EXPOSED
The modal worker here is a route driver or pizza/food delivery driver who drives, unloads, stocks shelves or hands off orders, takes reorders, and collects payment — physical work in unpredictable environments (traffic, loading docks, apartment stairwells) that today's robotics cannot do end to end. What AI does erode is the paperwork wrapper: route planning, order suggestion, inventory counts, invoicing, and upselling scripts are already software. The real threat is not a robot taking the wheel but route consolidation, app-mediated gig substitution, and the fact that no license or liability rule protects the role.
Part 2020 shock, part continued decline in the years since.
Median pay $25,860 → $38,770 +19.9% 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
+8.8% 451,500 → 491,300 on the projections basis
Growing, and only partly exposed
The BLS expects +8.8% 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.
~51,300 openings a year on average, including replacing people who leave.
DriverIcemanMilkmanBreadmanRoutemanBobtailerRelay ManExpressmanRoute RiderBread JockeyDelivery ManRelay WorkerRoute DriverRoute JumperCity RoutemanRural CarrierCoal DelivererMilk DelivererPick Up DriverRoute SalesmanCatering DriverDelivery ExpertDelivery PersonDriver Salesman
Holding it up: embodiment . Weakest point: liability shield .
Tasks largely resist digitisation Wrestling a hand truck of beverage cases through a back door, restocking a cooler in planogram order, and finding apartment 3B up an unlit stairwell are steps no current system performs end to end, which is why this sits at 15 rather than 18 — the reorder-taking, invoicing, and route sequencing that used to be the driver's judgment is already handled by the handheld and the dispatch app.
Hands-on in uncontrolled environments An 18 reflects a day spent almost entirely outside any controlled space: driving in live traffic, backing into blind loading docks, lifting kegs and case goods, and walking unfamiliar properties in rain and ice — the only screen work is scanning and signature capture.
No licence, no signature requirement A 3 is the CDL-free reality of most of these routes: a standard license, maybe a DOT medical card and a food-handler card, and no personal exposure — when product is short or a delivery is late the company eats it, and the employer's insurance, not the driver's license, absorbs a collision.
Executes defined procedures on defined inputs At 6 the discretionary calls are bounded: whether to swap a damaged case, how to sequence three stops when the dock is blocked, whether to leave an order at the door — decisions reversed by a phone call to dispatch, with route, quantities, and pricing set before you load.
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 (15/20 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 (3/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (6/20) 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 17 of this occupation's 50 points (34%).
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 67/100 — SAFE.
Task-mix shift: when routing, invoicing, and reorder suggestion are fully automated, what remains is the non-scripted tier — exception handling at docks, damaged-freight disputes, merchandising judgment, customer relationship repair. Genuine two-tier role, so the residual job is more resistant than the average of today's
Direct-store-delivery drivers handling beer/wine already need state alcohol-delivery permits and are personally liable for ID checks; expansion of that model — e.g. state bills requiring a permitted, personally-liable human to hand off alcohol, cannabis, nicotine, or Schedule-II/refrigerated pharmacy deliveries, plus FMCSA/state rules blocking driverless handoff of controlled goods — would make a human signature legally load-bearing on a growing share of routes
Route drivers who own the account — deciding shelf allocation, credit for damaged goods, whether to refuse a delivery to an unsafe or non-compliant site — carry consequential calls; if bargaining (Teamsters DSD contracts) or company policy formalizes driver authority over credits, returns, and store-level ordering rather than routing it to an algorithm, this rises
State food-safety codes requiring a certified food handler (ServSafe-type card) to be physically present for temperature-controlled handoff and cold-chain logging, as some health departments already demand for catering and meal delivery
Narrow route only: named-driver continuity in DSD accounts (bakery, snack, beverage) where store managers demand a consistent human who knows the planogram; formalized in some vendor contracts as an assigned-rep clause
The limit. Ceiling is low. Wage pressure here comes from gig substitution and route consolidation, not robotics, and no licensing lever touches pizza or parcel delivery — the largest sub-segments. Even with every lever above, the modal food-delivery driver keeps a liability_shield near zero.
| New York-Newark-Jersey City, NY-NJ | 19,040 | $44,720 +15% |
| Atlanta-Sandy Springs-Roswell, GA | 12,020 | $41,080 +6% |
| Los Angeles-Long Beach-Anaheim, CA | 11,130 | $45,220 +17% |
| Chicago-Naperville-Elgin, IL-IN | 10,190 | $40,490 +4% |
| Dallas-Fort Worth-Arlington, TX | 9,970 | $44,440 +15% |
| Houston-Pasadena-The Woodlands, TX | 9,820 | $28,880 -26% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 9,000 | $38,790 +0% |
| Phoenix-Mesa-Chandler, AZ | 7,640 | $39,640 +2% |
| Dubuque, IA | 200 | $53,910 +39% |
| San Jose-Sunnyvale-Santa Clara, CA | 1,360 | $51,260 +32% |
| Merced, CA | 330 | $49,290 +27% |
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.