SAFE
A general pediatrician's day is hands-on physical exams of squirming infants, otoscope and auscultation findings, immunizations, growth and development assessment, and talking anxious parents through fevers, feeding, behavior, and vaccine hesitancy. AI already drafts notes, flags dosing errors, and can produce a solid differential from a symptom list, which compresses documentation and some triage — but the diagnostic exam, the prescription signature, and the parent's trust in a specific doctor are not transferable to software. Note that state licensure and personal malpractice liability are the hard legal floor here, and that floor is regulatory rather than technical.
Dipped in 2020, then grew past where it started.
Median pay $175,310 → $210,040 -4.2% 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
+0.8%
Percentage only. The projection counts a different population from the 39,390 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Hard to automate, and growing
The work resists current AI and the BLS projects +0.8% more of these jobs by 2034. Note that safe does not mean well paid — several of the fastest-growing resistant occupations are among the lowest paid on the register.
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.
~1,200 openings a year on average, including replacing people who leave.
DoctorPhysicianPediatristBaby DoctorPediatricianNeonatologistNeonatal DoctorMedical Doctor (MD)Pediatric PhysicianGeneral PediatricianPediatric HospitalistOutpatient PediatricianPrimary Care PediatricianDevelopmental PediatricianGroup Practice PediatricianMedical Pediatric PhysicianInternal Medicine PediatricianPediatric Hospitalist PhysicianDevelopmental-Behavioral PediatricianPediatric Emergency Medicine PhysicianEmergency Room Pediatrician (ER Pediatrician)DO Physician (Doctor of Osteopathic Medicine Physician)
Holding it up: liability shield . Weakest point: task resistance .
Tasks largely resist digitisation Charting, dosing checks, and well-visit anticipatory-guidance handouts are already largely templated or AI-drafted, which is why this sits at 14 rather than 18 — but eliciting a history from a nonverbal toddler, distinguishing a benign flow murmur from a pathological one by ear, and judging whether a lethargic 6-week-old needs a lumbar puncture tonight have no software substitute.
Hands-on in uncontrolled environments Every encounter requires laying hands on a patient who cannot hold still or cooperate — otoscopy on a screaming 18-month-old, palpating an abdomen, checking hip abduction for dysplasia, drawing blood from tiny veins — in exam rooms, newborn nurseries, and sometimes delivery-room resuscitations; it stops short of 19 because the setting is a controlled clinic rather than a roadside or a rooftop.
Licensed human required and personally liable You hold a state medical licence and DEA registration, you personally sign every immunization order and amoxicillin prescription, you are the named defendant when a missed diagnosis becomes a claim, and pediatric malpractice carries the longest statutes of limitation in medicine because the clock often does not start until the child turns 18.
Exists to be accountable for ambiguous calls Fever in an infant under 90 days, when to suspect non-accidental trauma and file the mandated child-abuse report, whether to treat a possible ADHD presentation with stimulants or wait, whether failure to thrive is caloric or genetic — these are decisions made on incomplete information with a caregiver as the only historian, and AAP guidelines narrow but do not decide them.
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 (14/20 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 (19/20) is whether the law requires a licensed human to sign. Trust premium (19/20) is whether buyers specifically pay for a person. Judgment and accountability (17/20) 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 55 of this occupation's 85 points (65%).
Embodiment (16/20) 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 91/100, still SAFE.
State medical boards or legislatures explicitly extending the physician-signature requirement to AI-generated pediatric outputs — e.g. rules on the model of California AB 3030 (2024, requiring disclosure and physician review of GenAI clinical communications) and Texas HB 1265-style provisions, plus AAP guidance that a licensed pediatrician must personally review and attest to any AI-suggested dosing, immunization schedule deviation, or developmental screening result. Also: malpractice carriers writing policy exclusions for undocumented physician review of AI recommendations.
Task-mix shift: as AI absorbs well-child documentation, growth-curve plotting and routine dosing, the residual role concentrates on the genuinely ambiguous tier — the febrile neonate, suspected non-accidental trauma and mandated-reporter calls, failure-to-thrive workups, custody and consent disputes, and vaccine-hesitant counseling. Watch for CPT/E&M billing shifts toward high-complexity visits and for state child-abuse reporting statutes continuing to name a licensed physician as the accountable reporter.
Same two-tier shift, on the capability side: if payers move pediatrics toward capitated/value-based arrangements (CMS Making Care Primary, state pediatric ACOs) where the pediatrician's paid work is complex-care coordination for medically fragile children and behavioral/mental-health management rather than volume of routine visits, the remaining task set is disproportionately the part software cannot deliver.
Little room upward — parental preference for a named pediatrician is already near-saturated at 19. The only realistic increment is countervailing: direct primary care and concierge pediatric memberships making the human-continuity premium explicitly priced rather than implicit.
The limit. At 85/100 with liability and trust both at 19, there is almost no headroom; the meaningful question for this occupation is not whether the score rises but whether headcount and visit volume shrink while the score stays high. A high register score does not protect against fewer pediatricians each carrying a larger, more acute panel.
| New York-Newark-Jersey City, NY-NJ | 4,510 | $216,040 +3% |
| Los Angeles-Long Beach-Anaheim, CA | 3,080 | $176,900 -16% |
| Boston-Cambridge-Newton, MA-NH | 1,860 | $221,170 +5% |
| Phoenix-Mesa-Chandler, AZ | 1,080 | $233,950 +11% |
| Houston-Pasadena-The Woodlands, TX | 1,040 | $124,190 -41% |
| Indianapolis-Carmel-Greenwood, IN | 920 | $170,020 -19% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 870 | $226,440 +8% |
| Chicago-Naperville-Elgin, IL-IN | 860 | $188,340 -10% |
| Sacramento-Roseville-Folsom, CA | 100 | $395,040 +88% |
| York-Hanover, PA | 40 | $344,960 +64% |
| St. Cloud, MN | 60 | $298,830 +42% |
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 85. 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.