Engineering Teachers, Postsecondary

25-1032.00

Median wage $109,270/yr40,270 employed (US)Rank #390 of 923 scored · top 42% by substitution

Teach courses pertaining to the application of physical laws and principles of engineering for the development of machines, materials, instruments, processes, and services. Includes teachers of subjects such as chemical, civil, electrical, industrial, mechanical, mineral, and petroleum engineering. Includes both teachers primarily engaged in teaching and those who do a combination of teaching and research.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution30

Exposure28

Augmentation62

Substitution — the headline: capability discounted by cost, barriers and adoption.

Exposure — technical capability alone, regardless of whether anyone deploys it.

Augmentation — how much AI assists without replacing. High here + moderate substitution = a changing job, not a disappearing one.

Tasks on the substitution scale

24 rated tasks, binned by substitution score.

Position among all scored occupations

Distribution of 923 occupation scores; the marker is this occupation.

Tasks with substitution ≥ 70

13%

Run 1.0.0-draft.1 · computed 2026-08-05 · rater panel: claude-sonnet-5, claude-haiku-4-5-20251001 · intervals span rater disagreement.

Why this score

The five weighted dimensions of the composite, averaged across this occupation's tasks (importance-weighted, panel mean). Exact weights and formulas: /api/v1/methodology.

Task automatabilityw 35%28

panel mean rating 2.1/5 → substitution pressure 28/100

Technical feasibility todayw 20%27

panel mean rating 2.1/5 → substitution pressure 27/100

Cost vs. human wagew 15%33

panel mean rating 2.3/5 → substitution pressure 33/100

Adoption barriersw 20%inverted — strong barriers lower the score35

panel mean rating 3.6/5 (barrier strength) → substitution pressure 35/100

Sector adoption velocityw 10%28

panel mean rating 2.1/5 → substitution pressure 28/100

Task breakdown (24 tasks)

Substitution pressure per task, weighted by O*NET importance in the composite. Expand a task for the full rater audit trail — every rating, every model, every rationale.

Maintain student attendance records, grades, and other required records.

CI 90–95 · exposure 100 · augmentation 75 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Postsecondary institutions have near-universal adoption of LMS and SIS platforms that automate these records. Displacement of manual record-keeping is already advanced across higher education.
Sector adoption velocity	claude-sonnet-5	5/5	Higher education has near-universal adoption of digital gradebooks and attendance systems already embedded in standard LMS platforms.
Augmentation potential	claude-haiku-4-5-20251001	4/5	Beyond full automation, AI systems assist faculty by generating summary reports, flagging at-risk students from attendance patterns, and providing analytics dashboards that improve decision-making without removing human oversight.
Augmentation potential	claude-sonnet-5	4/5	Existing systems substantially reduce faculty administrative burden by automating calculations, flags, and reporting, though instructors still review and finalize grades.
Task automatability	claude-haiku-4-5-20251001	5/5	Maintaining attendance records, grades, and required documentation is a structured, rule-based task that can be fully automated by learning management systems and student information systems. Current AI can handle data entry, record-keeping, and compliance logging with >50% time savings and equal or better accuracy.
Task automatability	claude-sonnet-5	5/5	Attendance and gradebook management is a highly structured, rule-based data entry and calculation task that off-the-shelf LMS/SIS automation and AI tools already handle end-to-end with minimal human input.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Educational institutions have regulatory requirements (FERPA, accreditation standards) and established processes around student records, creating some friction and oversight requirements. However, these barriers apply to the systems not the automation itself, and many institutions have already adopted them.
Adoption barriers	claude-sonnet-5	2/5	Some institutional policy requires instructor sign-off on final grades, but the mechanical record-keeping itself faces minimal legal or licensing barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Once an LMS/SIS is deployed, the marginal cost of automated record-keeping is negligible (per-record inference and storage), vastly cheaper than paying staff to manually enter, track, and manage records.
Cost vs. human wage	claude-sonnet-5	5/5	Automated record-keeping software costs a small fraction of the faculty time it would take to manually maintain these records, representing an order-of-magnitude cost advantage.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature products (Canvas, Blackboard, Banner, Workday, etc.) reliably perform this task at scale in thousands of institutions today. Educational institutions have deployed these systems widely, demonstrating reliable production performance.
Technical feasibility today	claude-sonnet-5	5/5	Mature production systems (Canvas, Blackboard, PowerSchool, gradebook software) reliably automate attendance tracking and grade recording at scale across universities today.

Compile bibliographies of specialized materials for outside reading assignments.

CI 76–81 · exposure 75 · augmentation 100 · importance 2.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Academic institutions show moderate adoption: many faculty still manually compile or delegate to library liaisons, though AI-assisted discovery is gradually entering library workflows. Adoption is faster in digitally native fields (computer science, engineering) but remains inconsistent across the sector; production deployment is still in the pilot-to-transition phase.
Sector adoption velocity	claude-sonnet-5	3/5	Higher education is adopting AI research tools moderately, with many faculty already using AI-assisted search but broad, systematic institutional deployment still emerging.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI substantially boosts faculty productivity by rapidly assembling candidate sources, suggesting relevant papers, and automating formatting while the instructor filters for quality, relevance, and pedagogical value. The human remains firmly in the loop, making final curation decisions that require domain judgment and teaching goals.
Augmentation potential	claude-sonnet-5	5/5	AI significantly speeds up literature discovery and organization, letting instructors quickly generate and refine curated reading lists while retaining final curatorial judgment.
Task automatability	claude-haiku-4-5-20251001	4/5	AI can reliably search academic databases, extract citations, format them in standard styles, and compile reading lists with minimal human intervention. The task is largely mechanical—identifying relevant sources, organizing them, and producing standardized output—which current LLMs and search tools handle effectively, easily exceeding 50% time savings.
Task automatability	claude-sonnet-5	4/5	AI can search, identify, and compile relevant specialized literature and format bibliographies very quickly, meeting most of the task's requirements with light human review.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal or licensing barriers exist; bibliographies do not require human authorization. The primary friction is faculty preference for human curation to ensure pedagogical coherence and domain fit, plus institutional inertia around library services. These are soft barriers, not binding constraints.
Adoption barriers	claude-sonnet-5	1/5	No licensing, regulatory, or liability barriers prevent using AI to compile bibliographies; it is a low-stakes administrative task.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI-driven bibliography compilation costs negligible inference and integration overhead (a few cents per task via APIs or free tools), while a faculty member assembling a curated reading list would invest 1–2 hours of loaded labor cost ($50–150). AI achieves an order-of-magnitude cost advantage.
Cost vs. human wage	claude-sonnet-5	5/5	Generating a bibliography via AI search tools costs a small fraction of a professor's time compared to manual literature searching and formatting.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Production systems like academic database tools (Google Scholar, CrossRef, Zotero with AI plugins), ChatGPT, and specialized reference-management software with AI-assisted discovery are actively deployed in universities and libraries. These systems reliably generate and format bibliographies at scale, though some oversight for domain accuracy and relevance remains standard practice.
Technical feasibility today	claude-sonnet-5	4/5	Tools like AI-powered literature search assistants and citation managers (e.g., Elicit, Consensus, Semantic Scholar with AI summarization) are deployed and widely used to compile reading lists and bibliographies today.

Prepare course materials, such as syllabi, homework assignments, and handouts.

CI 71–76 · exposure 70 · augmentation 100 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Early adoption is growing in higher education, with pilots and ad-hoc use common, but systematic institutional adoption remains limited. Faculty adoption varies by discipline and institution, placing this in the middling-to-emerging range.
Sector adoption velocity	claude-sonnet-5	3/5	Higher education is adopting generative AI for content creation at a moderate pace, with growing but uneven institutional policies and faculty buy-in.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI assistants substantially boost faculty productivity by drafting templates, adapting materials across sections, and generating variations (accessible versions, different difficulty levels), while faculty retain full control over content and pedagogy.
Augmentation potential	claude-sonnet-5	5/5	AI is highly effective as a drafting and brainstorming aid for course materials, letting instructors quickly generate and refine content while retaining final control.
Task automatability	claude-haiku-4-5-20251001	4/5	AI can generate syllabi, homework assignments, and handouts at high quality with minimal human input, meeting or exceeding typical 50% time savings. Current tools like ChatGPT and Claude excel at structured content generation, formatting, and adapting templates to course specifications.
Task automatability	claude-sonnet-5	4/5	LLMs can generate syllabi, homework problems, and handouts from a course outline with substantial time savings, though engineering-specific technical accuracy needs instructor review.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal formal barriers exist; faculty retain discretion over course design and institutional IP. Academic culture favors human originality and institutional oversight, creating some friction but no hard legal or regulatory requirement blocking AI use.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement for course material creation, though institutional norms and accreditation standards mean faculty typically review and approve final content.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI inference cost for generating a syllabus or assignment set is negligible (pennies) compared to the 2–4 hours of faculty time at a $50–100+ hourly loaded rate, yielding 100–200x cost advantage.
Cost vs. human wage	claude-sonnet-5	5/5	Generating drafts of syllabi and assignments via AI costs a few cents to dollars versus hours of faculty or TA time at academic wages.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed products (ChatGPT, Claude, specialized educational tools) reliably produce course materials in production settings. Some institutions already use AI for syllabus and assignment drafting, though human review and customization remain standard practice.
Technical feasibility today	claude-sonnet-5	3/5	General-purpose AI tools (ChatGPT, Copilot, course-design assistants) are widely used to draft such materials, but no specialized product reliably handles advanced engineering content without expert vetting.

Compile, administer, and grade examinations, or assign this work to others.

CI 48–59 · exposure 55 · augmentation 75 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Postsecondary institutions are piloting AI-assisted grading and exam generation, but production deployment of end-to-end automated assessment remains limited and uneven. Higher education is digitizing faster than traditional sectors but slower than tech/finance; faculty adoption of these tools is still exploratory rather than mainstream.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adoption of AI for grading and exam creation is still in pilot phases at most institutions, with slow, uneven rollout compared to sectors like finance or tech.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly assists faculty by auto-drafting diverse exam questions, scheduling exam administration, and pre-scoring objective components, freeing instructor time for calibrating rubrics and providing feedback on complex engineering work. This substantially raises faculty productivity while keeping the human responsible for assessment validity and judgment.
Augmentation potential	claude-sonnet-5	4/5	AI substantially helps instructors draft questions, create rubrics, and pre-grade objective sections, meaningfully speeding up the overall task while the instructor retains final judgment.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can auto-generate exam questions, administer tests via learning platforms, and perform objective grading (multiple choice, short answer matching), achieving meaningful time savings on the grading component. However, subjective problem-solving assessment, rubric calibration for engineering design work, and oversight of exam integrity still require human judgment, preventing end-to-end automation at ≥50% time savings at equal quality.
Task automatability	claude-sonnet-5	3/5	AI can generate exam questions and grade objective/short-answer responses fairly well, but compiling exams aligned to specific course objectives and grading complex engineering problem sets with partial credit still needs human oversight.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Academic institutions face moderate adoption friction: accreditation expectations around learning outcome validation, faculty resistance to outsourcing assessment (seen as core pedagogical authority), and institutional policies on academic integrity checks. However, no legal licensing requirement prevents AI use in exam administration or grading, only organizational and professional norms.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement mandates a human grade exams, but academic integrity policies, institutional accreditation standards, and instructor accountability create moderate friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Commercial LMS solutions and AI grading tools cost a few dollars per student per course, while faculty time for exam construction and grading at postsecondary wages (≥$40/hr) is substantial; AI-assisted approaches deliver significant cost savings, though instructor oversight and quality control still carry modest labor costs.
Cost vs. human wage	claude-sonnet-5	3/5	AI tools can cut time on drafting questions and grading multiple-choice items cheaply, but verification and grading of open-ended engineering problems still requires paid faculty/TA time, keeping costs moderate rather than dramatically lower.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Mature products (learning management systems like Canvas, Blackboard, and dedicated exam platforms) reliably administer and auto-grade objective assessments at scale in production. AI-powered question generation and essay scoring exist in deployment but with known error rates on open-ended engineering work, limiting full reliability on the subjective components.
Technical feasibility today	claude-sonnet-5	3/5	Products like AI-assisted grading tools and question banks exist and are used in some LMS platforms, but reliable automated grading of technical/engineering work with multi-step derivations remains narrow and error-prone.

Evaluate and grade students' class work, laboratory work, assignments, and papers.

CI 51–54 · exposure 50 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Early-stage pilot adoption is common in higher education (plagiarism detection is mainstream; LLM grading is emerging), but production replacement of instructor grading is still rare and cautious. The sector is moving faster than laggards but slower than finance or corporate services.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education, especially engineering departments, adopts grading automation slowly and unevenly compared to fast-moving sectors like finance or tech.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly assists grading by drafting detailed feedback, identifying common errors across cohorts, and automating scoring of objective components, thereby raising instructor productivity and feedback quality. The instructor remains the decision-maker on final marks, particularly for nuanced lab and project work.
Augmentation potential	claude-sonnet-5	4/5	AI substantially speeds up grading of written work, provides feedback drafts, and flags issues, meaningfully boosting instructor productivity while they retain final judgment.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can automate parts of grading—scoring multiple-choice and structured problem sets with clear rubrics, flagging plagiarism, and generating feedback on technical accuracy. However, evaluation of laboratory work quality, experimental methodology, and papers requiring nuanced judgment on originality and conceptual understanding would still require significant human oversight, preventing a clean ≥50% time-saving outcome end-to-end.
Task automatability	claude-sonnet-5	3/5	AI can grade written assignments and objective work with rubrics, but engineering lab work, design projects, and open-ended problem sets often require domain judgment and verification of physical/practical correctness that current AI handles unevenly.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Universities and accrediting bodies increasingly expect human accountability in final grades and student feedback; many institutions have policies requiring faculty sign-off. Institutional inertia and faculty preference for direct student interaction provide friction but not hard legal bars to adoption.
Adoption barriers	claude-sonnet-5	3/5	Grading is typically an instructor's academic responsibility with institutional policies requiring instructor oversight and final sign-off, creating moderate friction, though no strict licensing barrier prevents AI assistance.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	API costs for AI-powered grading (inference + integration overhead) are low relative to instructor labor; bulk deployment in universities shows favorable economics. However, overhead for custom rubric setup and human review to catch errors prevents a full 5-level rating.
Cost vs. human wage	claude-sonnet-5	4/5	Once set up, AI grading of assignments is far cheaper per unit than faculty/TA time, though initial rubric calibration and oversight add cost.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	LLM-based grading systems and plagiarism detection tools are deployed in educational institutions with measurable accuracy on defined rubrics, but they perform poorly on open-ended lab reports and subjective components. Existing products exist but carry material error rates when applied to the full range of engineering assessment types.
Technical feasibility today	claude-sonnet-5	3/5	AI grading tools (e.g., automated essay/code graders, plagiarism and rubric-based assistants) are deployed in some universities, but adoption for engineering-specific lab reports and design work is narrower and error-prone.

Write grant proposals to procure external research funding.

CI 29–59 · exposure 38 · augmentation 88 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Engineering faculty in academia are adopting AI writing assistants slowly and cautiously, primarily for drafting support rather than end-to-end proposal generation. The conservative, high-stakes nature of grant funding and institutional gatekeeping mean adoption remains limited to draft assistance in most departments.
Sector adoption velocity	claude-sonnet-5	3/5	Academic research settings show growing but uneven adoption of AI writing tools; usage is common informally but institutional policies and disclosure norms are still developing, placing this in the middle range.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at accelerating proposal drafting—generating initial text, outlining sections, suggesting language for methods and timelines—allowing faculty to focus on strategic vision and novelty claims. This is a high-value assistive use case where AI materially speeds iteration without replacing human judgment.
Augmentation potential	claude-sonnet-5	5/5	AI significantly speeds up drafting, editing, formatting, and literature synthesis for grant proposals, making it a strong productivity multiplier for postsecondary engineering faculty who retain final intellectual control.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can draft text and outline grant sections, grant proposals demand deep domain expertise, original research framing, institutional knowledge of funding agencies, and credible justification of feasibility—all requiring human judgment and oversight. Current systems cannot reliably generate the strategic vision and evidence synthesis needed to meet funding agency expectations at scale.
Task automatability	claude-sonnet-5	3/5	AI can draft substantial portions of grant proposals (background, literature review, boilerplate sections) but requires significant human input for original research ideas, budget justification, and institutional specifics, limiting full end-to-end automation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Funding agencies (NSF, NIH, DoE) require that proposals be signed and take legal responsibility for accuracy and feasibility claims. Faculty members must personally certify the work, and institutions maintain strict oversight over grant submission, creating organizational and liability barriers to full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement bars AI-assisted writing, but funding agencies often require named PI certification of originality/integrity and institutional review, creating moderate friction against pure automation.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI-assisted drafting can reduce the time an engineer-faculty member spends on initial writing, but the proposal still requires substantial faculty review, revision, and sign-off. The cost savings are modest—perhaps 20–30% of effort—placing this in the comparable-cost range.
Cost vs. human wage	claude-sonnet-5	4/5	AI drafting assistance costs a small fraction of the time a faculty member would spend, though the final product still requires expert review, keeping it just short of an order-of-magnitude reduction in total task cost.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some AI writing tools (like GPT-4) can produce grant-like text, but no deployed product reliably writes fundable grant proposals end-to-end. Proposals require precise articulation of novel contributions, budget justification, and risk mitigation that exceeds current AI capabilities without extensive human revision.
Technical feasibility today	claude-sonnet-5	3/5	AI writing tools (ChatGPT, Claude, specialized grant-writing assistants) are used in production for drafting and editing proposals, but reliability varies and human review/rewriting is standard practice, not full autonomous generation.

Select and obtain materials and supplies, such as textbooks and laboratory equipment.

CI 25–47 · exposure 33 · augmentation 63 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Educational institutions, particularly postsecondary engineering departments, move slowly on administrative automation. Most still rely on manual requisition forms and departmental purchasing protocols rather than AI-assisted systems, reflecting low digitization of academic procurement workflows.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administrative/procurement functions adopt AI tools slowly, with most institutions still using traditional purchasing and requisition systems.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by searching supplier catalogs, summarizing product specifications, comparing prices across vendors, and flagging equipment that meets safety standards. However, the human instructor must retain final judgment on educational fit and quality, limiting the transformative potential of augmentation.
Augmentation potential	claude-sonnet-5	4/5	AI tools can meaningfully speed up research on textbook options, equipment comparisons, and vendor pricing, helping instructors make more informed selections faster.
Task automatability	claude-haiku-4-5-20251001	2/5	A small portion of this task—identifying items and placing orders—could be partially automated, but the task requires human judgment on material quality, curriculum alignment, budget constraints, and supplier relationships. The full workflow of selecting appropriate materials and obtaining them involves substantial human decision-making that current systems cannot reliably handle end-to-end.
Task automatability	claude-sonnet-5	3/5	AI can research and recommend textbooks, compare lab equipment specs, and draft purchase orders, but final selection requires curriculum judgment and vendor/budget coordination that still needs human decision-making.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Educational institutions have procurement policies, vendor approval processes, and budget authorization requirements that legally or procedurally require human decision-makers. Safety regulations for laboratory equipment and institutional purchasing controls create meaningful friction against full automation.
Adoption barriers	claude-sonnet-5	2/5	Institutional purchasing policies and budget authorization requirements create some friction, but there's no licensing or legal requirement that a human must personally select materials.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The procurement and integration overhead for an AI system would likely exceed savings for a task that typically occurs episodically (semester or annual purchasing cycles) and requires specialized knowledge that an AI system would need continuous human oversight to validate.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted research and drafting of purchase requests can cut time somewhat, but actual procurement, budgeting, and vendor negotiation still require paid staff time, keeping costs roughly comparable.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can assist with inventory lookups and basic procurement workflows, no mature product reliably handles the domain-specific selection judgment required (evaluating textbooks for engineering curricula, vetting lab equipment for safety and pedagogical fit). Existing e-procurement systems lack the contextual engineering education expertise needed.
Technical feasibility today	claude-sonnet-5	2/5	No deployed product autonomously manages procurement of academic materials end-to-end; existing tools (search, comparison, ordering systems) are used piecemeal with human oversight.

Review manuscripts for professional journals.

CI 25–43 · exposure 33 · augmentation 63 · importance 2.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Postsecondary institutions and journal publishers use AI mainly for desk-reject or plagiarism screening, not end-to-end review. Adoption remains cautious because peer review is a critical quality control and career-legitimacy function in academia.
Sector adoption velocity	claude-sonnet-5	2/5	Academic publishing is adopting AI slowly for plagiarism/statistics checks, but substantive peer review remains largely traditional and human-driven with cautious experimentation.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist reviewers by summarizing papers, flagging methodological issues, checking reference accuracy, and suggesting improvements to writing clarity, reducing administrative burden and improving review depth on the human's core judgment tasks.
Augmentation potential	claude-sonnet-5	4/5	AI tools substantially assist reviewers by summarizing papers, checking references, identifying statistical issues, and drafting review comments, improving efficiency while the human retains final judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can assist with manuscript screening (plagiarism detection, structural feedback, initial quality flags), but peer review requires domain expertise, subjective judgment on novelty/significance, and accountability for editorial decisions—tasks current systems cannot reliably perform end-to-end at the quality threshold.
Task automatability	claude-sonnet-5	3/5	AI can summarize manuscripts, check methodology consistency, detect statistical errors, and flag plagiarism or missing citations, saving significant reviewer time, but final judgment on novelty, significance, and technical soundness still requires domain expertise.review is not fully replaceable end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Peer review carries high stakes: editorial liability, professional reputation, and gatekeeping authority over publication. Most journals and funders still legally or normatively require qualified human reviewers to take responsibility for substantive assessment.
Adoption barriers	claude-sonnet-5	3/5	No formal licensing requirement, but strong norms of academic integrity, conflict-of-interest management, and reputational accountability mean journals still require named human reviewers with subject expertise.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for manuscript analysis (plagiarism, language checking) are inexpensive, but they supplement rather than replace the human reviewer's cognitive work. A full replacement would require bespoke integration and human oversight that narrows cost advantage.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted screening tools are cheap relative to reviewer time for narrow subtasks, but comprehensive review requiring expert judgment still needs paid/unpaid academic labor, keeping overall cost comparable.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product routinely performs full peer review; some journals use AI for desk-reject screening or plagiarism checks, but these are narrow components. Reliable full-manuscript technical and conceptual review remains a human responsibility.
Technical feasibility today	claude-sonnet-5	2/5	Some journals use AI tools for initial screening (plagiarism, statistics checks, format compliance) but no deployed product performs full substantive peer review reliably at scale.

Participate in student recruitment, registration, and placement activities.

CI 25–30 · exposure 25 · augmentation 50 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Adoption of AI in postsecondary recruitment remains limited and pilot-focused. Most institutions continue to rely on traditional human-driven recruitment and placement models, with AI tools used only for narrow administrative tasks like email filtering or basic CRM features.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administration adopts AI tools for outreach and CRM slowly compared to fast-moving sectors like finance or tech, with pilots more common than full production deployment.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist recruitment coordinators by flagging qualified candidates, drafting outreach emails, and flagging placement opportunities, but the core interpersonal and judgment work remains human-driven. Systems like applicant screening dashboards provide useful support without replacing human decision-making.
Augmentation potential	claude-sonnet-5	3/5	AI can help draft recruitment materials, manage communications, and analyze applicant data, meaningfully assisting the faculty member while they retain decision-making control.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with email campaigns and data management, student recruitment and placement require relationship-building, personalized persuasion, and complex decision-making about fit that current systems cannot reliably perform end-to-end. Registration processes are partially automatable (form processing), but placement requires human judgment about individual contexts and employer relationships.
Task automatability	claude-sonnet-5	2/5	This task blends interpersonal advising, event participation, and administrative judgment that AI cannot fully replicate end-to-end; only sub-components like drafting communications or scheduling can be automated.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Postsecondary institutions face regulatory requirements around FERPA, accreditation standards for placement outcomes, and contractual obligations to employers and students. Liability for poor placement decisions and the expectation that faculty/staff personally vouch for students create substantial organizational and legal friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No strict licensing requirement blocks AI use, but institutional policy, FERPA-related privacy concerns, and expectations of personal faculty involvement in recruitment/placement create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Current AI recruitment tools require significant human oversight, customization, and validation of matches. The loaded cost of a recruitment coordinator or placement officer is often lower than the integrated cost of AI platform, integration, and mandatory human review.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools can cheaply handle outreach messaging or scheduling, but the faculty member's personal engagement, mentorship, and judgment in placement decisions still require paid human time, keeping overall cost comparable to human labor.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some components (email outreach, document processing) have deployable tools, but no mainstream product reliably handles full recruitment pipelines or placement matching. Most institutions rely on human coordinators supplemented by basic CRM tools rather than AI-driven end-to-end recruitment systems.
Technical feasibility today	claude-sonnet-5	2/5	Some CRM and chatbot tools assist with recruitment outreach and FAQ answering, but no deployed product handles the full recruitment/registration/placement workflow reliably in production for faculty roles.

Advise students on academic and vocational curricula and on career issues.

CI 25–30 · exposure 25 · augmentation 75 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Higher education has lagged in AI adoption for core pedagogical and advising functions; most institutions piloting AI advising tools use them as supplementary resources, not replacements, reflecting both cultural preference for human contact and regulatory caution.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools unevenly and cautiously, with advising functions typically among the slower areas to see AI integration compared to research or administrative tasks.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can meaningfully assist faculty advisors by surfacing curriculum options, degree requirements, career statistics, and prerequisite mappings, allowing the advisor to focus on personalized dialogue and motivation. This augmentation clearly enhances faculty productivity while the human maintains judgment.
Augmentation potential	claude-sonnet-5	4/5	AI can help faculty advisors by summarizing degree requirements, drafting career resources, and providing quick information lookups, meaningfully speeding up parts of the advising process while the human retains the relationship.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can provide generic information about curricula and career paths, advising students requires understanding individual goals, learning styles, constraints, and career fit—nuanced judgment that current systems cannot reliably replicate end-to-end. AI might assist with research components but cannot replace the personalized, adaptive counseling that defines this task.
Task automatability	claude-sonnet-5	2/5	Advising involves relationship-building, understanding individual student circumstances, and institution-specific judgment calls that current AI cannot fully replicate end-to-end, though it can support parts of the information-gathering process.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Institutional and regulatory frameworks expect faculty advisors to take responsibility for student guidance; many universities require human sign-off on academic plans, and students often seek human mentorship on career trajectories. Legal and fiduciary liability for poor advice creates friction against full automation.
Adoption barriers	claude-sonnet-5	3/5	No strict licensing requirement for advising, but strong organizational and student expectation for a human faculty mentor, plus liability concerns around career guidance, create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Faculty advising labor is embedded in salary structures; the cost of an AI system plus institutional integration and human oversight to validate recommendations is unlikely to be substantially cheaper than the marginal cost of a professor providing advice during office hours.
Cost vs. human wage	claude-sonnet-5	2/5	While AI query tools are cheap, the actual value in this task comes from personalized mentorship and institutional knowledge that still requires paid faculty time, keeping cost savings modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Chatbots and career-guidance tools exist but operate at a shallow, information-retrieval level; they lack the contextual judgment, follow-up, and individualized assessment needed for genuine academic and vocational advising. Production deployments in education show these systems supplement rather than perform the task reliably.
Technical feasibility today	claude-sonnet-5	2/5	Chatbot advising tools exist in some universities for basic scheduling/curriculum questions, but nuanced career and academic advising by faculty remains largely undeployed as an AI-driven product.

Keep abreast of developments in the field by reading current literature, talking with colleagues, and participating in professional conferences.

CI 16–38 · exposure 17 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While some academics use AI tools for literature discovery and summarization, full replacement of active reading and conference participation remains rare and culturally resisted. Adoption is occurring for assistance rather than substitution in higher education.
Sector adoption velocity	claude-sonnet-5	3/5	Academia is a mixed-adoption sector; AI research assistants and summarization tools are gaining traction among academics, though full integration into professional development routines is still emerging.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can meaningfully augment this task by identifying relevant papers, summarizing findings, and flagging emerging topics, allowing faculty to focus on deeper synthesis and selective engagement. Many academics already use AI-powered literature tools to increase the breadth they can monitor while preserving their own judgment.
Augmentation potential	claude-sonnet-5	4/5	AI significantly aids literature review, summarization, and staying current by filtering and synthesizing large volumes of research, meaningfully boosting efficiency while the human remains engaged in networking and critical evaluation.
Task automatability	claude-haiku-4-5-20251001	1/5	Staying current requires judgment about which developments are significant, synthesis across disparate sources, and genuine peer engagement. While AI can summarize papers or identify trends, it cannot independently determine what a specific educator needs to know or meaningfully participate in collegial dialogue.
Task automatability	claude-sonnet-5	2/5	AI can help surface and summarize literature, but the core activities of ongoing learning, networking, and conference participation require sustained personal engagement and judgment that current systems cannot fully replace end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Professional development and staying current are expectations embedded in faculty roles and tenure/promotion criteria. Institutions and colleagues expect the faculty member themselves to engage with the field; outsourcing this entirely conflicts with professional norms and the legitimacy of one's expertise.
Adoption barriers	claude-sonnet-5	2/5	No licensing barrier prevents AI assistance, but professional norms and value placed on human networking/conference participation create moderate organizational and cultural friction against substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for literature monitoring and summarization exist but require subscription costs and human oversight to filter and interpret results. Total cost per unit of useful insight is comparable to or may exceed the time cost of a faculty member's selective reading and professional participation.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools for literature scanning are cheap, but since the task cannot be fully automated, oversight and human engagement costs remain, keeping overall cost comparable to or only marginally better than a human doing it directly.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI can assist with literature aggregation and summarization (existing systems do this), but no deployed product reliably identifies relevant developments for a specific researcher or authentically participates in professional conferences on their behalf. The human judgment component remains essential.
Technical feasibility today	claude-sonnet-5	2/5	Products like AI-powered literature summarizers and research digest tools exist but are used as supplements, not deployed to autonomously perform this ongoing professional development task.

Plan, evaluate, and revise curricula, course content, and course materials and methods of instruction.

CI 25–25 · exposure 25 · augmentation 63 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Postsecondary education remains one of the slowest-adopting sectors for AI-driven automation due to shared governance, faculty autonomy norms, and accreditation constraints. While AI-assisted content tools are emerging in pilots, production replacement of curriculum planning is rare and limited to minor supplementary tasks.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools unevenly and cautiously, with curriculum decisions especially slow due to committee-based governance and accreditation cycles.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist faculty by drafting course outlines, suggesting content structures, analyzing student performance data, and generating assessment rubrics, thereby raising productivity in parts of the planning and evaluation workflow. However, the human faculty member must remain central to synthesizing, validating, and making strategic trade-offs.
Augmentation potential	claude-sonnet-5	4/5	AI tools substantially help instructors draft syllabi, generate practice problems, suggest readings, and revise course materials faster, while faculty retain final curricular judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist in generating course outlines, content summaries, and evaluation rubrics, curriculum planning requires deep understanding of learning outcomes, institutional constraints, accreditation standards, and student needs that demands sustained human judgment. Current AI systems cannot reliably handle the end-to-end synthesis and strategic decisions needed to meet the ≥50% time-saving bar at equal quality.
Task automatability	claude-sonnet-5	2/5	AI can draft syllabi and suggest materials, but planning and revising curricula requires institutional judgment, accreditation alignment, and pedagogical expertise that current systems cannot fully replace end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Curriculum decisions are subject to accreditation standards, institutional governance, and faculty senate review requirements. Many institutions have formal processes requiring human faculty sign-off and approval, and regulatory bodies (e.g., engineering accreditation boards) maintain strict oversight of course content and learning outcomes, creating substantial legal and organizational barriers to full automation.
Adoption barriers	claude-sonnet-5	4/5	Curriculum design typically requires faculty governance, accreditation body approval (e.g., ABET), and departmental sign-off, creating strong institutional and regulatory barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI content-generation and analysis tools are inexpensive, but the cost of human oversight, validation, and rework to ensure curriculum quality and institutional fit typically exceed the savings from automation, especially given the liability and reputational stakes.
Cost vs. human wage	claude-sonnet-5	2/5	While AI can cheaply generate draft content, the human faculty oversight, subject-matter expertise, and institutional review needed still dominate cost, keeping overall savings modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools exist for content generation and draft materials, but no deployed product reliably performs the full curriculum-planning and revision cycle independently. Educational institutions still rely on human faculty committees and subject-matter experts to validate and own curriculum changes, reflecting the high stakes and context-dependency of this work.
Technical feasibility today	claude-sonnet-5	2/5	AI writing tools are used to draft course outlines and content suggestions, but no deployed product independently plans and revises full engineering curricula in production at universities.

Provide professional consulting services to government or industry.

CI 25–25 · exposure 25 · augmentation 75 · importance 2.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Adoption of AI in professional consulting is primarily assistive (research, drafting) rather than replacement. High-stakes consulting remains largely human-driven; pilots are common but production-level displacement of the consulting role itself is minimal and slow.
Sector adoption velocity	claude-sonnet-5	2/5	Postsecondary engineering faculty and technical consulting fields show only moderate, uneven AI adoption, largely for research support rather than replacing consulting engagements.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly augments engineering professors and consultants by accelerating literature review, generating initial design analyses, and drafting technical reports. These tools enhance productivity and breadth of work while the human consultant retains judgment and client-facing authority.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully speed up literature review, data analysis, and report drafting for consulting engagements, significantly aiding the professional while they retain responsibility for judgment and client interaction.
Task automatability	claude-haiku-4-5-20251001	2/5	Professional consulting for engineering requires deep contextual judgment, stakeholder management, and synthesis of complex domain knowledge. While AI can assist with research, analysis, and drafting recommendations, the core task of advising on strategic decisions and standing behind advice demands human expertise and accountability that AI cannot fully replace.
Task automatability	claude-sonnet-5	2/5	Consulting requires synthesizing specialized engineering expertise, contextual judgment, and client-specific problem-solving that current AI cannot reliably replicate end-to-end.dulze AI can support research and drafting but not the substantive consulting itself.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Government and industry clients typically require a licensed professional engineer or recognized domain expert to sign off on technical advice; liability and regulatory frameworks (particularly in safety-critical domains) create strong legal and organizational barriers to full AI automation of consulting deliverables.
Adoption barriers	claude-sonnet-5	4/5	Professional engineering consulting often requires credentialed expertise, liability for advice given, and in many cases licensed PE sign-off, creating strong barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The loaded cost of a senior engineering professor or consultant ($100k–$300k+ annually, fully loaded) far exceeds the inference cost of AI, but the latter cannot substitute for the credibility, liability absorption, and relationship capital that a human consultant provides, making the ratio unfavorable for AI displacement.
Cost vs. human wage	claude-sonnet-5	2/5	Because a qualified human expert must still validate and stand behind the recommendations, AI at best reduces some drafting/research time rather than replacing the paid consulting engagement, keeping costs comparable overall.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably performs end-to-end engineering consulting independently. AI can support analysis and generate draft reports, but deployed consulting remains human-led; AI tools serve as assistants rather than primary consultants in production environments.
Technical feasibility today	claude-sonnet-5	2/5	AI tools exist for technical research assistance and report drafting, but no deployed product independently provides authoritative engineering consulting to government or industry clients.

Prepare and deliver lectures to undergraduate or graduate students on topics such as mechanics, hydraulics, and robotics.

CI 14–30 · exposure 25 · augmentation 63 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Higher education has demonstrated slow adoption of automation for core teaching roles due to organizational structure, labor agreements, and institutional resistance to replacing faculty with machines. Pilots of AI-assisted content exist, but widespread deployment of AI-led instruction remains negligible.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools for course prep and assistance, but actual lecture delivery automation remains rare and cautious due to academic norms and quality concerns.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist instructors by generating draft lecture notes, creating visualizations of engineering concepts, and answering follow-up student questions in a tutoring system; however, the primary task of live delivery remains human-centered and augmentation is partial rather than transformative.
Augmentation potential	claude-sonnet-5	4/5	AI substantially aids lecture preparation, generating explanations, examples, diagrams, and practice problems for topics like mechanics or robotics, boosting instructor productivity.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can generate lecture slides and outline content on technical topics, delivering lectures requires real-time student engagement, adaptive explanation, and live problem-solving that current systems cannot reliably replicate. AI may automate components like slide generation, but cannot meaningfully replace the 50% threshold for full lecture delivery.
Task automatability	claude-sonnet-5	2/5	AI can help draft lecture content and slides, but live delivery, adapting to student questions, and pedagogical presence require human execution, so end-to-end automation with equal quality is not yet achieved.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal and regulatory barriers exist: accreditation bodies, institutional governance, and faculty contracts typically require that instruction be delivered by credentialed faculty. Institutions face reputational and legal risk from automated instruction, creating hard barriers to substitution.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement for lecturing itself, but accreditation standards, tenure norms, and student/institutional expectations for human instructors create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of building, integrating, and maintaining an AI system capable of delivering lectures (including oversight and error correction) would substantially exceed the loaded cost of a faculty member, especially considering liability and quality assurance requirements.
Cost vs. human wage	claude-sonnet-5	2/5	Preparing lecture materials with AI is cheap, but full delivery still requires a paid faculty member, so overall cost savings versus a professor's salary are limited.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably delivers full lectures to engineering students end-to-end. AI chatbots and content generators exist but lack the ability to manage classroom dynamics, respond to nuanced technical questions in real-time, or adapt explanations based on student comprehension signals in a live setting.
Technical feasibility today	claude-sonnet-5	2/5	Some universities use AI-generated content or recorded/AI-avatar lectures experimentally, but no mature product reliably delivers full engineering lectures in production at scale.

Collaborate with colleagues to address teaching and research issues.

CI 11–30 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Higher education has been slow to adopt AI agents for core academic governance and collaboration. Pilots exist for administrative support, but production adoption of AI in faculty deliberation remains minimal; institutions remain conservative on delegating collegial processes.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools slowly for collaborative and governance-related academic work, though usage of AI for individual productivity tasks is growing.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully augment collaboration by drafting meeting summaries, retrieving relevant research, or organizing discussion points, helping colleagues work more efficiently. However, the core interpersonal and creative elements limit how much AI can amplify human productivity on the central task.
Augmentation potential	claude-sonnet-5	3/5	AI can assist by summarizing research literature, drafting meeting agendas, or synthesizing curriculum ideas, improving productivity within collaborative discussions.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can assist with organizing meeting agendas, synthesizing literature, and drafting collaborative documents, but cannot replace the nuanced interpersonal negotiation, judgment, and creative problem-solving required to genuinely address teaching and research issues with colleagues. The task is fundamentally dialogical and requires human authority and accountability.
Task automatability	claude-sonnet-5	1/5	Collegial collaboration on teaching and research strategy requires relationship-building, institutional context, and real-time judgment that current AI cannot replace end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: academic governance structures mandate faculty participation in shared decision-making; institutional norms require colleagues to own collaborative outcomes; and implicit contractual obligations tie instructor roles to collegial participation in departmental problem-solving. Legal authority and accountability cannot be delegated to AI.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement blocks AI involvement, but strong organizational and cultural norms around faculty governance and peer collaboration create friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools can reduce preparation costs for collaboration (drafting agendas, notes), but the core task—synchronous, interactive problem-solving between humans—still requires paid colleague time. Integration overhead and the need for human judgment mean overall costs remain dominated by human labor.
Cost vs. human wage	claude-sonnet-5	2/5	Since AI cannot perform the core interpersonal collaboration, there is no meaningful cost substitution; only marginal costs saved on ancillary documentation.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can generate discussion prompts or summarize prior collaboration patterns, no deployed product reliably facilitates real colleague collaboration on substantive academic issues. This task requires presence, mutual understanding, and negotiated commitment that current AI systems cannot independently achieve.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product substitutes for human collaborative discussion among faculty on pedagogy or research direction; AI at most supports scheduling or note-taking.

Conduct research in a particular field of knowledge and publish findings in professional journals, books, or electronic media.

CI 7–28 · exposure 17 · augmentation 75 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While academia uses AI tools for writing support and literature management, actual automation of research and publication remains minimal in practice. Adoption is largely assistive (draft editing, search) rather than generative of independent research output or displacement of researchers.
Sector adoption velocity	claude-sonnet-5	3/5	Academia has moderate AI adoption for literature review, coding, and writing assistance, but full research automation in production is rare and cautious due to integrity concerns.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly augments researcher productivity through literature search, rapid hypothesis exploration, manuscript drafting, and statistical analysis assistance, allowing researchers to focus on conceptualization and interpretation while staying in the loop for all critical decisions.
Augmentation potential	claude-sonnet-5	4/5	AI tools significantly help with literature review, data analysis, drafting manuscripts, and coding, meaningfully boosting researcher productivity while humans retain oversight and originality.
Task automatability	claude-haiku-4-5-20251001	1/5	Research discovery, hypothesis formation, and novel knowledge generation fundamentally require human creativity, domain expertise, and scientific judgment. AI systems can assist with literature review and drafting, but cannot independently conduct original research or generate publishable findings that meet peer review standards.
Task automatability	claude-sonnet-5	2/5	AI can assist with literature review, data analysis, and drafting but cannot independently design novel engineering experiments, generate original insights, or conduct hands-on lab/field research end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Academic publishing has strong institutional barriers: peer review requires human expert evaluation, authorship carries legal/ethical accountability, institutional review boards govern research protocols, and professional reputation depends on personal intellectual contribution—all legally and culturally tied to the human researcher.
Adoption barriers	claude-sonnet-5	4/5	Academic credentialing, peer review, authorship ethics, and institutional research standards create strong barriers against AI substituting for the credited researcher.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The overhead of AI-assisted research (literature synthesis, drafting support, verification oversight) combined with the irreplaceable human expertise needed for novel research direction and interpretation makes the cost-benefit unfavorable compared to researcher wages for original work.
Cost vs. human wage	claude-sonnet-5	2/5	Substantial human expertise, experimentation, and peer-reviewed validation are still required, so AI only reduces costs for narrow sub-tasks like literature synthesis, not the full research cycle.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can draft sections of papers and help organize literature, no deployed system can autonomously conduct original research and produce publication-ready manuscripts. Current products lack the ability to design experiments, interpret ambiguous results, and make the theoretical contributions expected in peer-reviewed venues.
Technical feasibility today	claude-sonnet-5	2/5	Products like AI writing and literature-search tools are used in research support, but no deployed system autonomously conducts original engineering research and publishes it reliably.

Initiate, facilitate, and moderate class discussions.

CI 14–16 · exposure 16 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Postsecondary institutions show minimal adoption of AI-led classroom facilitation; discussions remain synchronous, human-led, and faculty-owned. Pilot projects are rare and mostly experimental, with no evidence of production displacement in mainstream engineering education.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools for content creation and grading assistance, but live classroom facilitation remains largely untouched by AI adoption trends.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can augment by generating discussion questions, tracking and summarizing threads in asynchronous forums, or identifying common student misconceptions from written responses—but the instructor remains essential for steering the discussion and making pedagogical judgments in real time.
Augmentation potential	claude-sonnet-5	3/5	AI can help generate discussion questions, summarize prior conversations, or suggest talking points, providing moderate support to instructors preparing for or supplementing discussions.
Task automatability	claude-haiku-4-5-20251001	2/5	Moderating live class discussions requires real-time understanding of context, student dynamics, and nuanced interpersonal judgment. While AI can help generate discussion prompts or summarize threads asynchronously, it cannot reliably facilitate the spontaneous, adaptive social-cognitive work that defines live classroom moderation today.
Task automatability	claude-sonnet-5	2/5	Facilitating live, dynamic classroom discussion requires real-time social judgment, reading student engagement, and adaptive pedagogy that current AI cannot reliably replicate end-to-end.'},
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: institutional accreditation and quality standards typically require a credentialed faculty member to take responsibility for course content and student learning outcomes, and the human-contact and relationship-building aspects of classroom discussion are often viewed as integral to the educational mission.
Adoption barriers	claude-sonnet-5	4/5	Postsecondary teaching typically requires an in-person, credentialed instructor to lead discussion, with strong institutional and accreditation expectations for human-led instruction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Integration costs, real-time monitoring, and inevitable human oversight to catch errors, tone-deafness, or pedagogical missteps would exceed the cost of an instructor already present and compensated to teach the class.
Cost vs. human wage	claude-sonnet-5	1/5	Since AI cannot perform the live facilitation role, there is no viable cost comparison—human instructor cost remains necessary for the core task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably initiates, facilitates, and moderates live academic class discussions as an end-to-end substitute. Chatbots can participate in written forums with significant moderator oversight, but cannot independently manage the interpersonal and pedagogical demands of real-time classroom facilitation.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously moderates live in-person engineering classroom discussions; existing tools are limited to online forum moderation or discussion prompts, not real classroom facilitation.

Supervise undergraduate or graduate teaching, internship, and research work.

CI 7–16 · exposure 5 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Higher education adopts AI selectively for administrative tasks but remains conservative about automating core supervisory and mentoring responsibilities; regulatory and accreditation requirements slow adoption of any system that would displace faculty oversight.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools slowly for core supervisory and mentorship functions, though pilots for administrative support exist; the substantive supervisory role sees minimal AI penetration.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist faculty by automating rubric application, tracking student milestones, organizing feedback summaries, or flagging at-risk students, meaningfully raising administrative efficiency while the faculty member remains the decision-maker.
Augmentation potential	claude-sonnet-5	3/5	AI can help draft feedback, track progress, or suggest research resources, giving moderate assistance, but the core supervisory judgment and mentoring remain human-driven.
Task automatability	claude-haiku-4-5-20251001	1/5	Supervising teaching, internship, and research work requires ongoing human judgment, mentorship, assessment of student/trainee progress, and adaptive guidance that cannot be meaningfully automated end-to-end today. AI cannot replace the relationship-based feedback and accountability essential to this task.
Task automatability	claude-sonnet-5	1/5	Supervising students' hands-on teaching, internships, and research requires ongoing personalized mentorship, evaluation of judgment, and relationship-building that current AI cannot autonomously perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: faculty are professionally and legally accountable for student and trainee supervision; institutions require credentialed humans to evaluate research and sign off on degree requirements, making full automation infeasible regardless of technical capability.
Adoption barriers	claude-sonnet-5	4/5	Accreditation, institutional policy, and mentorship/liability norms generally require a qualified faculty member to formally supervise and evaluate student research and teaching.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of oversight infrastructure, liability, and human verification would likely exceed or equal the savings from partial AI assistance, since supervision itself remains a human responsibility that cannot be fully delegated.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this supervisory role, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While AI can assist with grading rubrics, scheduling, or documentation, no deployed product reliably supervises the full spectrum of undergraduate/graduate work, internships, and research—this requires human accountability and professional judgment that remains a legal and ethical requirement.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product substitutes for a faculty supervisor overseeing students' practical and research work; this remains entirely research-stage or nonexistent as an automation product.

Supervise students' laboratory work.

CI 0–16 · exposure 8 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Higher education institutions have not adopted AI for primary laboratory supervision; the task remains firmly in human hands due to safety requirements and traditional pedagogy. Adoption velocity is near-zero.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education is adopting AI for teaching materials and grading, but physical lab supervision remains largely untouched by automation trends.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI could assist by monitoring for common errors in real-time, providing instructors with alerts or data dashboards about equipment status or protocol compliance, moderately enhancing the supervisor's effectiveness without removing human oversight.
Augmentation potential	claude-sonnet-5	3/5	AI can help prepare lab materials, answer student questions, analyze data, or provide feedback on reports, augmenting the instructor's overall workload even though direct supervision isn't automated.
Task automatability	claude-haiku-4-5-20251001	2/5	AI could assist with monitoring equipment operation and flagging protocol deviations through video/sensor data, but cannot replace the real-time safety supervision, adaptive instruction, and human judgment required for hands-on laboratory work. The task inherently requires human presence and decision-making.
Task automatability	claude-sonnet-5	1/5	Supervising hands-on lab work requires physical presence, real-time safety monitoring, and hands-on correction of student technique with equipment—something current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Institutional liability, legal duty of care, and safety regulations mandate that a licensed human instructor directly supervise laboratory work. No amount of AI assistance can remove the human requirement due to student safety and institutional accountability.
Adoption barriers	claude-sonnet-5	4/5	Safety regulations, liability for lab accidents, and institutional requirements for qualified faculty/TA oversight of hazardous equipment create strong barriers to removing human supervision.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of deploying reliable AI monitoring systems (cameras, sensors, software, integration, liability coverage) across laboratory facilities exceeds the salary cost of a single instructor supervising a cohort.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute for physical lab supervision, so cost comparison favors the human by default since AI cannot perform the core task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed products reliably perform full laboratory supervision today. While computer vision systems can detect some violations, no production system handles the dynamic, safety-critical aspects of supervising student experiments at scale.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product supervises physical engineering labs; at best AI can assist with pre-lab quizzes or report grading, not live supervision.

Maintain regularly scheduled office hours to advise and assist students.

CI 0–16 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Universities move slowly on replacing core instructional and advising functions; while some adopt automated scheduling and chatbots for routine questions, systematic replacement of office-hours advising is not observed in practice.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI advising tools slowly and unevenly, mostly for administrative FAQs, with in-person office hours remaining the norm.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could provide modest assistance—scheduling tools, pre-session student-data summaries, or post-session documentation drafts—but the core advising role remains human-driven and the productivity gains from such augmentation are limited.
Augmentation potential	claude-sonnet-5	3/5	AI can help by answering routine questions beforehand, scheduling, or providing supplementary resources, freeing office hours for higher-value discussion, but the core task remains human-led.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires real-time, personalized interaction with individual students to provide tailored academic and professional guidance, which is fundamentally resistant to automation. Current AI systems cannot reliably replicate the human judgment, contextual understanding, and empathetic presence that characterize effective office hours.
Task automatability	claude-sonnet-5	1/5	This task requires a human presence for personalized, relationship-based advising and mentorship, which AI cannot substitute for in a way that meets equal quality with 50% time savings.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong barriers protect this task: accreditation and institutional governance requirements typically mandate faculty office hours as a direct advisor contact, students have legitimate expectations for human interaction, and liability concerns around academic guidance favor human sign-off.
Adoption barriers	claude-sonnet-5	3/5	No formal licensing requirement bars AI assistance, but institutional norms, student expectations of personal mentorship, and accreditation standards around faculty accessibility create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	An AI system providing comparable office-hour functions (scheduling, monitoring, integration with student information systems, oversight) would likely cost more than the faculty time currently deployed, given the need for custom training and ongoing refinement.
Cost vs. human wage	claude-sonnet-5	2/5	While AI chat tools are cheap, they cannot fully replace the task, so any AI deployment is supplementary rather than substitutive, keeping the human cost largely intact.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs this task end-to-end; office hours require sustained presence, dynamic response to individual circumstances, and trust-building that current AI chatbots or automated advising systems cannot meet at production scale.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product replaces a professor's scheduled office hours for personal academic advising; chatbots exist for FAQs but not for the substantive human interaction this task implies.

Act as advisers to student organizations.

CI 5–5 · exposure 0 · augmentation 25 · importance 2.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Higher education is a relatively laggard sector for AI adoption in core instructional and mentoring roles; student organization advising is deeply embedded in faculty identity and institutional culture with no meaningful automation trend.
Sector adoption velocity	claude-sonnet-5	1/5	Higher education advising roles show minimal AI displacement; this is a low-digitization, relationship-driven aspect of academic work.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with administrative tasks (scheduling, document management, capturing meeting notes), but current systems offer limited augmentation for the core advisory work—listening, judgment, mentorship—that defines this role.
Augmentation potential	claude-sonnet-5	2/5	AI can help with scheduling, drafting communications, or budgeting suggestions for the organization, but offers limited assistance to the core mentoring and advisory relationship.
Task automatability	claude-haiku-4-5-20251001	1/5	Advising student organizations requires interpersonal judgment, mentorship, conflict resolution, and understanding of individual student needs and institutional context—nuanced human capabilities that current AI systems cannot perform end-to-end with 50% time savings.
Task automatability	claude-sonnet-5	1/5	Advising student organizations involves mentorship, relationship building, institutional judgment, and personal presence that current AI cannot replicate end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Universities and accrediting bodies expect licensed faculty to serve as advisers to student organizations; there is both institutional expectation and often explicit governance requirements that a human faculty member hold advisory responsibilities.
Adoption barriers	claude-sonnet-5	4/5	Institutions typically require a designated faculty member to serve as an official advisor for liability, oversight, and accreditation purposes, creating a strong structural barrier.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The task is intrinsically bound to human presence and credibility; even if AI could assist with administrative tasks, the core advisory function cannot be cost-effectively replaced by AI inference alone, making the all-in cost favor the human.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this role, so cost comparison favors the human by default since AI cannot deliver the output.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs the advisory role for student organizations; AI cannot substitute for the trusted human relationship and institutional authority that students require when seeking guidance on organization leadership, direction, and problems.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product performs the role of a faculty organizational advisor; this remains entirely a human relational and administrative function.

Participate in campus and community events.

CI 0–5 · exposure 0 · augmentation 13 · importance 2.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	No meaningful AI adoption is occurring in this domain because the task is fundamentally incompatible with automation. Academic institutions continue to require personal faculty participation in events as part of professional and community obligations.
Sector adoption velocity	claude-sonnet-5	1/5	Higher education community engagement is a low-digitization, relationship-driven activity with essentially no AI adoption trend for physical event participation.
Augmentation potential	claude-haiku-4-5-20251001	1/5	AI offers minimal assistance to faculty participating in campus or community events; the task itself is about presence and engagement rather than information processing or decision support. Augmentation tools have negligible value for this human-centered activity.
Augmentation potential	claude-sonnet-5	2/5	AI can help with scheduling, event promotion materials, or follow-up communications, but offers minimal help with the core act of attending and engaging in person.
Task automatability	claude-haiku-4-5-20251001	1/5	Participating in campus and community events requires human presence, social interaction, relationship-building, and contextual judgment that AI cannot currently perform end-to-end. These are fundamentally human activities involving networking, mentorship, and community engagement that resist automation.
Task automatability	claude-sonnet-5	1/5	Participating in physical campus and community events (attending, networking, representing the department) is an in-person social/relational activity that AI cannot perform.task requires embodied human presence.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong barriers protect this task: institutional expectations that faculty participate in person, professional norms requiring human presence, and the intrinsic requirement for a human representative of the department/university. Organizational culture and mission deeply embed this expectation.
Adoption barriers	claude-sonnet-5	4/5	Institutional expectations, professional norms, and representation duties require a human faculty member's physical presence and social capital, creating strong organizational barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI has no economic advantage here since the task cannot be meaningfully automated. The human must attend personally, making the cost ratio heavily favoring human performance as the only viable option.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this task, so no cost comparison favors AI; the human cost is the only viable option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product can autonomously participate in campus or community events; this task is inherently tied to human presence and professional judgment. Current AI cannot replicate the social and relational work required.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product attends events or represents a person in community engagement; this is not a product category that exists.

Perform administrative duties, such as serving as department head.

CI 0–3 · exposure 0 · augmentation 38 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Higher education has lagged in administrative automation, especially for leadership roles. Institutional conservatism, governance structures, and the requirement for human accountability mean adoption of AI for department head duties remains near zero in practice.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administration adopts AI tools slowly for decision support, but the leadership/headship role itself sees essentially no displacement trend.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI can assist with ancillary tasks like meeting scheduling, memo drafting, or report compilation, but augmentation is limited because the core duties—hiring, evaluation, strategy, conflict resolution—remain quintessentially human and cannot be substantially enhanced by AI assistance without delegating actual authority.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with scheduling, drafting reports, summarizing data, and handling routine correspondence that supports the administrative aspects of the role, improving efficiency without replacing leadership duties.
Task automatability	claude-haiku-4-5-20251001	1/5	Department head duties involve strategic decision-making, personnel management, budget oversight, and institutional policy-making that require human judgment, accountability, and authority. Current AI cannot autonomously perform these functions end-to-end, nor can it sign binding institutional decisions or represent the department in governance contexts.
Task automatability	claude-sonnet-5	1/5	Serving as department head involves personnel management, strategic decisions, budgeting, conflict resolution, and interpersonal leadership that cannot be executed end-to-end by AI systems today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Department head positions are legally and institutionally tied to specific individuals who hold fiduciary, hiring, and governance authority. Educational institutions require a licensed faculty member in this role by design, creating a hard regulatory and structural barrier to automation.
Adoption barriers	claude-sonnet-5	5/5	Department head roles require formal institutional appointment, accountability, signing authority, and often tenure/faculty governance structures that legally and organizationally require a human in the position.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	A department head's salary is typically $80k–$150k+ annually. AI systems cannot replace this role, and the overhead of using AI for ancillary tasks (scheduling, report generation) would add cost rather than reduce it relative to human execution of the full role.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this leadership function, so no meaningful cost comparison to a human department head exists—AI cannot replace the role at any cost.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs the full scope of department head responsibilities in production. AI can assist with scheduling or data aggregation, but actual leadership, hiring decisions, conflict resolution, and budget authority remain non-automatable and require a human administrator.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs the role of department head; this is an inherently human leadership and administrative role requiring judgment, authority, and accountability.

Serve on academic or administrative committees that deal with institutional policies, departmental matters, and academic issues.

CI 0–0 · exposure 0 · augmentation 38 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Academic governance remains deeply rooted in human deliberation and cannot be displaced by AI; adoption patterns show no movement toward AI committee participation even in highly digitized institutions.
Sector adoption velocity	claude-sonnet-5	1/5	Higher education governance structures are slow-changing and there is no observable trend of AI systems replacing committee members.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could modestly assist by summarizing prior meeting minutes, drafting policy language, or organizing committee materials, but the core deliberative and decision-making work remains human-centered with limited opportunity for AI to enhance productivity in meaningful ways.
Augmentation potential	claude-sonnet-5	3/5	AI can help prepare briefing materials, summarize policy documents, draft meeting minutes, or analyze data to inform committee discussions, aiding the human's preparation and follow-through.
Task automatability	claude-haiku-4-5-20251001	1/5	Committee service requires nuanced judgment on institutional policy, negotiation among stakeholders with competing interests, and accountability for decisions affecting faculty and students. Current AI cannot meaningfully participate in deliberative processes or bear responsibility for governance decisions.
Task automatability	claude-sonnet-5	1/5	Committee service requires deliberation, negotiation, political judgment, and institutional relationship-building among human stakeholders, which AI cannot perform on someone's behalf today.'
Adoption barriers	claude-haiku-4-5-20251001	5/5	Strong legal and governance barriers protect this task: faculty committees derive legitimacy and accountability from human members with official standing. Institutional bylaws and accreditation standards typically require human judgment and legal responsibility for policy decisions.
Adoption barriers	claude-sonnet-5	5/5	Committee membership requires institutional authority, faculty governance status, and often tenure/rank-based eligibility; only a qualified human can be appointed to serve.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Committee service is a human governance responsibility that cannot be cost-effectively replaced by AI; the task requires someone with institutional standing and legal accountability. There is no meaningful AI alternative to offset the cost of human participation.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI alternative to human committee membership, so cost comparison is moot—the human role cannot be replaced by any AI-priced service.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed system can autonomously serve on academic committees or represent institutional interests in policy deliberation. This task fundamentally requires human judgment, accountability, and contextual understanding of organizational dynamics that AI systems cannot replicate in production.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product substitutes for a person's presence and voice on academic/administrative committees; AI has no role as a committee member.

Related occupations — Educational Instruction & Library

How to read this

A high substitution score does not mean this job disappears — it means a large share of its current tasks face replacement pressure, so the mix of tasks is likely to change. High augmentation alongside substitution typically means the occupation reorganizes around the protected tasks. Wide confidence intervals mean the rater panel disagreed: treat those scores as open questions, not verdicts.

What would change this score

New model capabilities (automatability, feasibility), falling inference costs (cost ratio), regulation and licensing shifts (barriers), and measured sector adoption (velocity) all re-enter at every index release. Each release is recomputed, versioned and kept queryable — scores are claims with a date on them, not permanent labels.