Why clearing and resetting tables stays with people
Will AI replace cafeteria attendants? Not in the way the headlines suggest. The heart of this job is physical and reactive: clearing dirty dishes and glassware from tables, wiping and resetting those tables before the next guest sits down, and keeping a bartender supplied with ice, clean glassware and cut garnishes. All of it happens in a tight, moving space.
A dining room never holds still. Chairs get pushed out, a glass tips, a child drops a plate, a guest asks where the restroom is while you are carrying a loaded tray. The work also changes shape by the hour: a slow mid-afternoon reset is nothing like a lunch rush in a school cafeteria. Hands do most of it, and judgment decides the order.
Scale and pay matter too. The Bureau of Labor Statistics counts about 542,750 of these jobs in the United States, with median pay of $33,980 a year (BLS, 2025), and projects employment to grow about 5.3% between 2025 and 2035. The machine that could take over the full task set is not a chatbot but a mobile robot with arms, and the cost panel on this page compares that outlay against a wage this low. That gap is the quiet reason change comes slowly here.
What AI runs, what it assists, and what it leaves alone
No task on the list above sits in the group AI can run on its own yet, which is why the share there reads 0%.
Nothing sits in the assisted group either, at 0%, so the split gives software no partial credit on this job’s duties.
Everything is in the needs-a-human group, at 100% of task time: bussing tables and carrying dishes, glassware and silverware to the dish area, resetting place settings, replenishing condiments and bar supplies, and serving water or bread to seated guests. Our Can AI do it? figure, which estimates the share of task time AI can handle today, lands at 5. You can read how that number is built on the coverage method page.
What the evidence actually shows
There is no direct head-to-head test of a robot or an AI system against trained attendants in this occupation, so our Is it better than a person? measure carries evidence grade D and we publish no parity number for it. Grade D means not measured, not measured and failed.
What would settle it is specific and testable: timed trials in a real dining room, with tables cleared and reset per hour, breakage of plates and glassware, accuracy of the finished setting, spill cleanup handled without help, and how often a staff member has to step in. Published trials of tray-carrying service robots in senior dining and campus settings describe fetch-and-carry work, not the full clear, wipe and reset cycle. Until something measures the whole cycle, the honest answer is that the comparison has not been run. Our grading scale is explained on the quality parity page, and the wider scoring method sits alongside it.
When the picture could shift
Most likely after 2045 (8 in 10 of our scenarios). That window is wide for a reason, and the chart above shows its shape; the replacement-year method explains what the range is measuring.
Two things could pull it earlier. First, cheaper mobile robots with better arms and grippers, since the blockers on this page are mostly about handling mixed, fragile items at speed. Second, rooms redesigned around machines: cafeterias with wide flat aisles, standard trays, self-bussing stations and a single dish type give a robot a far easier floor than a crowded restaurant.
Two things hold it back. The pay in this job is low, so a machine has to be very cheap and very reliable before the math works, which is what the cost comparison on this page is for. And the hardest parts are not tidy: liquid spills, broken glass, people standing in the path, and the constant switch between clearing, stocking and answering a guest. Those are the cases where a mobile robot still needs a person nearby.
How to stay needed in a dining room
Lean into the parts of the job that carry responsibility, not just motion. Three worth building on: owning the table reset so a section turns fast and correctly; managing bar replenishment, ice and garnish prep so service never stalls; and handling the guest-facing moments, including water, bread and the small requests that come mid-shift. Attendants who do those well get asked to train new hires.
Two skills travel furthest from here. One is service judgment: reading a room, sequencing tables and spotting a problem before a guest flags it. The other is basic equipment and systems comfort, from handheld ordering tools to a floor robot that needs charging, loading and unjamming. Rooms that buy machines still need staff who can keep them running.
What to do: ask for shifts that mix bar support with floor work, since the combination is harder to break into separate, automatable pieces.
Close-by jobs are worth a look before you move: dishwashers, hosts and hostesses and bartenders. You can put any two side by side on the compare tool, see the rest of the group on the other food preparation and serving workers family page, or read how the wider industry is scored on the restaurants sector page. For the robot side of the story, our guide to humanoid robots and physical jobs covers what the hardware can and cannot do, and the list of jobs that mostly need a person shows where this kind of work sits. The headline figure for this role is 86 out of 100 (higher is safer).