■ MODERATE RISK ■ Hospitality & Food
Partially, and it depends entirely on the price point. Casual and fast-casual service is being stripped back to app ordering and runners; proper table service is a skilled hospitality job that customers pay to keep.
“Tablets take orders. But recommending the special and reading the table? Still a human skill.”
Our AI replacement risk score — how we score jobs
Waiting tables looks simple from the customer's chair and is not. A good server sequences a full section in their head — who ordered first, whose starter is late, which table wants to linger and which has a theatre booking — while upselling without being irritating, catching allergies before the kitchen does, reading whether a couple wants conversation or invisibility, and handling the complaint that arrives when the kitchen is already eighty covers deep. Order entry is a small slice of that, and it is the slice technology has taken.
QR-code menus, table tablets, and pay-at-table apps went from novelty to default in a handful of years, and they removed real labour: ordering, upsell prompts, splitting bills, and payment. Food-running robots are now common in Asian chains and spreading in Western casual dining, moving plates from the pass to the table so staff cover more sections. Kitchen display systems and automated table management handle the sequencing that a head waiter used to do in their head. Combined, these routinely allow a restaurant to run the same covers with fewer floor staff.
The countervailing force is that hospitality is the product in anything above the mid-market. Fine dining, wine service, tasting menus, hotel restaurants — these sell attention, expertise, and a sense of being looked after, and a tablet actively degrades it. There is also an economic quirk: labour costs push operators toward automation, while tipping culture and customer expectation in some markets push the other way. Our risk score of 40 reflects an occupation splitting into a shrinking commodity tier and a durable skilled tier.
Automatability: our editorial assessment of current and near-term AI capability
The transformation is happening now and largely settles by the early 2030s. Self-ordering and pay-at-table have already cut floor headcount in casual dining, and running robots extend that in venues with the layout for them. Expect fewer servers per hundred covers across the mid-market, with the surviving roles more skilled and better paid. Premium table service remains a human craft indefinitely — its whole value proposition depends on it.
Running robots that ferry plates from kitchen to table are genuinely deployed, especially in large chain venues in Asia and increasingly elsewhere. They are useful for the walking, not the serving — a human still greets, advises, and handles anything unexpected. They reduce the number of staff needed per section rather than eliminating the role.
They already made some servers redundant, which is the honest answer. Ordering and payment were a meaningful chunk of the job and both are now self-service in much of casual dining. What remains is the part requiring judgement and warmth, and restaurants that removed that entirely have frequently found customers respond by choosing somewhere else.
As short-term or flexible work, yes — hospitality still hires constantly. As a long-term plan, only in the skilled tier: fine dining, sommelier roles, hotels, and events. The mid-market is being automated at the ordering end and squeezed on staffing ratios, so building genuine expertise is what turns the job from temporary income into a career.
Everything requiring you to read a room: knowing when to interrupt, recognising a proposal about to happen two tables over, judging that a guest is unhappy before they say so, and improvising a fix. Add product knowledge that changes a guest's order for the better. Those are perception and judgement tasks in a chaotic physical space, which is exactly where machines are weakest.