EXPOSED
The paper-and-screen half of this job — writing lecture slides, generating problem sets, autograding code, drafting syllabi, answering routine 'why won't my loop compile' questions — is exactly what LLMs do well, and CS is the one discipline where students already have a free tutor better at explaining pointers than most TAs. What survives is embodied classroom presence, research supervision, capstone and thesis advising, curriculum design under accreditation pressure, and the degree-granting institution's monopoly on credentials. The modal worker here is a teaching-heavy lecturer or community-college instructor, not a tenured R1 researcher; the lecturer tier is more exposed than the research tier.
Headcount grew steadily across the period.
Median pay $85,180 → $96,980 -8.9% 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
+5.3%
Percentage only. The projection counts a different population from the 35,480 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 +5.3% 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.
~3,500 openings a year on average, including replacing people who leave.
LecturerProfessorInstructorC++ ProfessorFaculty MemberCollege ProfessorAdjunct InstructorAssistant ProfessorAssociate ProfessorComputer InstructorCollege Faculty MemberComputer Science TeacherCybersecurity InstructorUniversity Faculty MemberComputer Science ProfessorJava Programming ProfessorComputer Science InstructorNetwork Technology InstructorComputer Engineering ProfessorComputer Networking InstructorComputer Programming ProfessorComputer Technology InstructorComputer Applications InstructorAdjunct Computer Science Professor
Holding it up: trust premium . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Grading autogradable programming assignments, refreshing lecture decks on recursion, and fielding office-hour debugging questions are all now one prompt away, but running a live 200-seat lecture where you read confusion off faces, sitting on an ABET curriculum committee, and supervising a two-semester capstone team through scope collapse are not, which is why this sits at the middle of the mixed band rather than in the automatable floor.
Some physical or field component You are physically in a room — proctoring closed-book exams so students can't prompt their way through, staffing hardware and networking labs with real switches and embedded boards, and holding scheduled in-person office hours — but none of it is uncontrolled or physically demanding, so it lands at the low end of the partial-physical band rather than up with clinical or field work.
Certification preferred, not legally required No state licence gates who teaches CS 101; a master's or PhD plus departmental hiring and regional accreditation standards for faculty credentials is the whole barrier, and no statute makes you personally answerable for a graduate who ships insecure code — which is credential preference, not legal protection.
Meaningful discretion You decide academic-integrity cases where the student swears the AI-flagged submission is theirs, set the curve that determines who stays in the major, and choose whether a language or framework enters the curriculum for the next four cohorts — real discretion with appeals processes and department chairs above you, which keeps it out of the top band.
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 (5/20) 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 (12/20) 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 31 of this occupation's 50 points (62%).
Embodiment (8/20) 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.
Chief Executives SAFE
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 69/100 — SAFE.
Accreditation bodies (ABET's Computing Accreditation Commission, and regional accreditors like SACSCOC/MSCHE) already require that courses be taught by faculty with documented terminal-degree or equivalent qualification, and that credit-hour instruction involve 'regular and substantive interaction' with a qualified instructor — the same US Dept. of Education distance-education rule (34 CFR 600.2) that killed unsupervised correspondence-style courses for Title IV eligibility. If ED or accreditors explicitly rule that AI-generated instruction and AI-only feedback do not satisfy 'regular and substantive interaction,' and that a named faculty member of record must personally certify grades and learning-outcome assessment, a licensed-equivalent signature requirement attaches to every credit-bearing CS section. Watch also state authorization reciprocity (NC-SARA) guidance on AI instructors.
Genuine two-tier structure: if the routine tier (lecture delivery, autograding, syntax debugging, problem-set generation) is fully automated, the residual job is the judgment tier — curriculum redesign for a post-AI labor market, ABET outcomes assessment, research supervision, capstone sponsorship with industry partners, and lab-based assessment where the student must demonstrate reasoning live. Concretely visible where departments replace take-home programming assignments with oral code defenses and proctored whiteboard assessment, which cannot be delegated to a model.
Academic-integrity adjudication in CS has become the highest-stakes ambiguity call on campus: AI-written code is undetectable, and an instructor's determination of plagiarism carries appealable, sometimes litigated, consequences for a student's visa status or degree. If institutions formalize the instructor of record as the accountable adjudicator (as many honor-code policies already do) rather than routing to automated detection, and courts continue to require human-reviewed evidence in student disciplinary appeals, the role owns more consequential calls, not fewer. Similarly, sole authority over letters of recommendation, thesis pass/fail, and capstone client deliverables.
Employer signaling is the mechanism, not student preference: if major tech employers or professional bodies begin discounting credentials from programs without human-verified assessment — analogous to how bar examiners and medical boards refuse unaccredited pathways — institutions gain reason to advertise faculty contact hours as the product. Watch for named-faculty small-cohort programs marketed explicitly against MOOC/AI alternatives, and union contracts (e.g., CFA, PSC-CUNY, AAUP chapters) that cap section sizes or bar AI-substituted instruction.
Narrow route: shift of CS instruction toward hardware, robotics, networking labs, and embedded systems where assessment requires physical presence at benches and equipment, plus proctored in-person examination as the response to take-home cheating. This raises presence requirements but does not make the work physically unpredictable in the way trades are.
The limit. The lecturer tier faces a hard budget ceiling that no dimension score captures: even a strong faculty-of-record signature requirement can be satisfied by one instructor supervising AI across many more sections, so the protection accrues to the role while the headcount falls. Enrollment is the binding constraint — CS undergraduate enrollment softening alongside entry-level SWE hiring reduces demand regardless of how defensible the teaching task is. Contingent and adjunct positions absorb that contraction first, and no accreditation rule requires that the faculty of record be full-time.
| New York-Newark-Jersey City, NY-NJ | 2,830 | $93,980 -3% |
| Dallas-Fort Worth-Arlington, TX | 1,070 | $99,670 +3% |
| Boston-Cambridge-Newton, MA-NH | 1,060 | $119,810 +24% |
| Chicago-Naperville-Elgin, IL-IN | 990 | $85,120 -12% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 940 | $97,940 +1% |
| Los Angeles-Long Beach-Anaheim, CA | 820 | $109,930 +13% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 760 | $98,370 +1% |
| Atlanta-Sandy Springs-Roswell, GA | 660 | $107,780 +11% |
| Santa Maria-Santa Barbara, CA | 70 | $173,640 +79% |
| Lawrence, KS | 40 | $170,660 +76% |
| San Diego-Chula Vista-Carlsbad, CA | 300 | $166,550 +72% |
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.