EXPOSED
The content layer of this job — leveled reading passages, grammar drills, GED practice sets, pronunciation feedback, placement quizzes, progress reports to funders — is exactly what language models and apps like Duolingo already produce cheaply, and self-study tools will absorb the motivated, higher-level learner. What survives is the part machines handle badly: standing in a room with adults who have interrupted schooling, low digital literacy, trauma, night-shift jobs and immigration stress, and keeping them coming back week after week. Funding is the real threat vector here as much as AI — adult ed programs are grant-dependent and thin-staffed, so cheap software gives budget-cutters an excuse.
This fall is concentrated in 2020 and has not recovered since.
Median pay $54,350 → $61,540 -9.4% 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
-13.7% 40,900 → 35,300 on the projections basis
Shrinking, but not obviously because of AI
The BLS projects -13.7% by 2034, but at 51/100 this work is only moderately exposed — not the profile of a job current AI can simply do. Occupations shrink for many reasons, and the score does not point at automation as this one's cause.
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.
~3,900 openings a year on average, including replacing people who leave.
TeacherInstructorAdult EducatorLiteracy CoachEnglish TeacherReading TeacherLiteracy TeacherBilingual TeacherEnglish InstructorReading SpecialistAcademic SpecialistLiteracy SpecialistAdult School TeacherBilingual InstructorAdult Literacy TeacherAdult Education TeacherAdult Literacy InstructorGeneral Education TeacherAdult Education InstructorAdult Education SpecialistAdult Basic Studies TeacherAdult Education CoordinatorCommunity Education SpecialistMath Teacher (Mathematics Teacher)
Holding it up: trust premium . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Roughly half the week — writing leveled passages, marking TABE/CASAS practice items, running pronunciation and grammar drills, generating BEST Plus placement scores and NRS progress narratives — is already replicable by an app, while the other half (multi-level classes where one adult decodes at 2nd-grade level and another needs workplace idioms, plus the phone calls chasing a student who missed three weeks for a shift change) has no software equivalent, which lands it mid-band rather than higher.
Some physical or field component You teach in physical rooms — church basements, community college annexes, county jails, employer break rooms — and do real bodywork there: modeling mouth position for /θ/, walking students through a paper GED registration or a bus route, hauling materials between two or three sites a night, but it is still a classroom rather than a job site with hazards, so it sits at the top of the 'some physical component' range, not in the field band.
Certification preferred, not legally required Credentialing is genuinely split: ABE and adult secondary teachers in K-12-operated programs often need a state teaching license, while a large share of ESL instruction at nonprofits and community-based providers hires on a bachelor's plus a TESOL certificate or nothing at all — and in neither case does a licence attach personal legal liability to your instructional calls, so the protection is a hiring filter rather than a shield.
Meaningful discretion You make consequential calls — placing someone at the right level, deciding a student is ready to sit the GED, noticing disclosed domestic violence or a learning disability and knowing where to refer — but they run inside CASAS/NRS assessment rules, mandated-reporter statutes and state curriculum frameworks that constrain the range of defensible answers, so the discretion is real but bounded, not the ambiguous ownership of a 15.
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 (11/20 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 (6/20) is whether the law requires a licensed human to sign. Trust premium (15/20) is whether buyers specifically pay for a person. Judgment and accountability (9/20) 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 51 points (59%).
Embodiment (10/20) 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 63/100, still EXPOSED.
Task-mix shift: if content generation (leveled passages, GED practice, pronunciation drills) is fully absorbed by apps, the residual job becomes intake triage, trauma-aware classroom management, and retention of learners with interrupted schooling — genuinely two-tier work, and the surviving tier is the hard tier. Watch WIOA Title II program designs that redefine the instructor role as 'learning navigator/coach' with AI content, as some state adult ed offices (e.g. Texas, Washington) have piloted.
If WIOA Title II accountability rules or state adult-ed manuals make the instructor the named decider on NRS educational functioning level placement and gain, including overriding AI-scored CASAS/TABE placements with documented rationale, the role owns a consequential call under ambiguity rather than transcribing a score.
ESL instruction attached to legally consequential attestations — USCIS-recognized citizenship-test preparation, I-9/ESL workplace compliance training, or court-mandated GED completion — where a credentialed instructor must sign attendance/competency certifications. Also state licensure floors: several states already require adult ed teachers to hold a standard teaching credential; extending that to grant-funded ESL contractors would raise the floor.
Trust premium here is paid by funders and employers, not learners: if refugee resettlement agencies, unions (e.g. SEIU/1199 training funds, Building Skills Partnership), or employer-sponsored workplace ESL contracts specify in-person human-taught cohort hours rather than software seat licenses, the premium is contractual rather than sentimental.
The limit. Every lever here is downstream of appropriations. Federal WIOA Title II adult education funding has been roughly flat-to-declining in real terms for two decades; a licensure or countersignature requirement raises per-seat cost in exactly the programs least able to absorb it, and the likely response is fewer sections, not better-protected instructors. Headcount can fall while the score rises.
| New York-Newark-Jersey City, NY-NJ | 2,750 | $83,980 +36% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 2,040 | $79,210 +29% |
| Miami-Fort Lauderdale-West Palm Beach, FL | 1,780 | $63,010 +2% |
| Chicago-Naperville-Elgin, IL-IN | 1,510 | $62,370 +1% |
| Charlotte-Concord-Gastonia, NC-SC | 1,120 | $57,510 -7% |
| Boston-Cambridge-Newton, MA-NH | 980 | $62,020 +1% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 880 | $62,940 +2% |
| Seattle-Tacoma-Bellevue, WA | 730 | $71,630 +16% |
| Modesto, CA | 70 | $129,080 +110% |
| Riverside-San Bernardino-Ontario, CA | 140 | $119,590 +94% |
| San Diego-Chula Vista-Carlsbad, CA | 170 | $107,420 +75% |
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 51. 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.