft gets wrong is a skill of its own.
- Judgment where errors are severe. Some tasks are severe if wrong. Knowing when not to trust an output matters more there than speed.
- The conversations. About 15% of the time is conversation and in-person work, which this site leaves with people: listening, explaining, persuading, calming.
- Briefing a model and checking its drafts. About 14% of the time is documents and data, the part models reach first: writing a clear brief, giving the right material, and checking what comes back.
Your own version of this job: read it against your job description. The headline is free; the full personal report, every task with its timeline and what to learn, is $19 (a sample).
The tasks
Sorted by . 17 tasks from .
| Task | ||||
|---|---|---|---|---|
| Cook foodstuffs according to menus, special dietary or nutritional restrictions, or numbers of portions to be served.Cooking in quantity to the menu and the diet restrictions, where a wrong ingredient can hurt someone.Done daily by most who do it. Robots: in a purpose-built cell, UT Cook / T-chef Tech / OakDeer cooking robots (deployed). Anthropic's rating. | 18% | 1.7 h38 min to 3.3 h | Hands-on 100% | Stays with people |
| Clean and inspect galley equipment, kitchen appliances, and work areas to ensure cleanliness and functional operation.Cleaning the appliances and looking them over as you go.Done daily by most who do it. Robots: in a purpose-built cell, Alpha Grill (Aniai) (sold). Anthropic's rating. | 9% | 38 min17 min to 1.3 h | Hands-on 100% | Stays with people |
| Plan menus that are varied, nutritionally balanced, and appetizing, taking advantage of foods in season and local availability.Working out the menu cycle on paper from nutrition, season and cost.Done daily by most who do it. | 9% | 2.5 h50 min to 5 h | Documents 100% | 99%Half by now |
| Apportion and serve food to facility residents, employees, or patrons.Dishing up portions on the line, with the residents or patrons in front of you.Done daily by most who do it. Robots: in a purpose-built cell, Chef Robotics ChefOS meal-assembly robot (sold). Anthropic's rating. | 9% | 50 min25 min to 1.7 h | Hands-on 70%, in person 30% | Stays with people |
| Wash pots, pans, dishes, utensils, or other cooking equipment.Washing up by hand and machine.Done daily by most who do it. Robots: in a purpose-built cell, MEIKO pot and pan washers (DV/FV series, UPster PF) (sold). Anthropic's rating. | 8% | 38 min17 min to 1.3 h | Hands-on 100% | Stays with people |
| Clean, cut, and cook meat, fish, or poultry.Cutting and cooking meat, fish and poultry.Done daily by most who do it. Robots: in a purpose-built cell, Stein TwinDrum PRoYIELD Spiral Oven (JBT Marel) (sold). Anthropic's rating. | 7% | 50 min25 min to 1.7 h | Hands-on 100% | Stays with people |
| Train new employees.Showing new hires the work in the kitchen.Done more than yearly by most who do it. | 6% | 2.5 h50 min to 6.7 h | In person 100% | Stays with people |
| Direct activities of one or more workers who assist in preparing and serving meals.Leading the helpers in the kitchen through the meal.Done daily by most who do it. | 6% | 38 min17 min to 1.3 h | In person 100% | Stays with people |
| Bake breads, rolls, or other pastries.Mixing, proving and baking in quantity.Done daily by most who do it. Robots: in a purpose-built cell, Bakisto (FANUC CRX cobot + WIESHEU TrayMotion/Dibas oven + Wanzl BakeOff i) (FANUC, WIESHEU, Wanzl) (demonstrated). Anthropic's rating. | 5% | 1.7 h38 min to 2.5 h | Hands-on 100% | Stays with people |
| Rotate and store food supplies.Moving stock so the oldest is used first.Done daily by most who do it. Robots: in a purpose-built cell, Symbotic system (robotic depalletizing + autonomous mobile bots + WMS) (deployed). Anthropic's rating. | 4% | 33 min17 min to 50 min | Hands-on 100% | Stays with people |
| Monitor and record food temperatures to ensure food safety.Taking temperatures with a probe and writing them on the log.Done several times a day by most who do it. Robots: in a purpose-built cell, Automated Core Temperature Probe (CME Automation Systems (CME Ltd), UK) (sold). Anthropic's rating. | 4% | 13 min8 min to 25 min | Hands-on 70%, systems 30% | 30%Desk part half by now |
| Monitor menus and spending to ensure that meals are prepared economically.Reading the spend against the menu and adjusting.Done daily by most who do it. | 3% | 50 min25 min to 1.7 h | Documents 100% | 100%Half by now |
Within reach, per task: the share of the task's time whose instances are short enough for a model to finish at the chosen date, pace and reliability. Half by: the date half of that time comes within reach. Made of: how the task's time splits among five kinds of doing. Only the documents and systems parts can count as within reach, so a task that is 30% documents can be at most 30% within reach.
Jobs that share skills with this one
From O*NET's related occupations. Within reach now and by the end of 2028, long-run pace, four in five.
Where this job works
The industry groups that employ the most of it (BLS, May 2025).
Where the most of these jobs are
The metro areas with the most people in it (BLS, May 2025).
Questions about this job
- How much of the work of Cooks, Institution and Cafeteria is within reach of AI now?
- An estimated 20% of the working time is within reach of AI models now and 20% by the end of 2028, on METR's long-run pace at four in five. Within reach means a model could finish work that long at that reliability. It is not a forecast of jobs lost or of savings.
- Which tasks of Cooks, Institution and Cafeteria come within reach first?
- Weighted by the time each takes, the tasks with the most within reach now (projected from METR's April 2026 measurement and Epoch's November 2026 index) are: Plan balanced menus (99% of the task's time now); Watch menu costs (100% of the task's time now); and Keep food use and cost records (100% of the task's time now).
- Which tasks of Cooks, Institution and Cafeteria stay with people?
- Tasks that are people work or body work are not counted as within reach at any date. The largest: Cook to menus and diet orders; Clean and inspect kitchen equipment; and Portion and serve meals. Together such tasks are about 73% of the working time, and the people and body parts of the other tasks stay with people too.
- What should Cooks, Institution and Cafeteria learn as AI takes on more of the work?
- By the job's own task-by-task review, the first things to learn are the hands-on craft; reviewing work you did not do yourself; and judgment where errors are severe. By the end of 2028 on the long-run pace at four in five, about 80% of the working time stays with people, 20% becomes checking a model's work and 1% could be handed over with lighter checks. It says which parts of the work change, never whether anyone keeps a job.
- Do people already use AI for the work of Cooks, Institution and Cafeteria?
- In the Claude.ai conversations Anthropic sampled for May 2026 and tied to a work task, 0.021% went to this job's tasks, the 300th largest share of the 447 jobs that showed up; 1 of its 17 tasks appeared. Anthropic's labor market measure puts the share of the job's tasks seen in real Claude use at work at 0%. These are measures of use, not of how much of the work AI could finish, and they cover Claude only.
- Can robots do the physical work of Cooks, Institution and Cafeteria?
- Anthropic's robot exposure rating (September 2026) finds 87% of the working time is physical work, and that a robot has been shown doing most of the task, somewhere, for 87% of the working time. Its index is 0.87 of 3. One example: UT Cook / T-chef Tech / OakDeer cooking robots for "cook to menus and diet orders". A robot shown somewhere is not one in every workplace; this site's own estimates still count no physical work as within reach.
- How is this estimated?
- From the 17 tasks O*NET lists for the occupation, each with an estimated length checked against GDPval and a share of the working time, reviewed one task at a time on September 30, 2026 for what each is made of (documents, business systems, conversation, in person, hands-on), and METR's measurements of how long a task the best AI models can finish, carried forward to the date from the latest, Claude Mythos Preview (early) in April 2026. Data checked November 10, 2026.
This job in your state or metro area
Embed these figures on your site
A small card with the share of working time within reach now and at the end of 2027, 2028 and 2030, the date and a link back. Free under CC BY 4.0; the card carries the credit.
Your own version of this job
The page above is the occupation. Your job is your title and what you actually do: give both, and they are read against these tasks in your browser, nothing sent anywhere. The headline is free: how much of your working time is within reach now and by the end of 2028, and how many of the tasks your description names.
The full personal report is $19, to print or save as a PDF: every task with when it comes within reach, which ones your description names, the parts of your description the task list does not cover, what stays with you, what becomes checking the machine and what to learn.