EXPOSED
Lecture prep, syllabus writing, problem-set generation, grading, and literature summaries for intro geology/meteorology/oceanography courses are already substantially automatable, and enrollment pressure in small earth-science departments is a bigger near-term threat than AI itself. What holds is the embodied and relational core: running field camps and research cruises, teaching students to read outcrops and calibrate instruments, supervising labs, and advising graduate researchers who need a human to vouch for their work. No licensure protects the role, so the moat is field instruction, funded research, and mentorship rather than credentialing.
Part 2020 shock, part continued decline in the years since.
Median pay $92,040 → $103,170 -10.3% 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
+2.6%
Percentage only. The projection counts a different population from the 9,900 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Growing, and only partly exposed
The BLS expects +2.6% more of these jobs by 2034, and at 54/100 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.
~1,000 openings a year on average, including replacing people who leave.
LecturerProfessorInstructorGeodesy TeacherGeology TeacherAdjunct ProfessorCollege ProfessorGeology ProfessorHydrology TeacherPetrology TeacherAdjunct InstructorGeophysics TeacherMineralogy TeacherOceanology TeacherResearch ProfessorSeismology TeacherAssistant ProfessorAssociate ProfessorAstronomy ProfessorClimatology TeacherHydrography TeacherMeteorology TeacherPetrography TeacherVolcanology Teacher
Holding it up: trust premium . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier A 12 sits above the desk-bound humanities lecturer because half your teaching load is irreducible — taking students to a road cut and making them argue about the contact, calibrating a sonde on a cruise, troubleshooting a mass spec — but the intro-course machinery of writing lectures on plate tectonics, generating problem sets on geostrophic balance, and grading multiple-choice exams for 200-seat GEOL 101 is now largely machine work, which keeps it out of the 14+ band.
Some physical or field component Field camp in the Wasatch, a two-week berth on a UNOLS vessel, and hands-on rock saw and thin-section lab supervision put you well past screen-only, but the reason this is 12 and not 17 is that the majority of contact hours in most academic years happen in a lecture hall and an office, with the uncontrolled-environment work concentrated in a summer session or one field methods course.
No licence, no signature requirement There is no license to teach oceanography — no state board, no PE stamp, no equivalent of the Professional Geologist registration that consulting geologists in states like Texas actually need — so the 3 reflects only that field trips and shipboard work carry institutional safety duties and small-boat or research-diver certifications that a non-human cannot hold.
Meaningful discretion Deciding a student's field mapping is publishable, calling off a traverse when weather turns, judging whether an anomalous isotope result is instrumentation or discovery, and setting the ambiguous line between struggling and unqualified in a qualifying exam are all yours to own — 13 rather than 16 because these calls are reviewable by committees, department chairs, and IRB/safety officers rather than being final and unappealable.
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 (12/20 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 (14/20) is whether buyers specifically pay for a person. Judgment and accountability (13/20) 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 30 of this occupation's 54 points (56%).
Embodiment (12/20) 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.
No occupation passed every test: close enough to atmospheric, earth, marine, and space sciences teachers, postsecondary on skills and subject matter, at least 10 points more resistant, no big jump in training, no new licence, no pay cut, and not shrinking on its own. That happens for 223 of the 654 occupations here that aren't SAFE, and it is worth stating plainly rather than leaving the section off.
The usual reason is that exposure travels with the skill profile. The jobs most similar to yours tend to be exposed for the same reasons yours is, so the near neighbours don't clear the gap — and the ones that do are a different kind of work, not a transfer of what you already know. Read that as a limit of this method, not a verdict that you're stuck: it only compares whole occupations, and it cannot see specialisation, industry, or anything you'd bring that isn't in a federal skill survey.
Here is that claim on your own job rather than in the abstract. These are the three occupations closest to this one by skill and subject matter — the places the work would most naturally transfer — with what the register scores them:
That is the whole problem in three lines. The nearest work is not meaningfully safer, so there is no move here that trades a similar skill set for a better verdict. This is not us running out of ideas — it is what the neighbourhood looks like.
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 70/100 — SAFE.
Task-mix shift plus formal authorship rules: as lecture prep and grading are absorbed, the residual job becomes doctoral committee service, funding-proposal PI responsibility, and research-integrity calls. Concretely, if NSF/journal policies (already tightening AI-authorship rules at AGU, Nature, and NSF's 2023-24 proposal guidance) require a named human to certify that a student's data, model output, and interpretation are not machine-fabricated, the advisor personally owns a consequential ambiguous judgment.
Accreditation or program-review requirements that make a supervised residential field camp (e.g., the geology field-camp requirement retained by many ABET-style geoscience programs and state PG licensure boards such as Texas's, which credit specific field hours) a non-waivable degree component — plus NSF/UNOLS ship-time and research-cruise berths tied to a named shipboard instructor of record. Every hour of instruction that must happen on an outcrop, a research vessel, or with a calibrated field instrument is an hour no model performs.
State Professional Geologist boards (about 30 states license PGs; ASBOG administers the FG/PG exams) tightening the rule that only courses taught or certified by a licensed PG count toward exam eligibility, making a licensed human the signer on field-hour and course-content attestations. Parallel route: institutional lab-safety and marine-operations rules naming a specific instructor as personally responsible for student safety during field and shipboard activities, backed by university insurers after a field-trip injury claim.
Genuine two-tier structure: intro-survey lecturing and problem sets versus field mapping instruction, instrument calibration, and dissertation supervision. If departments consolidate intro courses into shared/online sections and reallocate remaining faculty lines to field methods, capstone research, and graduate mentoring, the measured residual task set is the judgment tier. Watch for department restructurings that cut intro sections while preserving field-camp and thesis-advising loads.
Narrow route only: employers and graduate admissions continuing to weight a named human's letter of recommendation and field-camp evaluation, which AI cannot supply as a reputational stake. This does not extend to classroom teaching, where students are not observably paying extra for a human instructor.
The limit. No licensure gates the teaching role itself, so the liability route is indirect and caps low. The dominant risk is not AI capability but enrollment collapse in small earth-science departments — every lever above can move and the occupation can still shrink through line elimination.
| New York-Newark-Jersey City, NY-NJ | 1,120 | $97,860 -5% |
| Los Angeles-Long Beach-Anaheim, CA | 580 | $175,120 +70% |
| Boston-Cambridge-Newton, MA-NH | 320 | $124,440 +21% |
| San Diego-Chula Vista-Carlsbad, CA | 250 | $105,080 +2% |
| Austin-Round Rock-San Marcos, TX | 170 | $129,110 +25% |
| Portland-Vancouver-Hillsboro, OR-WA | 170 | $128,140 +24% |
| Albany-Schenectady-Troy, NY | 140 | $81,600 -21% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 140 | $128,690 +25% |
| Los Angeles-Long Beach-Anaheim, CA | 580 | $175,120 +70% |
| New Haven, CT | 60 | $163,850 +59% |
| Riverside-San Bernardino-Ontario, CA | 60 | $163,120 +58% |
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 54. 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.