Clean and sanitize work areas, equipment, utensils, dishes, or silverware.[2]
Where your edge is
Time Machine
Food Preparation Workers
Scrub through 151 years of this role's history — from when it first emerged, through every wave of technology that reshaped it, to the cited projections for where it's heading next.
19001925195019752000now
Drag the dot, click anywhere on the track, or use ← → arrow keys (Shift for 10-year jumps, PgUp/PgDn for 25).
2026
Known today as Food Preparation Workers (BLS SOC 35-2021)
US Employment
903K
BLS OEWS May 2024 national estimate for SOC 35-2021, sourced from O*NET which reflects the same BLS establishment-survey figure. Employment has recovered past the 2019 pre-pandemic peak. The institutional foodservice sector — driven by school feeding program expansion (29.4 million children served daily in 2023-24), healthcare system growth, and corporate dining recovery — has absorbed the commercial-sector job losses. This is the baseline used in the BLS 2024-34 employment projections.
Median Annual Wage
$34,220
Source: BLS-OEWS
Tool of the era · Sweetgreen Infinite Kitchen — robotic assembly integrated from design (2021 Spyce acquisition, 2023 deployment)
In August 2021, Sweetgreen acquired Spyce, a Boston startup founded by four MIT students that had operated a robotic fast-casual restaurant since 2018. Where Chowbotics tried to retrofit automation into existing cafeteria settings, the Spyce/Infinite Kitchen concept built the automation into the restaurant's design: a conveyor-based system that portions and assembles salad bowls from pre-cut ingredients held in temperature-controlled stations, integrated with digital ordering. Sweetgreen opened its first Infinite Kitchen location in Naperville, Illinois in May 2023. By Q1 2024, Infinite Kitchen locations were posting 28% restaurant-level profit margins — compared to 18.1% across the chain's standard locations. The key qualifier: the Infinite Kitchen's automation is confined to the assembly step — it does not cut, wash, or marinate. Human food preparation workers still perform all upstream prep. The automation eliminated the assembly-line worker for high-volume salad portioning at a purpose-built location; it has not eliminated the prep worker who cuts the vegetables those machines portion.
The Infinite Kitchen model suggests that robotic bowl assembly can succeed when designed in rather than retrofitted — but its labor displacement is narrower than the headline suggests. Sweetgreen still employs food prep workers for vegetable cutting, protein cooking, and sauce preparation; the automation handles portioning and assembly. This is a task-substitution story rather than an occupation-substitution story, consistent with the pattern seen in Chipotle's Autocado avocado robot (which cores and peels avocados but does not make guacamole) and in Miso Robotics' Flippy (which manages fry baskets but not the surrounding kitchen).
Projection cone · present → 2034
What credible sources project
Scrub the slider past now to anchor each scenario on the scrubber. The spread you see below is the range of futures credible sources project for this role.
Eloundou et al. — "GPTs are GPTs" (2023)
→ 2028
-1%
GPT-4 task-by-task LLM exposure labeling on O*NET tasks for SOC 35-2021. Food preparation workers score very low on LLM exposure — the core tasks (cutting, portioning, washing, assembling) are physical and sensory tasks that a large language model cannot perform. Eloundou explicitly notes that physical-preparation occupations have near-zero direct LLM task substitution risk. The mild negative projection reflects the marginal case: AI-assisted ordering and inventory systems in institutional foodservice may slightly reduce the labor input per meal served (better demand forecasting means less food prepared in excess), but the physical prep work itself is not LLM-automatable. Displayed near the BLS baseline.
O*NET / BLS Occupational Outlook 2024-34
→ 2034
-3%
O*NET-summarized BLS employment projections for SOC 35-2021. Classified as "Decline (-1% or lower)" — consistent with the National Employment Matrix figure. 148,000 projected annual openings despite net contraction reflects the enormous replacement-need driven by turnover: food preparation workers typically leave the occupation within 1-3 years, creating a persistent hiring demand even when headcount is not growing. The occupation remains one of the most numerically large in the US food service sector — above 860,000 workers — despite the projected contraction.
BLS National Employment Matrix 2024-34
→ 2034
-3.4%
BLS National Employment Matrix detailed projections for SOC 35-2021. Projected employment falls from 902,700 in 2024 to approximately 871,800 by 2034 — a decline of roughly 30,900 positions (-3.4%). The BLS projects mild headcount contraction driven by continued automation of assembly-line prep tasks in fast-casual restaurants (Sweetgreen Infinite Kitchen model, Chipotle Autocado/Hyphen), productivity improvements in institutional kitchens from better combi-oven and food-processor technology, and slower growth in commercial foodservice relative to the early 2020s recovery. The decline is offset by strong replacement needs: BLS projects 148,000 annual job openings through 2034, driven by the very high turnover characteristic of the occupation (annual turnover rates in food service consistently above 70%). This is a mild structural contraction, not an employment collapse.
Infinite Kitchen scale scenario (institutional automation)
→ 2034
-15%
Industry scenario: if the Sweetgreen Infinite Kitchen model — which achieves 28% restaurant-level margins vs. the 18% chain average — scales to the broader fast-casual and institutional sectors, the assembly-line subset of food preparation work faces meaningful displacement. Sweetgreen has indicated plans to expand Infinite Kitchen to additional locations; if 10-20% of high-volume fast-casual and institutional cafeteria operations adopt similar integrated robotic assembly by 2034, the displacement of assembly-worker roles within the 35-2021 tier could approach 10-15% net reduction beyond the BLS baseline. This scenario is more pessimistic than BLS but consistent with the Infinite Kitchen unit economics: unlike the Chowbotics model, the Infinite Kitchen's 28% vs 18% margin advantage is a real financial case for institutional adoption at scale, if and when capital costs come down.
Frey & Osborne (2013)
→ 2030
-30%
Gaussian-process classifier on O*NET task features. Frey & Osborne (2013) rated food preparation workers at high probability of computerization — the repetitive nature of chopping, portioning, and assembling cold preparations placed the occupation in the high-automation-risk tier. The Chowbotics Sally failure (2014-2022) was a direct test of the F&O thesis: a robot designed specifically to perform the most repetitive subset of food prep assembly failed commercially not because the task was technically impossible to automate but because the economics of automation against $14/hr labor did not close. The -30% estimate represents the pessimistic interpretation of F&O — displacement of roughly one-third of the workforce through automation of the most repetitive task subset. The actual observed trajectory (mild -3.4% BLS projection through 2034) has not validated the F&O pessimism, consistent with the economics failure documented in the Chowbotics case.
Today, in this role
What's shifting in the work right now
The historical view above shows how this role has moved. This is the present-day detail: which AI tools are picking up which tasks, where the edge still is, and the natural directions this work can grow.
What's changing in your day
Three parts of your work where AI is already doing real lifting — and what stays yours.
Assist cooks and kitchen staff with various tasks as needed, and provide cooks with needed items.[2]
Where your edge is
Take and record temperature of food and food storage areas, such as refrigerators and freezers.[2]
Where your edge is
Present-day sources
Sources
Every claim on this page traces back to one of the following. Updated 2026-05-30.
Share this year
Drops anyone you send it to straight into 2026.
Different role?
See the same long-arc view for your own profession.
Browse the directory by industry, or search by title or SOC code. New roles ship every few weeks — every profile cites every claim.
Browse all roles