COOKED
The core work — answering inbound calls, routing to the right extension, taking and relaying messages, looking up directory information, paging staff — is exactly what auto-attendants, IVR trees, and now conversational voice AI already do at usable quality and near-zero marginal cost. Employment has been falling for decades and voice AI removes the last argument for a human in the loop, which was that phone trees frustrate callers. What survives is thin: after-hours medical and legal answering services where a human triages urgency and reaches an on-call provider, and small operations that keep a live voice as a deliberate brand choice.
Part 2020 shock, part continued decline in the years since.
Median pay $30,610 → $38,630 +1.0% 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
-26.3% 36,600 → 27,000 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -26.3% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
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.
~2,800 openings a year on average, including replacing people who leave.
OperatorRoute ClerkDesk OperatorTest OperatorLine TechnicianSwitching ClerkTandem OperatorService OperatorThrough OperatorTraffic OperatorExchange OperatorSwitch TechnicianComplaint OperatorCombination OperatorInformation SpecialistCommunications OperatorCommunication SpecialistTelecommunications ClerkCommunications CoordinatorTelecommunications OperatorCommunication Center OperatorCommand and Control SpecialistPolice Communications OperatorTelegraph Office Telephone Clerk
Holding it up: trust premium . Weakest point: liability shield .
Core tasks are already automatable Answering, greeting with a scripted salutation, transferring to an extension, taking a name-and-number message, and paging overhead are all discrete, fully specifiable steps that auto-attendants have handled since the 1990s and voice AI now handles including the free-text message capture — the 3 rather than 0 reflects the residual after-hours medical triage script where the operator has to decide whether a caller's symptom description warrants waking the on-call doctor.
Fully desk- and screen-based The job is a headset, a console or softphone, and a directory database, done seated; the 2 rather than 0 covers hospital and hotel boards where you also physically monitor fire and elevator alarm panels and hand-deliver written message slips to the front desk.
No licence, no signature requirement No state licenses switchboard operation and no certification is asked for at hire — even HIPAA-covered medical answering services train you in-house in an afternoon and the liability for a mishandled patient call sits with the physician practice that contracted the service, not with you.
Executes defined procedures on defined inputs Routing decisions come from a directory and a call-handling protocol sheet supplied by the client, with escalation rules written out; the 3 covers judgment calls on urgency in after-hours legal and medical accounts and on whether a distressed or threatening caller gets 911, which are real but rare and bounded by the script.
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 (3/20 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 (0/20) is whether the law requires a licensed human to sign. Trust premium (5/20) is whether buyers specifically pay for a person. Judgment and accountability (3/20) 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 8 of this occupation's 13 points (62%).
Embodiment (2/20) 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 31/100, still COOKED.
Formal escalation protocols that name the operator as the accountable decision-maker for urgency classification — i.e., the answering service, not the AI vendor, owns the miss when an on-call doctor is not woken. Visible where HIPAA business-associate agreements and hospital contracts already assign breach and escalation liability to the answering vendor.
Narrow but real: legal intake and personal-injury firms paying premiums for live human first contact because conversion rates on distressed callers are measurably higher than with bots. Watch legal-intake vendors (e.g., Ngage/Alert Communications) pricing live answer above AI answer and clients choosing it.
If state medical boards or malpractice carriers require that after-hours patient calls be triaged by a human meeting a defined standard — the model already exists in nurse-triage licensure (e.g., state boards treating phone triage as nursing practice requiring RN licensure) — then the answering-service seat either upgrades into a licensed role or must hand off to one. A concrete watch item: malpractice insurers adding a policy condition that AI-only after-hours intake voids coverage for failure-to-escalate claims.
Task-mix shift in medical/legal answering: if routine routing is fully absorbed by voice AI, the residual human role is after-hours triage — deciding whether a caller's symptom description warrants waking the on-call physician, escalating a suicidal caller, or handling a caller who is incoherent, non-native-speaking, or in crisis. This tier is genuinely two-layered (dispatch vs. triage) and the remaining seats are the triage tier. Watch TigerConnect/MedConnect-style HIPAA answering vendors advertising 'human escalation tier' rather than full automation.
911/NG911 and PSAP-adjacent transfer work: if state 911 boards extend telecommunicator certification requirements (already mandatory in many states, and NENA standards) to private answering services that hand off emergency calls, a certified human becomes legally required in the chain.
The limit. Even with every lever, this stays a low-scoring occupation. The levers protect a small fraction of the 34,280 seats — medical/legal after-hours triage and emergency-adjacent transfer — and the protected version of the job is arguably no longer 'switchboard operator' but a certified triage or intake role with different hiring. Volume decline is driven by call volume itself moving to text, app, and portal channels, which no liability rule reverses.
| Los Angeles-Long Beach-Anaheim, CA | 2,110 | $45,870 +19% |
| New York-Newark-Jersey City, NY-NJ | 1,470 | $47,350 +23% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 1,270 | $35,080 -9% |
| Dallas-Fort Worth-Arlington, TX | 1,200 | $35,390 -8% |
| Chicago-Naperville-Elgin, IL-IN | 860 | $44,540 +15% |
| Houston-Pasadena-The Woodlands, TX | 740 | $34,330 -11% |
| San Francisco-Oakland-Fremont, CA | 700 | $61,780 +60% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 590 | $38,410 -1% |
| San Jose-Sunnyvale-Santa Clara, CA | 440 | $79,260 +105% |
| Urban Honolulu, HI | 240 | $63,190 +64% |
| San Francisco-Oakland-Fremont, CA | 700 | $61,780 +60% |
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 13. 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.