Anthropology and Archeology Teachers, Postsecondary

25-1061.00

Median wage $99,650/yr5,240 employed (US)Rank #383 of 923 scored · top 41% by substitution

Teach courses in anthropology or archeology. 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

Augmentation60

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

27 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

11%

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%26

panel mean rating 2.0/5 → substitution pressure 26/100

Cost vs. human wagew 15%36

panel mean rating 2.4/5 → substitution pressure 36/100

Adoption barriersw 20%inverted — strong barriers lower the score36

panel mean rating 3.6/5 (barrier strength) → substitution pressure 36/100

Sector adoption velocityw 10%26

panel mean rating 2.0/5 → substitution pressure 26/100

Task breakdown (27 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 95–95 · exposure 100 · 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	5/5	Postsecondary institutions are highly digitized and have been adopting integrated student-information and learning-management systems for decades. AI-assisted record-keeping is now standard across universities, colleges, and online programs.
Sector adoption velocity	claude-sonnet-5	5/5	Higher education has near-universal adoption of digital LMS platforms for attendance and grade recording, a mature and long-standing practice.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI substantially assists faculty and admins by auto-populating, validating, flagging anomalies (e.g., unusual attendance patterns), and generating reports, freeing humans from manual data entry while keeping them in oversight. This augmentation is already widespread.
Augmentation potential	claude-sonnet-5	4/5	AI-enhanced LMS tools help instructors track trends, flag at-risk students, and auto-calculate grades, meaningfully reducing administrative burden while the instructor retains oversight.
Task automatability	claude-haiku-4-5-20251001	5/5	Record-keeping of attendance, grades, and student data is entirely structured, digital, and repetitive—core strengths of current AI and learning management systems. AI can extract, verify, input, and maintain these records with 100% accuracy and near-zero latency, easily exceeding the 50% time-saving threshold.
Task automatability	claude-sonnet-5	5/5	Recording attendance and grades into gradebooks/LMS is a structured data-entry task fully handled by existing learning management systems and spreadsheet automation with no quality loss.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While FERPA and other privacy regulations govern access to records, they do not legally require a human to maintain them—only to protect and audit them. Integration into existing institutional systems requires some organizational setup, but substitution itself faces minimal legal or liability barriers.
Adoption barriers	claude-sonnet-5	2/5	Some institutional policy requires instructor sign-off on final grades, but the mechanical recordkeeping itself faces minimal regulatory or licensing barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	LMS tools with AI-assisted record management cost pennies per student per term, while a human administrator or assistant managing equivalent records costs thousands annually. The ratio is at least 10–100x in AI's favor.
Cost vs. human wage	claude-sonnet-5	5/5	Automated recordkeeping via existing software costs a tiny fraction of the faculty time it would take to manually maintain these records.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Student information systems (Blackboard, Canvas, Workday) with AI-assisted data entry and automated flagging are deployed at scale across virtually all postsecondary institutions. These systems reliably handle millions of student records daily in production.
Technical feasibility today	claude-sonnet-5	5/5	LMS platforms like Canvas, Blackboard, and Google Classroom already automate attendance tracking, gradebook calculations, and record storage in production at virtually every university.

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

CI 76–84 · exposure 75 · 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	4/5	Higher-education institutions are rapidly piloting and adopting AI-assisted course material generation; many faculty already use ChatGPT for syllabus drafting and assignment design. Adoption is measurably faster in information-rich, digitized sectors like postsecondary education.
Sector adoption velocity	claude-sonnet-5	3/5	Higher education is adopting generative AI for course prep at a moderate pace—many individual faculty use it informally, but institutional integration and policy remain uneven and cautious.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI strongly augments instructor productivity by handling drafting, formatting, and iteration, freeing faculty to focus on domain expertise, pedagogical innovation, and customization. The human remains in creative and quality-control roles while time-to-completion drops substantially.
Augmentation potential	claude-sonnet-5	5/5	AI is highly effective as a drafting assistant for syllabi, assignments, and handouts, letting instructors iterate quickly while retaining control over final content and academic judgment.
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 the 50% time-saving threshold. Current LLMs excel at structuring educational content, creating learning objectives, and drafting assignments; human review for subject-specific accuracy and pedagogical fit remains necessary but represents a minority of effort.
Task automatability	claude-sonnet-5	4/5	Generating syllabi, homework assignments, and handouts from a course description or textbook is well within current LLM capability, requiring mainly instructor review and customization rather than full creation from scratch.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No licensing or legal requirement mandates that a human prepare course materials; universities and faculty retain discretion. Minimal barrier exists beyond faculty preference for creative control and institutional norms favoring human authorship of pedagogical content.
Adoption barriers	claude-sonnet-5	1/5	There is no licensing or legal requirement that a human personally draft syllabi or handouts; faculty have full discretion to use AI tools with no regulatory obstacle.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI inference cost for generating a full syllabus or assignment set is negligible (cents to dollars), while a faculty member's loaded hourly cost to create equivalent materials from scratch is $50–150+, yielding a cost advantage of one to two orders of magnitude.
Cost vs. human wage	claude-sonnet-5	5/5	Generating drafts of course materials via AI costs a few cents to dollars in compute versus significant instructor hours, an order-of-magnitude cost reduction for the drafting portion of the task.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Multiple deployed products (ChatGPT, Claude, specialized EdTech platforms) reliably generate course materials in production. Error rates are low for structural and formatting tasks; subject-matter accuracy requires review but the baseline output is professional and deployable with modest oversight.
Technical feasibility today	claude-sonnet-5	4/5	Products like ChatGPT, Claude, and specialized ed-tech tools are already used widely by instructors to draft syllabi and assignments, though instructors still edit for accuracy, institutional policy, and pedagogical fit.

Compile bibliographies of specialized materials for outside reading assignments.

CI 71–81 · exposure 70 · augmentation 100 · 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	3/5	Academic adoption is growing (LLMs now used in course prep), but many institutions still rely on manual processes or traditional library services. Adoption is faster in research-heavy and well-resourced universities, slower in small colleges and abroad, reflecting middling overall sectoral velocity.
Sector adoption velocity	claude-sonnet-5	3/5	Higher education is adopting AI research tools unevenly; some faculty use AI-assisted literature tools while others are cautious, so adoption is moderate and uneven across academia.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI dramatically assists faculty by suggesting sources, organizing by theme, catching gaps, and handling formatting—allowing instructors to focus on pedagogical judgment and specialized curation rather than mechanical compilation. The human remains in full control of assignment design and source selection.
Augmentation potential	claude-sonnet-5	5/5	AI significantly speeds up literature discovery and organization, letting instructors quickly draft and refine bibliographies while retaining final curatorial judgment.
Task automatability	claude-haiku-4-5-20251001	4/5	AI can generate comprehensive, well-organized bibliographies from source material with high accuracy, automatically filtering by topic, publication date, and relevance. A human may need only to verify completeness and add specialized niche sources, yielding >50% time savings on this primarily information-retrieval and formatting task.
Task automatability	claude-sonnet-5	4/5	AI systems with literature search and citation tools can compile topical bibliographies quickly, though a human should still vet relevance and quality for a specific course. This meets the time-saving bar for most of the work.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal or regulatory requirement mandates human compilation; institutional inertia and faculty preference for human curation are the main friction points. Integration into course systems is straightforward, and liability risk is low since instructors retain full discretion over final assignments.
Adoption barriers	claude-sonnet-5	1/5	No licensing, regulatory, or human-contact requirement restricts using AI to help compile reading lists; it's a low-stakes administrative/academic task.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Inference cost for bibliography generation is negligible (pennies per task), while human faculty time spent on manual literature hunting and formatting would cost $30–100+ per hour of work, making AI orders of magnitude cheaper.
Cost vs. human wage	claude-sonnet-5	5/5	Generating a bibliography draft via AI costs a fraction of a cent to a few dollars in compute versus the faculty time otherwise spent searching and compiling references.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed products (ChatGPT, Claude, specialized research APIs like Semantic Scholar integration) reliably compile and format bibliographies at scale with minimal errors. Some edge cases around rare or very recent materials remain, but production-quality outputs are routine in academic and library settings.
Technical feasibility today	claude-sonnet-5	3/5	Tools like reference managers, AI search assistants, and citation databases exist and are used, but hallucinated or irrelevant citations remain a known problem requiring verification.

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

CI 47–87 · exposure 66 · augmentation 75 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Procurement automation and e-purchasing systems are widespread in higher education and corporate sectors, with many institutions actively adopting AI-assisted vendor selection and ordering tools.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administrative processes adopt AI slowly; procurement/purchasing workflows in academia remain largely manual and paperwork-driven.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI systems can significantly assist instructors by recommending suppliers, flagging cost-effective alternatives, tracking inventory, and automating routine reordering, allowing humans to focus on pedagogy rather than logistics.
Augmentation potential	claude-sonnet-5	4/5	AI can efficiently help identify suitable textbooks, compare specifications, and summarize equipment options, significantly speeding up the research phase of this task.
Task automatability	claude-haiku-4-5-20251001	5/5	Selecting and obtaining materials and supplies is a transactional task involving catalog browsing, price comparison, inventory checking, and purchase order completion—all readily automatable by current AI systems and procurement agents that can integrate with supplier APIs and institutional purchasing systems.
Task automatability	claude-sonnet-5	3/5	AI can research, compare, and recommend textbooks/supplies and even draft purchase orders, but final selection requires judgment about course fit and vendor/budget coordination that still needs human decision-making.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While some institutions require human review of purchases above certain thresholds or have approval workflows, these barriers are relatively weak and increasingly automated; no legal licensing requirement exists for placing supply orders.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but institutional purchasing policies, budget approvals, and departmental preferences create moderate procedural friction.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI-driven procurement and vendor management systems cost a fraction of a human's time to identify suppliers, compare prices, and place orders, offering order-of-magnitude savings once integrated into institutional systems.
Cost vs. human wage	claude-sonnet-5	3/5	AI-assisted research and comparison shopping is cheap, but human time is still needed for approvals, vendor relationships, and physical equipment logistics, keeping overall costs comparable to doing it manually.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Procurement automation via e-procurement platforms and AI purchasing assistants is mature and deployed at scale in many universities and institutions; however, some institutions still use legacy systems or require manual approval steps that limit end-to-end automation.
Technical feasibility today	claude-sonnet-5	2/5	No deployed product autonomously handles academic procurement end-to-end; existing tools (search, recommendation engines, procurement software) only assist parts of the process.

Write letters of recommendation for students.

CI 54–80 · exposure 70 · augmentation 100 · 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	3/5	Academic institutions are actively piloting AI for administrative writing (including letters), but policies remain in flux and human-authored letters remain the norm; adoption is in the middle phase (pilots and early production in some departments) rather than widespread displacement.
Sector adoption velocity	claude-sonnet-5	3/5	Higher education adoption of AI writing tools is growing quickly among individual faculty, but institutional policies and norms around letters specifically remain cautious and inconsistent.
Augmentation potential	claude-haiku-4-5-20251001	5/5	AI substantially augments faculty productivity by drafting personalized letters that the professor reviews, edits, and signs, reducing composition time while preserving human judgment on student merit and content—a clear human-in-the-loop productivity multiplier.
Augmentation potential	claude-sonnet-5	5/5	AI is highly effective for augmenting this task—turning bullet points or rough notes into polished, well-structured letters that the professor reviews and personalizes.
Task automatability	claude-haiku-4-5-20251001	5/5	AI can generate well-structured, substantive letters of recommendation by synthesizing student performance data, coursework, and achievements into compelling narratives that meet the ≥50% time-saving threshold; the task is primarily text generation with clear inputs (student records, context) and outputs (formal letter).
Task automatability	claude-sonnet-5	3/5	AI can draft a competent recommendation letter given input about a student's performance, but the task requires genuine, specific knowledge of the student that only the professor has, limiting full automation without substantial human input.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Letters require faculty authentication and professional judgment about student merit; many universities and professional bodies are now establishing guidelines or restrictions on AI-generated letters, creating moderate friction around whether a human must personally author or merely review and sign the output.
Adoption barriers	claude-sonnet-5	3/5	There's no legal requirement for a human to write the letter, but strong norms of authenticity, personal knowledge, and academic integrity create social/institutional friction against pure AI generation.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	A few dollars in API costs per letter versus the faculty time (30–60 minutes per letter at loaded wages of $50–100/hour) represents an order-of-magnitude cost advantage for AI-generated output.
Cost vs. human wage	claude-sonnet-5	4/5	Drafting time is significantly reduced using AI compared to a human writing from scratch, since inference costs are trivial relative to a professor's time even with review overhead.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Multiple deployed systems (ChatGPT, Claude, specialized academic tools) reliably generate recommendation letters in production; however, some institutions restrict AI use for these letters, and quality varies with input detail, so adoption is not yet completely frictionless.
Technical feasibility today	claude-sonnet-5	3/5	LLM tools are widely used by faculty to draft recommendation letters based on notes or CVs, but professors must supply substantive details and edit for accuracy, so it's not a hands-off production process.

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

CI 51–61 · exposure 55 · 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	4/5	Postsecondary education has high digitization and rapid adoption of LMS-integrated auto-grading and AI-assisted assessment design, especially for large enrollment courses. Many universities are piloting or deploying AI exam generation and objective grading in production; adoption is faster in STEM fields.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education, especially humanities/social science departments, has been slower and more cautious in adopting AI grading tools compared to tech-forward sectors, with many pilots but limited scaled deployment.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI significantly augments faculty productivity by drafting diverse question banks, automatically scoring objective items, flagging potential plagiarism, and suggesting rubric frameworks—freeing instructors to focus on subjective evaluation and pedagogical feedback. This maintains the instructor in the loop while substantially reducing administrative burden.
Augmentation potential	claude-sonnet-5	4/5	AI substantially helps instructors draft question banks, generate rubrics, and provide first-pass feedback on essays, meaningfully speeding up the overall exam workflow while the instructor retains final grading authority.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can generate exam questions, create answer keys, and automatically grade objective tests (multiple choice, short answer matching) reliably today, yielding ~50% time savings on grading routine assessments. However, holistic evaluation of subjective responses (essays, research critiques) still requires significant human judgment, limiting full automation to perhaps half the task.
Task automatability	claude-sonnet-5	3/5	AI can draft exam questions and grade objective/short-answer items reliably, but grading nuanced archaeology/anthropology essay responses and ensuring alignment with course-specific learning objectives still requires substantial instructor input, so only partial time savings are realized end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	3/5	No legal mandate requires human grading, but institutional policies, accreditation standards, and faculty preference for pedagogical control create meaningful friction. Many institutions still require faculty sign-off on final grades and may resist algorithmic grading of subjective work due to fairness and liability concerns.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement blocks AI use, but academic integrity policies, grading appeals processes, and institutional norms around instructor accountability for grades create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI inference for generating questions, administering online exams, and auto-grading objective items costs pennies per exam, far below the loaded cost of faculty time (~$50–80/hour for postsecondary instructors). Even accounting for LMS infrastructure and oversight, the ratio favors AI by an order of magnitude on routine grading.
Cost vs. human wage	claude-sonnet-5	4/5	For generating and grading multiple-choice or short-answer exams, AI inference costs are far below faculty/TA time costs, though oversight for essay grading narrows the gap somewhat.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Learning management systems (Canvas, Blackboard) integrate automated grading for objective formats at scale in production; Large Language Models can draft questions and rubrics with reasonable quality. The main limitation is subjective grading where human oversight remains necessary, but deployed tools handle the administrative and objective components reliably.
Technical feasibility today	claude-sonnet-5	3/5	Products like AI-assisted quiz generators and essay-grading tools (e.g., Gradescope AI features, LLM-based graders) are deployed in some institutions, but reliability for discipline-specific, interpretive humanities content is inconsistent and adoption is narrow.

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

CI 51–59 · exposure 50 · 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	3/5	Higher education is moderately adopting AI-assisted grading through pilots and LMS integrations, but full replacement remains cautious. Adoption is faster in large institutions and online programs, slower in traditional humanities-heavy departments.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education is a slower-adopting sector for grading automation, with pilots more common in STEM/writing-heavy intro courses than in specialized humanities/social science upper-level courses.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI excels at augmenting instructor grading: flagging assignments for human review, generating detailed feedback drafts, scoring objective components, and surfacing outliers. This substantially raises instructor productivity while preserving human judgment on subjective evaluation.
Augmentation potential	claude-sonnet-5	4/5	AI can efficiently draft rubric-based feedback, flag issues, and summarize common errors across a class, meaningfully speeding grading while the instructor still finalizes grades.
Task automatability	claude-haiku-4-5-20251001	3/5	AI can automate significant portions of grading (syntax, factual accuracy, basic structure) and provide rubric-based scoring, but requires human oversight for nuanced evaluation of arguments, originality, and discipline-specific standards. This likely achieves 50% time savings on routine grading with current systems.
Task automatability	claude-sonnet-5	3/5	AI can draft grades and feedback for essays and assignments with substantial time savings, but nuanced grading of disciplinary reasoning in anthropology/archaeology papers still requires human judgment for full reliability at equal quality.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Academic institutions have governance structures, faculty resistance, and concerns about accountability and fairness in AI grading that slow adoption. However, no hard legal requirement mandates human grading, and many institutions are experimenting without regulatory blockers.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement blocks AI-assisted grading, though academic integrity norms and institutional policy often require instructor-of-record final sign-off on grades.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	AI grading inference costs are minimal (cents per assignment), while instructor labor for detailed grading represents substantial loaded cost. All-in integration and oversight is low relative to faculty wages, making AI economically favorable.
Cost vs. human wage	claude-sonnet-5	4/5	AI-assisted grading tools are far cheaper per assignment than faculty time, though human review/oversight for accuracy adds some cost back.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Multiple products (learning management systems with AI assistance, GPT-powered grading tools) now grade student work in production, but error rates remain material—particularly on subjective assessment and contextual understanding—and deployment is uneven across institutions.
Technical feasibility today	claude-sonnet-5	3/5	AI writing/grading assistants (e.g., Gradescope, GPT-based tools) are used in production for some grading tasks, but they have material error rates on nuanced, discipline-specific written work like archaeology essays.

Review manuscripts for publication in books and professional journals.

CI 29–37 · exposure 33 · augmentation 63 · 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	Academic publishing is slow to adopt automation in core review decisions; most institutions still require human experts, and resistance to non-human editorial authority remains high across anthropology and archeology journals.
Sector adoption velocity	claude-sonnet-5	2/5	Academic publishing is cautious and slow to adopt AI in peer review due to confidentiality, ethics policies, and trust concerns, despite AI adoption growing in adjacent research tasks.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist reviewers by flagging methodological inconsistencies, extracting key claims, suggesting references, or checking formatting, reducing reviewer workload while keeping final judgment with the human expert.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully assist reviewers by summarizing manuscripts, checking references, flagging inconsistencies, and improving efficiency, while the human retains ultimate judgment and sign-off.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can assist with structural review, grammar checking, and reference formatting, but cannot reliably assess novelty, scholarly contribution, or disciplinary validity—the core gatekeeping functions of peer review. Meaningful automation would require human experts to validate the final recommendations.
Task automatability	claude-sonnet-5	3/5	AI can draft summaries, check citations, flag methodological issues, and assess clarity, but nuanced disciplinary judgment about theoretical contribution and originality still requires human expertise, limiting full end-to-end automation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Peer review legitimacy depends on expert human judgment and accountability; publishers and disciplines have strong norms against algorithmic gatekeeping, and editorial liability falls on named human reviewers, creating legal and professional barriers to full automation.
Adoption barriers	claude-sonnet-5	4/5	Peer review is an academic institutional norm requiring domain expert credibility and accountability; journals require named, credentialed reviewers, creating strong professional and reputational barriers to AI substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	AI-assisted tools (grammar, similarity checking) cost pennies per manuscript, but supplementary review—not replacement—is the realistic use case, so the economic advantage over expert human reviewers is modest.
Cost vs. human wage	claude-sonnet-5	3/5	AI tools can cheaply handle preliminary checks, but the substantive scholarly review still requires paid expert time, so overall cost savings versus a human reviewer are moderate, not dramatic.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While automated plagiarism detection and language tools exist in production, no deployed system can independently conduct scholarly peer review with credible quality or editorial trust. Academic publishers still rely on human reviewers for substantive decisions.
Technical feasibility today	claude-sonnet-5	2/5	Some AI writing-assistant and reviewer-support tools exist (e.g., grammar/plagiarism checkers, summarization tools) but no deployed product reliably performs substantive peer review of anthropology/archaeology manuscripts in production.

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

CI 25–37 · exposure 25 · augmentation 75 · 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	2/5	Postsecondary institutions, especially in humanities-heavy fields like anthropology, have been slow to adopt AI for core academic functions. Adoption remains in pilot/experimental phases rather than production deployment at scale.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools unevenly and cautiously, especially in humanities/social science departments, with pilots more common than systematic curriculum automation.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can meaningfully assist faculty by generating draft syllabi, suggesting learning outcomes, recommending readings, and helping revise course materials—tasks that benefit from AI support while faculty retain full pedagogical authority and judgment. This augmentation substantially raises productivity on content generation and iteration.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully assist by generating draft materials, suggesting readings, summarizing new research, and helping revise course content, significantly speeding up parts of the planning process.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can assist with generating course outlines, summarizing content, and drafting learning objectives, but curriculum planning requires subject-matter judgment, alignment with institutional standards, and understanding of pedagogical fit that remains largely human-dependent. End-to-end automation with 50% time savings at equal quality is not achievable with current systems.
Task automatability	claude-sonnet-5	2/5	AI can draft syllabi and suggest readings, but true curriculum planning requires disciplinary judgment, alignment with accreditation, and institutional context that current systems cannot autonomously handle end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Faculty governance structures, accreditation bodies, and institutional curricula committees typically require faculty review and approval of curriculum changes; many institutions have formal policies requiring human faculty sign-off on pedagogy. These governance and liability barriers are substantial.
Adoption barriers	claude-sonnet-5	3/5	Curriculum design is generally faculty-governed and subject to departmental/accreditation approval, creating organizational friction, though it isn't a formally licensed task requiring legal sign-off.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Inference cost for AI-assisted curriculum planning is very low (dollars per prompt), while a faculty member's time on curriculum work costs tens to hundreds of dollars per hour. The cost ratio strongly favors AI, though integration and oversight overhead applies.
Cost vs. human wage	claude-sonnet-5	2/5	While AI drafting assistance is cheap, the human faculty time for expert review, revision, and institutional approval remains substantial, keeping the overall cost comparable to human-led effort.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools like ChatGPT and Claude can draft syllabus sections and suggest readings, but no deployed product reliably handles the full cycle of curriculum evaluation and revision as practiced in postsecondary institutions. Production use is limited to narrow components, not the full task.
Technical feasibility today	claude-sonnet-5	2/5	Some LLM-based tools assist with lesson planning and syllabus drafting, but no deployed product reliably designs and revises full postsecondary anthropology/archaeology curricula without heavy faculty oversight.

Prepare and deliver lectures to undergraduate or graduate students on topics such as research methods, urban anthropology, and language and culture.

CI 23–34 · exposure 25 · augmentation 75 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Postsecondary education adoption of AI for core instructional delivery is still in early pilots. Most institutions use AI for administrative tasks and content drafting, not replacement of faculty lecturers. Cultural and regulatory momentum remains slow.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools slowly for core teaching functions, with pilots for content generation but little movement toward AI-led lecture delivery.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can powerfully augment lecturers by generating slide decks, drafting explanations of complex concepts, drafting exam questions, and suggesting contemporary examples—all while the instructor retains editorial control and live delivery. This is already being adopted in some institutions.
Augmentation potential	claude-sonnet-5	4/5	AI substantially helps professors prepare lecture materials, generate examples, summarize readings, and create assessments, meaningfully boosting prep efficiency while the instructor still delivers content.
Task automatability	claude-haiku-4-5-20251001	2/5	Generating lecture content and slides is now feasible at scale with LLMs, but delivering lectures—especially engaging, interactive ones responsive to student questions and non-verbal cues—requires human presence. Current AI can assist with preparation (~40% time savings) but cannot fully replace the live delivery component that defines the task.
Task automatability	claude-sonnet-5	2/5	AI can help draft lecture content and slides, but live delivery, in-class engagement, Q&A, and adaptive teaching require human presence and judgment that current systems cannot fully replace end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers include tenure, union protections, accreditation requirements (many institutions expect faculty with research credentials to teach), and student/parent expectations of human instruction. Educational institutions also face reputational risk and enrollment pressure if teaching is fully automated.
Adoption barriers	claude-sonnet-5	3/5	No formal licensing requires a human lecturer, but strong institutional norms, accreditation expectations, and student/faculty preference for human instruction create real friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Lecture preparation AI (ChatGPT, Claude) is cheap per output, but the marginal cost of eliminating a tenured faculty member is offset by employment contracts, institutional commitment, and the salaries are already paid. For new lecture creation alone, AI might be 5–10x cheaper, but replacement does not follow a simple per-task cost model.
Cost vs. human wage	claude-sonnet-5	3/5	AI can cheaply assist with content prep, but the delivery portion still requires paid faculty time, so overall cost savings are moderate rather than transformative.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI systems can draft lecture notes and outlines reliably, but no deployed product reliably delivers full lectures or manages the real-time pedagogical feedback loop. Lecture capture and asynchronous content generation exist, but synchronous live instruction with student interaction remains a human function in practice.
Technical feasibility today	claude-sonnet-5	2/5	Tools like ChatGPT can generate lecture outlines and even narrated content, but no deployed product reliably delivers full interactive university lectures in place of a professor at scale.

Advise students on academic and vocational curricula, career issues, and laboratory and field research.

CI 25–30 · exposure 25 · augmentation 63 · importance 4.7/5 · 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	Higher education adoption of AI advising tools is still in pilot phase at most institutions; most advising remains human-performed. While some administrative and scheduling automation has been deployed, comprehensive AI-driven student advising is not yet standard practice, indicating slow actual adoption.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools slowly for advising due to institutional inertia, though some universities pilot AI chatbots for basic student services.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist faculty advisors by surfacing curriculum options, flagging student progress data, and retrieving research literature or field-site information, thereby freeing time for deeper mentorship conversations. However, the augmentation is incremental rather than transformative because the core value—personalized guidance and research mentorship—remains human-centered.
Augmentation potential	claude-sonnet-5	4/5	AI can help faculty draft advising materials, summarize career paths, suggest curricula, and prepare guidance for field/lab research, meaningfully boosting efficiency while the advisor remains central.
Task automatability	claude-haiku-4-5-20251001	2/5	Advising students on academic curricula and career guidance can be partially automated (information retrieval, suggesting program options), but the task fundamentally requires understanding individual student circumstances, aspirations, and complex interpersonal judgment. The research guidance aspect demands subject-matter expertise in methodology and field-specific knowledge that current AI can support but not fully replace end-to-end.
Task automatability	claude-sonnet-5	2/5	AI can provide generic information but genuine academic and career advising requires personalized knowledge of a student's history, institutional context, and relationship-based trust that current systems cannot replicate end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Postsecondary institutions have strong norms, accreditation expectations, and often formal requirements that faculty provide academic advising and mentorship. Students expect and often require direct human interaction with advisors, and institutions face liability for poor advising outcomes. These institutional and regulatory barriers substantially protect this task from full automation.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement exists, but institutional norms, accreditation expectations, and student preference for a mentor with disciplinary expertise create moderate friction against full AI substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI can reduce overhead on information-delivery aspects of advising, but the personalized and expert judgment components (evaluating student fit for subfields, discussing research ethics and design) require human faculty time. The cost savings are modest because human oversight and final decision-making remain essential.
Cost vs. human wage	claude-sonnet-5	2/5	While AI query costs are low, the need for extensive human oversight, verification, and relationship-building to make advice trustworthy keeps effective costs comparable to or only modestly below a professor's time.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	While chatbots can answer routine advising questions and provide curriculum information, no deployed product reliably handles the full scope of this task—personalised career counseling, nuanced discussion of research methodologies, and adaptive guidance based on student progress and constraints. Deployed advising systems are narrow (schedule checking) rather than comprehensive.
Technical feasibility today	claude-sonnet-5	2/5	Chatbots and advising tools exist for basic FAQ-style guidance, but no deployed product reliably handles nuanced academic/career mentorship or field research planning at scale in real institutions.

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

CI 16–39 · exposure 17 · augmentation 63 · 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	Academic sectors adopt digital tools slowly and tend to resist outsourcing professional judgment on what is important in a field; most faculty still rely on manual reading, conference attendance, and peer conversation rather than AI-driven curation.
Sector adoption velocity	claude-sonnet-5	2/5	Academia adopts AI tools unevenly; literature review aids are gaining traction but the broader professional engagement practice remains largely traditional.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI literature summaries and preprint screening could usefully help a faculty member filter and process large document volumes, but the human must remain the decision-maker on relevance and significance, making this an assistive rather than transformative augmentation.
Augmentation potential	claude-sonnet-5	4/5	AI significantly aids in literature discovery, summarization, and trend tracking, meaningfully boosting a scholar's efficiency while they remain the one engaging with colleagues and conferences.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires sustained professional judgment about what is relevant, synthesis across diverse sources, and meaningful peer engagement. While AI can summarize literature, it cannot replicate the selective curation, critical evaluation, and relationship-building that characterize staying abreast in a specialized field.
Task automatability	claude-sonnet-5	2/5	AI can summarize literature and surface relevant papers, but the task inherently requires ongoing personal engagement, judgment, and networking that cannot be fully offloaded end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Professional development and maintaining expertise are core expectations of academic roles; there is strong institutional and cultural expectation that faculty directly engage with their field rather than delegate this to AI systems, creating significant adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing barrier exists, but the professional and social nature of the task (conferences, colleague interaction) creates organizational and cultural friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of AI-assisted reading tools is modest, but they must be integrated into workflows and do not eliminate the human time spent evaluating, conversing, and networking—making overall cost-per-outcome only marginally better than human-only approaches.
Cost vs. human wage	claude-sonnet-5	3/5	AI literature summarization tools are cheap relative to time spent, but human conference attendance and colleague discussion carry costs that AI cannot replace, making overall savings moderate.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI can assist with literature summarization and conference abstract processing, but no deployed system reliably identifies what is professionally important or substitutes for the human judgment needed to evaluate developments and maintain collegial networks.
Technical feasibility today	claude-sonnet-5	2/5	Tools like AI-powered literature summarizers and research digest apps exist but are not reliably used as full substitutes for staying current in a specialized academic field.

Participate in student recruitment, registration, and placement activities.

CI 25–30 · exposure 25 · augmentation 50 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While some higher-education institutions are piloting chatbots and CRM automation, the sector has been slow to adopt AI for core student-facing recruitment and placement roles; most rely on human advisors and traditional processes.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administrative functions like admissions marketing are adopting AI tools moderately, but the specific faculty-driven recruitment/placement activities lag due to relationship-dependent nature.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI tools can augment recruitment coordinators by automating email campaigns, scheduling interviews, and flagging registration anomalies, raising efficiency on administrative parts of the task while humans retain decision-making on placements and advising.
Augmentation potential	claude-sonnet-5	3/5	AI can help draft recruitment materials, manage student data, and provide administrative support for registration and placement tracking, meaningfully aiding but not transforming the task.
Task automatability	claude-haiku-4-5-20251001	2/5	Student recruitment, registration, and placement require significant human judgment and relationship-building. AI can assist with email outreach, scheduling, and basic forms processing, but cannot authentically recruit students or handle placement decisions that depend on nuanced career counseling and personalized guidance.
Task automatability	claude-sonnet-5	2/5	This task involves interpersonal engagement, event participation, and institutional coordination that AI can support but not perform end-to-end reliably today.atable partly through communications drafting but not full participation.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Higher education enrollment and placement carry regulatory and reputational risks; institutions must legally retain human advisors in key placement and student-success decisions. FERPA and accreditation standards also impose human accountability requirements.
Adoption barriers	claude-sonnet-5	3/5	No formal licensing barrier, but institutional norms, personal relationships in placement/recruitment, and faculty governance create moderate friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration and oversight of AI systems in student recruitment and placement would require significant configuration and human review, making the total cost comparable to or exceeding the cost of a dedicated academic advisor or student services staff member doing this work.
Cost vs. human wage	claude-sonnet-5	2/5	Faculty time for these activities is a small fraction of their role, and AI tools (CRM, chatbots) add cost/integration without replacing the human presence required for recruitment events and placement mentoring.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some CRM and scheduling tools exist, but no deployed product reliably handles the full pipeline of recruitment, registration, and placement in an academic context at equal quality. Current systems lack the judgment required for personalized advising and placement matching.
Technical feasibility today	claude-sonnet-5	2/5	Some products assist with recruitment marketing, chatbots for admissions inquiries, and registration systems, but faculty participation in recruitment events and placement counseling is not replaced by deployed AI products.

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

CI 23–30 · exposure 25 · augmentation 63 · 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	1/5	Higher education and academic publishing are slow adopters of AI-driven automation, with strong cultural and institutional norms favoring human scholarship. Adoption of AI as a research tool remains limited to ancillary tasks like writing assistance.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education and humanities/social science research are relatively slow adopters of AI for core research tasks compared to fields like finance or software, though writing-assistance tools are spreading.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist by helping researchers search literature, organize citations, draft sections, and visualize data analysis, but the core research tasks—fieldwork, interpretation, novel contribution—remain fundamentally human.
Augmentation potential	claude-sonnet-5	4/5	AI substantially aids literature reviews, data analysis, drafting manuscripts, and formatting for journals, meaningfully boosting researcher productivity while the human retains intellectual ownership.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can assist with literature review, data analysis, and draft writing, conducting original research requires field expertise, novel intellectual contributions, and judgment that current AI cannot reliably perform end-to-end. AI cannot independently design fieldwork, interpret anthropological/archaeological evidence, or produce genuinely novel findings.
Task automatability	claude-sonnet-5	2/5	AI can assist with literature review, drafting, and data analysis, but original archaeological/anthropological fieldwork, novel data collection, and scholarly interpretation require human expertise and cannot be fully automated end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers protect this task: academic institutions require human researchers for tenure and credentials, field work demands legal/ethical authorization and physical presence, peer review assumes human authorship, and liability for false or plagiarized findings rests with the researcher.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement per se, but academic norms, peer review, authorship credit, and institutional expectations of human intellectual contribution create meaningful friction against full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The loaded cost of a postsecondary anthropology/archaeology researcher (salary + benefits + institutional overhead) remains far lower than the total cost of AI systems that would need to replicate field research, data collection, analysis, and novel insight generation.
Cost vs. human wage	claude-sonnet-5	2/5	AI tools are cheap for drafting and summarization, but the core research (fieldwork, data collection, expert analysis) still requires costly human labor, so overall cost savings are limited.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	AI tools can draft papers and summarize literature, but no deployed product reliably conducts original research or produces publication-ready scholarly work that passes peer review in these fields. These disciplines demand situated knowledge and interpretive depth beyond current AI capabilities.
Technical feasibility today	claude-sonnet-5	2/5	Products like AI writing assistants and literature-review tools exist and are used by academics, but no deployed system reliably conducts original research and produces publishable, peer-reviewed findings independently.

Write grant proposals to procure external research funding and review others' grant proposals.

CI 25–25 · exposure 25 · 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	2/5	Academic and research institutions are adopting AI writing assistants cautiously, mainly for drafting support and editing rather than independent proposal generation. Concerns about authenticity, funder policies, and institutional risk management have kept adoption slow and limited to augmentation roles in this specialized domain.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education and humanities/social science research are slower adopters of AI tools compared to industries like finance or tech, with usage mostly limited to individual experimentation rather than institutional workflows.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI writing assistants meaningfully improve productivity by helping researchers draft sections, refine language, check for clarity, and organize ideas—transforming the editing and iteration cycle. A human researcher using AI to draft and revise proposal text can work faster while maintaining full control over content, novelty claims, and strategic direction.
Augmentation potential	claude-sonnet-5	4/5	AI is genuinely useful for brainstorming, drafting sections, editing prose, summarizing literature, and checking proposals against guidelines, meaningfully speeding up the writing and reviewing process while the scholar retains final judgment.
Task automatability	claude-haiku-4-5-20251001	2/5	Grant writing requires deep domain expertise, novel argumentation, and persuasive narrative tailored to specific funding agencies and review committees. While AI can draft sections or improve clarity, the core task of conceptualizing research, justifying novelty, and crafting compelling arguments for scarce funding fundamentally depends on human judgment and specialized knowledge that current systems cannot reliably replace end-to-end.
Task automatability	claude-sonnet-5	2/5	AI can draft boilerplate sections and improve language, but crafting a compelling, discipline-specific research narrative with novel contributions and accurate budget justification requires deep domain expertise and original judgment that current AI cannot fully replicate end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Funding agencies and institutional review committees typically require that grant proposals and reviews be authored or certified by qualified humans with proper credentials and accountability. Many funders explicitly prohibit or restrict AI-generated content; liability for false claims in proposals and the need for expert judgment create strong organizational and regulatory barriers to full automation.
Adoption barriers	claude-sonnet-5	4/5	Grant proposals require named PI credentials, institutional certification, and reviewers must be qualified experts per funder rules (e.g., NSF/NEH panels), creating strong authorization and integrity barriers to AI substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Grant writing demands substantial human expertise (senior researchers, grant administrators, subject-matter specialists) whose loaded costs are high. AI assistance reduces iteration and editing time modestly, but does not yet approach an order-of-magnitude cost advantage when accounting for the need for expert human review and revision of AI output.
Cost vs. human wage	claude-sonnet-5	2/5	AI drafting tools are cheap per use, but the human oversight, subject-matter expertise, and iteration needed to produce a competitive proposal or credible review keeps overall cost comparable to faculty time rather than dramatically cheaper.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed AI product reliably writes competitive grant proposals from scratch or conducts rigorous peer review of proposals independently. Current tools (LLMs, writing assistants) can assist with drafting and editing, but producing fundable proposals or authoritative reviews at scale remains beyond production-grade capability in real organizations.
Technical feasibility today	claude-sonnet-5	2/5	Writing assistants and LLMs are used informally to help draft proposal sections, but no deployed product reliably produces fundable grant proposals or performs peer review at production scale in academia.

Initiate, facilitate, and moderate classroom discussions.

CI 16–30 · exposure 20 · augmentation 50 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Academic institutions are slow adopters of instructor replacement; higher education remains staff-intensive and resistant to automation of core teaching functions. Discussion facilitation is seen as a core instructor competency, not a candidate for external automation.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education is a moderate-adoption sector for AI tools (e.g., grading, content generation) but live classroom facilitation by AI remains largely unexplored in production settings.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by generating discussion prompts, summarizing key points in real-time, or flagging participation patterns, moderately boosting instructor productivity. However, the augmentation is supplemental rather than transformative, as the instructor remains the primary discussion driver.
Augmentation potential	claude-sonnet-5	3/5	AI can help instructors generate discussion questions, summarize readings, or provide talking points, offering moderate productivity support before or after class sessions.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI can generate discussion prompts and summarize points, facilitating and moderating live classroom discussions—requiring real-time responsiveness to student nuance, dynamic group management, and instructor judgment—falls well short of the 50% time-saving threshold. The core human role (presence, adaptability, authority) cannot be meaningfully replaced today.
Task automatability	claude-sonnet-5	2/5	Live discussion facilitation requires real-time social reading, spontaneous follow-up, and classroom management that current AI cannot reliably replicate end-to-end; at best AI can help prep discussion prompts.6.rating
Adoption barriers	claude-haiku-4-5-20251001	4/5	Institutional pedagogy, accreditation standards, and student expectations strongly favor human instructor presence and authority in classroom discussions. Legal and professional norms treat the instructor's role as non-delegable, creating substantial organizational and regulatory friction.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement mandates a human specifically, but strong institutional and pedagogical norms favor human instructors leading discussions and there is no legal barrier akin to licensure.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The integrated cost of AI discussion tools (licensing, moderation oversight, integration) remains high relative to the marginal cost of an instructor already present in the classroom who naturally facilitates discussion as part of their role.
Cost vs. human wage	claude-sonnet-5	1/5	Since no viable AI substitute exists for live facilitation, the comparison to human faculty cost is not favorable; any attempt would require extensive human oversight negating savings.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed product reliably moderates live academic discussions end-to-end. Chatbots can simulate discussion participation, but orchestrating an actual classroom discussion with real students requires contextual awareness, conflict resolution, and pedagogical timing that current AI does not handle in production.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously leads live in-person postsecondary seminar discussions; existing chatbots handle asynchronous text Q&A, not dynamic group facilitation.

Maintain regularly scheduled office hours to advise and assist students.

CI 9–20 · exposure 8 · 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	1/5	Postsecondary education, especially in humanities and social sciences, has remained resistant to automating core faculty-student interactions. While chatbots may supplement information dissemination, actual adoption of AI for office hour replacement remains negligible in practice.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education has been slow to adopt AI for personalized advising roles; some institutions pilot chatbots for basic FAQs but faculty office hours remain untouched.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI tools could assist faculty by preparing summaries of student questions, suggesting relevant resources, or drafting initial responses that the faculty member refines. Such augmentation could modestly boost office hour efficiency, though the human faculty member would remain the primary advisor.
Augmentation potential	claude-sonnet-5	3/5	AI can help professors prepare materials, answer routine student questions beforehand, or triage inquiries, moderately increasing efficiency without replacing the interactive advising itself.
Task automatability	claude-haiku-4-5-20251001	2/5	Office hours require real-time, context-sensitive conversation with individual students about their specific academic and personal circumstances. While AI could handle some routine scheduling or FAQ-style questions, the core advisory function—understanding student concerns, offering tailored guidance, and building trust—requires human presence and judgment that current AI cannot reliably replicate at acceptable quality.
Task automatability	claude-sonnet-5	1/5	This task requires physical/virtual presence and real-time personal interaction and mentorship with students, which is inherently a human relational function that AI cannot substitute for in a meaningful end-to-end way.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Institutional and professional norms strongly expect faculty to hold office hours and maintain direct relationships with students. Universities have accreditation and pedagogical standards emphasizing faculty-student contact. Student expectations and potential enrollment/satisfaction impacts create significant friction against automated substitution.
Adoption barriers	claude-sonnet-5	4/5	Institutional norms, accreditation expectations, and student-support policies effectively require faculty availability for advising; there's strong organizational and reputational pressure for a human presence, though not a strict legal license requirement.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	A chatbot or AI assistant might reduce response latency for simple questions at low cost, but the full replacement would require sophisticated conversational AI, integration with university systems, and ongoing monitoring. The total cost per student interaction would likely approach or exceed the marginal cost of a faculty member's time, especially when liability and student satisfaction are factored in.
Cost vs. human wage	claude-sonnet-5	2/5	While AI chat tools are cheap, they cannot substitute for the actual advising function, so the relevant cost comparison for genuine task completion still favors the human, though AI can reduce some ancillary support costs.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably substitutes for a faculty member's office hours. Chatbots exist for limited support, but universities have not adopted AI systems to conduct actual faculty advising at scale or with measurable equivalence to human office hours.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product replaces a professor's office hours; chatbots can supplement FAQ-type queries but cannot hold accredited advising sessions or provide personalized academic mentorship.

Provide professional consulting services to government or industry.

CI 4–20 · exposure 8 · augmentation 63 · 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	1/5	Adoption of AI for autonomous consulting delivery is minimal. While some firms use AI tools to augment consultant work, autonomous consulting agents are not in production. The sector values human credibility and liability assignment.
Sector adoption velocity	claude-sonnet-5	2/5	Academic consulting in niche social science fields adopts AI slowly; this is a small, specialized market with limited AI-driven tool integration reported so far.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist consultants by drafting reports, summarizing literature, analyzing datasets, and brainstorming frameworks, meaningfully raising productivity on research and synthesis phases while the human consultant retains client relationship and judgment responsibility.
Augmentation potential	claude-sonnet-5	4/5	AI can meaningfully assist with literature reviews, report drafting, data organization, and preliminary analysis, boosting consultant productivity while the expert retains final judgment.
Task automatability	claude-haiku-4-5-20251001	1/5	Providing professional consulting requires domain expertise, client relationship management, nuanced judgment, and contextual advice tailored to specific organizational problems. Current AI systems lack the credibility, accountability, and deep domain mastery required to deliver consulting services independently.
Task automatability	claude-sonnet-5	2/5	Consulting on anthropology/archeology matters requires site-specific judgment, contextual expertise, stakeholder negotiation, and professional liability that AI cannot autonomously replicate end-to-end today.'
Adoption barriers	claude-haiku-4-5-20251001	5/5	Consulting services face hard barriers: professional liability, client expectations for human expertise and accountability, potential regulatory requirements for anthropological/archeological expertise (especially on cultural heritage matters), and the legal requirement that a licensed/credentialed professional typically must sign off on recommendations.
Adoption barriers	claude-sonnet-5	4/5	Government contracts and regulatory compliance (e.g., cultural resource management, environmental review) often require credentialed experts to attest to findings, creating strong professional and legal barriers to full automation.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	While AI analysis tools are cheap, integrating them into a consulting engagement with proper oversight, client interaction, and liability management approaches human cost. The full consulting service (not just analysis) remains far more expensive to deliver via AI than employing human consultants.
Cost vs. human wage	claude-sonnet-5	2/5	AI could cheaply draft background research or reports, but the core deliverable—expert judgment and signed-off recommendations—still requires paid human expert time, so overall cost savings are limited.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably performs consulting services autonomously today. Consulting requires iterative client engagement, responsibility for outcomes, and professional certification/reputation that AI cannot assume.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product provides autonomous professional archeological/anthropological consulting services to government or industry clients; this remains squarely a human expert function.

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

CI 0–16 · exposure 8 · augmentation 50 · 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	1/5	Higher education remains a laggard sector in AI adoption for core pedagogical functions. Institutions move slowly due to accreditation concerns, resistance to erosion of faculty roles, and the high stakes of student outcomes; pilot deployments are minimal and production use of AI for supervision is virtually nonexistent.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI tools slowly for core mentorship and supervisory duties, with pilots mostly limited to administrative or grading support.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can usefully assist faculty with administrative overhead—scheduling meetings, tracking student progress, generating initial feedback on written work, or flagging at-risk students—but these augmentations support rather than transform the core supervisory task, which remains fundamentally human-centered.
Augmentation potential	claude-sonnet-5	3/5	AI can help with scheduling, feedback drafting, research resource curation, and monitoring progress, but the core supervisory judgment remains human-driven.
Task automatability	claude-haiku-4-5-20251001	2/5	Supervision of student work requires ongoing judgment, mentoring, and adaptive feedback on individual progress—tasks that depend on understanding nuanced student needs, motivation, and intellectual development. While AI could assist with scheduling, progress tracking, or initial feedback drafting, the core supervisory relationship and real-time guidance cannot be meaningfully automated today.
Task automatability	claude-sonnet-5	1/5	Supervision requires ongoing mentorship, judgment calls on student development, and situational adaptation that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Educational supervision carries hard legal and institutional barriers: faculty members hold fiduciary duty and accreditation responsibility for student learning outcomes, institutional policies require human oversight, and many jurisdictions have compliance requirements around student safety and research ethics that mandate human authorization and sign-off.
Adoption barriers	claude-sonnet-5	4/5	Academic accreditation, mentorship responsibilities, and institutional policies typically require a qualified faculty member to supervise students, creating strong structural barriers.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The loaded cost of a faculty member supervising students is embedded in institutional salary structures and cannot be replaced by current AI at lower total cost while maintaining educational quality and legal/fiduciary responsibility for student outcomes.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this supervisory function, so cost comparison favors the human entirely.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs educational supervision at scale. While AI can grade essays or answer routine questions, autonomous supervision of research and internships—involving accountability for student safety, learning outcomes, and intellectual integrity—remains a research-stage problem with no production systems.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously supervises students' teaching, internships, or research; this remains a human relational and evaluative role.

Hire new faculty.

CI 0–16 · exposure 5 · augmentation 38 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Higher education, particularly tenure-track hiring, remains a traditionally conservative sector with slow AI adoption; most departments still rely on manual review processes and committees, with only early-stage pilots of AI-assisted screening in a few institutions.
Sector adoption velocity	claude-sonnet-5	1/5	Academic hiring processes are notoriously slow to change, governed by tradition, tenure committees, and accreditation standards, with minimal AI adoption in the core decision process.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can provide useful assistance by automating resume screening, flagging potentially strong candidates, scheduling interviews, and summarizing application materials, which reduces administrative burden on search committees and allows faculty to focus on substantive evaluation and interviews.
Augmentation potential	claude-sonnet-5	2/5	AI can help with administrative aspects like screening CVs, scheduling interviews, or drafting job postings, but offers limited assistance for the core judgment-based hiring decision.
Task automatability	claude-haiku-4-5-20251001	1/5	Hiring faculty involves subjective judgment about research fit, teaching philosophy, interpersonal dynamics, and institutional alignment that requires human deliberation. While AI can assist with resume screening or scheduling, the core decision-making and relationship assessment cannot be fully automated to meet the 50% time-saving threshold at equal quality.
Task automatability	claude-sonnet-5	1/5	Hiring faculty requires nuanced human judgment about research fit, collegiality, teaching ability, and departmental politics that AI cannot perform end-to-end today.directed.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Faculty hiring is heavily protected by university governance structures, shared governance committees, accreditation requirements, and norms that demand human faculty judgment in evaluating peers; legal liability for discrimination suits also creates strong institutional pressure to maintain human oversight of hiring decisions.
Adoption barriers	claude-sonnet-5	5/5	Faculty hiring involves legal compliance (EEO, tenure review), institutional governance, shared governance rules, and requires authorized human committees and administrators to make and sign off on decisions.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	AI tools for candidate screening and scheduling reduce some administrative burden but cannot replace the salary costs of hiring committees, search coordinators, and senior faculty time spent on interviews and deliberation, keeping total costs close to or above human-only alternatives.
Cost vs. human wage	claude-sonnet-5	1/5	The task is inherently a human committee decision-making process; there is no AI substitute product whose cost could be compared favorably to human labor for this specific judgment task.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	No deployed AI system reliably performs end-to-end faculty hiring; products exist only for narrow subtasks like resume parsing or initial application filtering, with significant error rates and scope limitations in evaluating academic fit and scholarly potential.
Technical feasibility today	claude-sonnet-5	1/5	No deployed products conduct faculty hiring decisions; at most AI is used for resume screening or scheduling, not the actual hiring judgment and committee deliberation.

Collaborate with colleagues to address teaching and research issues.

CI 0–7 · exposure 0 · augmentation 38 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Higher education adoption of AI for core academic governance and collaboration remains minimal; institutions are cautious about automating faculty relationships and institutional decision-making, and such automation would face strong organizational and professional resistance.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education is a moderate-to-slow adopter of AI for governance and collegial functions, with pilots mostly in administrative or research-support tools rather than collaborative decision-making itself.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI might assist by drafting meeting notes or organizing research data for discussion, but the essence of collaborative problem-solving among colleagues—trust, accountability, and negotiated judgment—cannot be substantially enhanced by current AI tools in this context.
Augmentation potential	claude-sonnet-5	3/5	AI tools can assist by summarizing research, drafting agendas, or synthesizing literature to support these discussions, but the core collaborative interaction remains human-led.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires nuanced interpersonal negotiation, synthesis of diverse disciplinary perspectives, and context-dependent judgment about academic priorities. Current AI systems cannot meaningfully replace the collaborative problem-solving and relationship-building essential to addressing institutional or research issues among human colleagues.
Task automatability	claude-sonnet-5	1/5	This is an interpersonal, relationship-driven collaborative activity involving negotiation, shared decision-making, and institutional context that AI cannot perform end-to-end.'
Adoption barriers	claude-haiku-4-5-20251001	5/5	Academic governance, tenure decisions, and research partnerships involve confidential matters, professional judgment, and institutional accountability that legally and ethically require direct human participation and sign-off. Colleagues cannot delegate their voice in such matters to an automated system.
Adoption barriers	claude-sonnet-5	4/5	Collaboration among colleagues is embedded in academic governance, tenure processes, and departmental norms requiring human participation and accountability.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Any AI system capable of meaningful collaboration would require significant human oversight and final decision-making; the all-in cost (inference, integration, and mandatory human validation) would exceed the hourly cost of the faculty member directly engaging in collaboration.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI product performing this task, so no meaningful cost comparison exists; the human cost is the only viable option.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs academic collaboration and issue resolution on behalf of faculty members. While AI can draft meeting agendas or summarize discussions, actually representing a colleague's interests and advancing consensus on complex teaching or research challenges remains beyond production-grade systems.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product substitutes for human faculty collegial collaboration on teaching and research strategy; this remains a human social process.

Participate in campus and community events.

CI 0–7 · exposure 0 · augmentation 25 · 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	There is no adoption of AI for this task because it is intrinsically human-centered; faculty participation in institutional and community life remains entirely within human purview.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education is adopting AI for administrative and instructional support, but this specific interpersonal/civic engagement task shows little to no AI displacement or adoption.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could marginally assist with event planning, scheduling, or promotion logistics, but offers minimal productivity gain for the core task of actual participation and engagement.
Augmentation potential	claude-sonnet-5	2/5	AI can help with logistics like scheduling, drafting event materials, or summarizing outcomes, but offers minimal assistance for 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, interpersonal interaction, and contextual judgment that cannot be automated. AI systems cannot physically attend events or meaningfully engage with audiences in real time.
Task automatability	claude-sonnet-5	1/5	Participating in in-person campus and community events is a physical presence and social/relational activity that AI cannot perform on behalf of a person; there is no meaningful time-saving substitution possible.
Adoption barriers	claude-haiku-4-5-20251001	5/5	This task is protected by the requirement for human physical and social presence; institutions and communities expect actual faculty members to participate in events, and no substitution is legally or organizationally feasible.
Adoption barriers	claude-sonnet-5	4/5	Community and campus engagement inherently requires human presence, institutional representation, and relationship-building, which are effectively hard barriers against automation even though not formally licensed.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	AI participation is not applicable to this task, making cost comparison meaningless. The task requires human time investment with no AI alternative.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute providing this output, so cost comparison favors the human by default (AI cannot deliver the task at all).
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No AI product can autonomously participate in campus or community events; this task is fundamentally dependent on human attendance and social engagement.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product attends or represents an individual at physical events; this remains entirely outside current product capabilities.

Perform administrative duties, such as serving as department head.

CI 0–7 · exposure 0 · augmentation 38 · importance 3.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Academic institutions have strong structural and regulatory reasons to require human department heads; adoption of AI for autonomous administrative governance in higher education remains negligible.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education administration adopts AI slowly for governance and leadership roles, though it may use AI tools for routine administrative support tasks.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI can assist with scheduling, email filtering, or report generation, but these assistive applications remain marginal to the core duties of strategic leadership and institutional decision-making that define the role.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with scheduling, drafting memos, summarizing meeting notes, and handling routine correspondence, moderately easing the administrative burden of the role.
Task automatability	claude-haiku-4-5-20251001	1/5	Department head duties require strategic decision-making, budget allocation, personnel management, and institutional representation that demand human judgment, institutional knowledge, and accountability. Current AI cannot autonomously manage these responsibilities end-to-end.
Task automatability	claude-sonnet-5	1/5	Department head duties involve interpersonal leadership, budget decisions, personnel management, conflict resolution, and institutional politics that require human judgment and accountability, not something AI can execute end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Legal, fiduciary, and contractual requirements mandate that a human with institutional authority and accountability must serve as department head; universities cannot delegate this role to an automated system.
Adoption barriers	claude-sonnet-5	4/5	Department head roles typically require institutional appointment, faculty governance approval, and formal accountability structures, creating strong organizational and quasi-regulatory barriers to automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The loaded cost of a department head (faculty salary plus benefits and institutional overhead) is far lower than the accumulated cost of AI systems, human oversight, and liability insurance needed to cover autonomous departmental administration.
Cost vs. human wage	claude-sonnet-5	1/5	Since AI cannot substitute for the leadership and accountability functions of a department head, there is no meaningful AI cost basis for comparison—the human role remains necessary.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product can reliably perform department head administrative duties independently; while AI can assist with scheduling or document drafting, the legal and fiduciary responsibilities of a department head require a human decision-maker.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product performs the role of department chair; at most AI tools assist with scheduling or drafting reports, but the administrative leadership role itself is not automated by any product.

Act as advisers to student organizations.

CI 0–7 · exposure 0 · augmentation 25 · importance 3.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Postsecondary advising of student organizations remains a low-digitization, human-contact-dependent activity with no measurable AI adoption in production; sector inertia and regulatory structure prevent rapid change.
Sector adoption velocity	claude-sonnet-5	2/5	Higher education adopts AI unevenly and mostly for research/teaching support tasks, not for interpersonal advisory or mentorship roles like this one.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist with scheduling, documentation, or resource recommendations for student organizations, but the core advising function—mentorship, judgment, and accountability—offers limited scope for meaningful augmentation.
Augmentation potential	claude-sonnet-5	2/5	AI can help with scheduling, drafting communications, or organizing event logistics for the group, but it offers little assistance for the core mentoring and relational aspects of advising.
Task automatability	claude-haiku-4-5-20251001	1/5	Advising student organizations requires relationship-building, judgment about student development, and real-time responsiveness to individual circumstances—tasks that demand human presence and discretion that current AI cannot replicate end-to-end.
Task automatability	claude-sonnet-5	1/5	Advising student organizations requires ongoing personal mentorship, relationship-building, institutional knowledge, and judgment calls about student conduct and events that cannot be executed end-to-end by AI today.'
Adoption barriers	claude-haiku-4-5-20251001	5/5	Educational institutions have legal, accreditation, and liability requirements that a human faculty adviser must fulfill; student welfare, duty of care, and institutional accountability create hard barriers to AI substitution.
Adoption barriers	claude-sonnet-5	4/5	Institutions typically require a designated faculty member to be officially responsible for student organizations, including liability, safety, and disciplinary oversight, creating strong organizational and quasi-regulatory barriers to substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The loaded cost of a faculty member advising student organizations is already embedded in their salary; AI would need to both replace that advising and reduce overall institutional cost, neither of which is currently feasible.
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 at all.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably serves as a primary adviser to student organizations; this role fundamentally requires a licensed faculty member's judgment, accountability, and ongoing personal engagement with students.
Technical feasibility today	claude-sonnet-5	1/5	There are no deployed AI products that serve as faculty advisers to student clubs; this remains a purely human, relational role in practice.

Supervise students' laboratory or field work.

CI 0–5 · exposure 0 · augmentation 25 · 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	1/5	Postsecondary education, especially hands-on fields like anthropology and archeology, has low digitization and strong institutional conservatism; adoption of AI for direct student supervision remains negligible.
Sector adoption velocity	claude-sonnet-5	1/5	Higher education fieldwork supervision is a low-digitization, physically-embedded task with minimal AI adoption pressure or observed displacement.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI might assist with documentation or preliminary analysis of student work, but offers minimal augmentation to the core supervisory task of monitoring technique, ensuring safety, and providing real-time corrective feedback.
Augmentation potential	claude-sonnet-5	2/5	AI can help with prep materials, checklists, or data logging tools used during fieldwork, but offers little direct assistance to the supervisory act itself.
Task automatability	claude-haiku-4-5-20251001	1/5	Supervising students in laboratory or field work requires real-time presence, safety oversight, personalized feedback on technique, and adaptive response to unexpected situations—none of which current AI systems can perform end-to-end without a human supervisor remaining responsible.
Task automatability	claude-sonnet-5	1/5	Supervising students in physical lab or field settings (e.g., archaeological digs, artifact handling) requires real-time physical presence, safety oversight, and hands-on guidance that AI cannot perform end-to-end today.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Educational institutions have legal, regulatory, and accreditation requirements that students receive direct supervision by qualified faculty; liability for student safety and learning outcomes creates hard barriers to automation.
Adoption barriers	claude-sonnet-5	4/5	Institutional liability, safety regulations, and accreditation standards typically require a qualified faculty member to be physically present and responsible for student safety in field/lab settings.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The labor cost of human supervision is modest per student-hour and cannot be meaningfully replaced; AI monitoring tools, if developed, would add cost rather than reduce it without eliminating the need for human oversight.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute for physical on-site supervision, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably supervises hands-on student work in labs or field sites; this task fundamentally requires embodied human judgment, accountability, and duty of care that existing systems do not support.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product provides autonomous supervision of students' physical fieldwork or lab safety; this remains firmly in the human domain.

Conduct ethnographic field research.

CI 0–5 · exposure 0 · augmentation 50 · 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	1/5	Academic anthropology and archaeology remain low-digitization, human-centered disciplines where field research is foundational and irreplaceable. Adoption of AI for fieldwork automation is negligible; ethnographic method is defended as fundamentally human-centric.
Sector adoption velocity	claude-sonnet-5	1/5	Academic fieldwork in anthropology is a low-digitization, physically embedded practice with essentially no AI adoption trend for the core fieldwork activity itself.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist by helping organize, transcribe, code, and analyze field notes and interview recordings post-collection, improving a researcher's productivity in synthesis and pattern detection. However, the act of field observation and relationship-building itself remains solely human.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with transcription, translation, coding of field notes, literature review, and organizing data, meaningfully aiding researchers even though it cannot replace the fieldwork itself.
Task automatability	claude-haiku-4-5-20251001	1/5	Ethnographic field research requires sustained human presence, cultural immersion, rapport-building with subjects, and reflexive judgment about meaning and context that current AI cannot replicate. AI cannot meaningfully participate in participant observation or conduct unstructured interviews with the embodied, interpersonal trust required for ethnographic work.
Task automatability	claude-sonnet-5	1/5	Ethnographic fieldwork requires in-person immersion, building trust with communities, participant observation, and real-time interpretive judgment that current AI cannot physically or socially perform.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Ethnographic field research is intrinsically a human, interpretive social science practice that requires the researcher's own positionality, ethical responsibility, and deep cultural engagement. Academic and professional standards, institutional review boards (IRBs), and disciplinary practice all mandate human-led fieldwork with documented reflexivity.
Adoption barriers	claude-sonnet-5	4/5	Ethical review boards, informed consent requirements, cultural sensitivity, and the need for embodied human presence create strong institutional and practical barriers to any automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Ethnographic fieldwork is a high-touch, person-months-to-years endeavor requiring researcher salary and travel; even with AI assistance on analysis, the core cost is the human anthropologist's time in the field, which remains irreplaceable.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this task, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product conducts ethnographic field research autonomously. While AI can assist with transcription and coding of field notes, the core task—observing social practices, conducting interviews, and interpreting meaning through cultural immersion—remains entirely human-dependent in practice.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product conducts field research of this kind; this remains entirely a human, on-site, relationship-dependent activity.

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

CI 0–0 · exposure 0 · augmentation 38 · importance 3.1/5 · 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	Academic governance remains firmly rooted in human faculty participation with no meaningful adoption of AI substitutes; institutional inertia and regulatory requirements keep this task in the human domain.
Sector adoption velocity	claude-sonnet-5	1/5	Higher education governance structures are slow-moving, tradition-bound, and show essentially no displacement of human committee roles by AI systems.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could potentially assist with research preparation, policy analysis summaries, or document drafting before or after committee meetings, but it offers limited assistance during the deliberative core of the task itself.
Augmentation potential	claude-sonnet-5	3/5	AI can help draft meeting agendas, summarize policy documents, or prepare briefing notes for committee members, providing moderate assistance despite not replacing the deliberative role.
Task automatability	claude-haiku-4-5-20251001	1/5	Committee service requires nuanced judgment about institutional policies, stakeholder interests, and complex trade-offs that demand human discretion, debate, and accountability. AI cannot autonomously participate in deliberative bodies or cast binding votes on academic governance.
Task automatability	claude-sonnet-5	1/5	Committee service requires real-time deliberation, political judgment, negotiation, and representation of colleague/departmental interests that AI cannot perform end-to-end.dim
Adoption barriers	claude-haiku-4-5-20251001	5/5	Faculty governance relies on institutional bylaws, accreditation standards, and legal frameworks that explicitly require human faculty to serve on committees; many institutions have quorum and voting requirements that mandate human participation and accountability.
Adoption barriers	claude-sonnet-5	5/5	Committee membership is an institutional/governance role often requiring tenure status, elected or appointed authority, and formal accountability—hard structural and legal/organizational barriers exist.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Committee participation is a non-delegable duty of faculty membership tied to professional salary; there is no cost comparison because AI cannot substitute for the required human presence and decision-making authority.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI product performing this role, so no cost comparison favors AI; the human's institutional presence and accountability cannot be replaced by inference costs.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably serves on academic committees in production; this task is fundamentally tied to human participation, voting authority, and fiduciary responsibility that current AI systems cannot legally or practically assume.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product substitutes for a human serving on and voting/deliberating within academic governance committees.

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.