Biological Science Teachers, Postsecondary

25-1042.00
Median wage $84,620/yr50,190 employed (US)Rank #373 of 923 scored · top 40% by substitution

Teach courses in biological sciences. 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

Substitution31
Exposure28
Augmentation62

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

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

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

Tasks on the substitution scale

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%27

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

Technical feasibility todayw 20%28

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

Cost vs. human wagew 15%35

panel mean rating 2.4/5 → substitution pressure 35/100

Adoption barriersw 20%inverted — strong barriers lower the score38

panel mean rating 3.5/5 (barrier strength) → substitution pressure 38/100

Sector adoption velocityw 10%28

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

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.

95

CI 95–95 · exposure 100 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5This task is already deeply embedded in automated systems across virtually all postsecondary institutions; adoption of LMS-based attendance and grade recording is near-universal in higher education, representing the fastest possible velocity.
Sector adoption velocityclaude-sonnet-55/5Higher education has near-universal adoption of digital gradebooks and attendance systems already integrated into standard campus IT infrastructure.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assists faculty by auto-populating records from student information systems, flagging attendance patterns, and generating reports, thereby reducing manual data entry burden and improving record accuracy while the instructor retains oversight of grades and policy decisions.
Augmentation potentialclaude-sonnet-54/5Automated systems significantly reduce instructor burden for record maintenance, though instructors still input grades and verify accuracy, keeping them in the loop.
Task automatabilityclaude-haiku-4-5-202510015/5This is a routine administrative task involving data entry and record management that aligns perfectly with AI automation: capturing attendance, entering grades, and storing records are well-defined, repetitive processes that current learning management systems (Blackboard, Canvas, etc.) already automate at scale with no quality degradation.
Task automatabilityclaude-sonnet-55/5Recording attendance and grades is a structured data-entry task fully handled by existing LMS/gradebook software with automated calculation, syncing, and reporting, meeting the ≥50% time-saving bar easily.
Adoption barriersclaude-haiku-4-5-202510012/5While institutional IT policies and data security/FERPA compliance requirements create some friction, these are well-established and institutionally embedded; no licensing requirement mandates human record-keeping, and automation is already standard practice.
Adoption barriersclaude-sonnet-52/5Some institutional policy and data-privacy (FERPA) compliance requirements exist, but no licensing requirement mandates a human personally maintain these records, so barriers are modest.
Cost vs. human wageclaude-haiku-4-5-202510015/5LMS infrastructure costs per student per record are negligible (fractions of a cent per transaction) compared to the loaded wage of a faculty member or administrative staff member spending time on manual entry and maintenance.
Cost vs. human wageclaude-sonnet-55/5Institutional software licenses cost a small fraction of the faculty/staff time required to manually maintain these records, making automated systems an order of magnitude cheaper per record-keeping unit.
Technical feasibility todayclaude-haiku-4-5-202510015/5Deployed products demonstrably perform this task reliably in production across thousands of educational institutions worldwide; learning management systems with attendance tracking, grade books, and automated record-keeping are mature, widely adopted infrastructure in postsecondary education.
Technical feasibility todayclaude-sonnet-55/5Learning management systems (Canvas, Blackboard, PowerSchool, etc.) already perform automated grade calculation, attendance tracking, and recordkeeping reliably in production across nearly all universities.

Compile bibliographies of specialized materials for outside reading assignments.

84

CI 71–97 · exposure 83 · augmentation 100 · importance 3.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Academic institutions and faculty increasingly use digital tools (Zotero, Mendeley, ChatGPT, semantic search) for bibliography work. Adoption is rapid in higher education and research-heavy sectors, though some traditional departments lag.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI tools for research and content preparation at a moderate pace, with many faculty experimenting but institutional/course-level integration still uneven.
Augmentation potentialclaude-haiku-4-5-202510015/5AI dramatically assists professors by rapidly filtering and organizing literature, allowing them to focus on pedagogical judgment (selecting the most relevant sources for learning objectives) rather than manual compilation—a classic case of human-in-the-loop productivity gain.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up literature discovery and list compilation, letting instructors quickly draft and refine bibliographies while still applying subject expertise to vet quality and relevance.
Task automatabilityclaude-haiku-4-5-202510015/5AI systems can now reliably compile bibliographies from academic databases, search scholarly literature, format citations, and organize materials by topic or reading level—all core components of this task. Modern LLMs and research tools can generate comprehensive, properly formatted bibliographies in seconds, easily exceeding the 50% time-saving threshold with minimal human oversight.
Task automatabilityclaude-sonnet-54/5AI systems can quickly generate topic-relevant reading lists and bibliographies, especially with access to academic databases or search tools, though verification of accuracy and currency is still needed for specialized biology content.
Adoption barriersclaude-haiku-4-5-202510012/5There are minimal legal or regulatory barriers; no license requirement exists to compile reading lists. The main friction is institutional inertia and instructor preference for human curation to ensure pedagogical fit, but these are soft barriers, not hard ones.
Adoption barriersclaude-sonnet-51/5There is no licensing, regulatory, or liability barrier to using AI-assisted tools for compiling reading lists; it's a low-stakes administrative/preparatory task.
Cost vs. human wageclaude-haiku-4-5-202510015/5API costs for AI-powered bibliography compilation are negligible (cents per task), compared to a professor's hourly wage (typically $40–80+). Integration overhead is minimal given widespread adoption of free or low-cost tools.
Cost vs. human wageclaude-sonnet-55/5Generating a bibliography via AI tools takes minutes and costs a fraction of a cent compared to the faculty time otherwise spent searching and compiling references.
Technical feasibility todayclaude-haiku-4-5-202510015/5Multiple mature, production-ready systems perform this task: Zotero, Mendeley, and AI-enhanced academic search tools (Google Scholar, semantic search via OpenAI plugins, ChatGPT) reliably compile and format bibliographies at scale. Libraries and educational institutions actively deploy these for bibliography generation.
Technical feasibility todayclaude-sonnet-53/5Products like AI research assistants and citation tools can generate bibliographies today, but they sometimes hallucinate citations or miss the latest specialized literature, requiring instructor review.

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

76

CI 76–76 · exposure 75 · augmentation 100 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Educational institutions show mixed adoption: early adopters in online/distance education use AI heavily, but traditional postsecondary biology departments remain cautious about AI-generated lab materials due to accuracy and liability concerns around laboratory safety.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting generative AI for content creation at a moderate pace, with growing use but institutional caution around academic integrity and course quality controls slowing full integration.
Augmentation potentialclaude-haiku-4-5-202510015/5AI significantly accelerates faculty productivity by drafting and iterating materials, allowing instructors to focus on customization, learning design, and validation rather than writing from scratch—a clear assistive transformation of the workflow.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up drafting syllabi, homework problems, and handouts, letting instructors focus on customization, pedagogy, and lab-specific safety considerations.
Task automatabilityclaude-haiku-4-5-202510014/5AI can generate syllabi, homework assignments, and handouts from course outlines with minimal human review, achieving >50% time savings. However, customization for specific learning objectives, institutional requirements, and laboratory safety protocols typically requires human oversight, preventing a full 5-rating.
Task automatabilityclaude-sonnet-54/5AI can draft syllabi, homework assignments, handouts, and lab protocols quickly given course parameters, saving significant time though instructor review and customization for specific lab equipment/safety is still needed.'
Adoption barriersclaude-haiku-4-5-202510012/5No licensing requirement mandates a human prepare these materials; institutional policies and faculty preference for control over content provide modest friction, but nothing legally prevents automation or requires human sign-off.
Adoption barriersclaude-sonnet-52/5No licensing requirement governs drafting course materials, though institutional academic policies and accreditation standards may require faculty ownership/review of final content, creating mild friction.
Cost vs. human wageclaude-haiku-4-5-202510015/5API costs for generating these text-based materials are negligible ($0.01–0.10 per assignment) compared to faculty labor ($40–80/hour for material preparation), representing orders of magnitude cost savings.
Cost vs. human wageclaude-sonnet-55/5Generating draft course materials via AI costs pennies compared to the faculty time (often billed at high hourly rates) needed to write these from scratch.
Technical feasibility todayclaude-haiku-4-5-202510014/5LLM-based tools (ChatGPT, Claude, specialized educational platforms) reliably generate course materials in production settings. Deployed systems can create homework, handouts, and syllabi at scale, though educators commonly refine outputs for quality and accuracy assurance.
Technical feasibility todayclaude-sonnet-54/5Deployed LLM tools (ChatGPT, Claude, specialized ed-tech platforms) are routinely used by instructors today to generate syllabi and assignment drafts, though lab-specific materials requiring safety protocols and specific equipment need more human customization.

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

64

CI 50–79 · exposure 62 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Higher education and postsecondary institutions have rapidly adopted LMS platforms and auto-grading tools over the past decade; COVID-19 acceleration pushed adoption deeper. Most colleges now use some form of automated exam administration, though adoption rates for subjective grading remain more mixed.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI grading and question-generation tools at a moderate pace, with many pilots and some scaled use in large courses, but adoption is uneven across institutions and disciplines.
Augmentation potentialclaude-haiku-4-5-202510014/5AI strongly augments faculty productivity by handling routine compilation, question-bank curation, batch grading, and analytics (identifying weak topics, outlier grades). Instructors can focus on reviewing flagged essays and providing targeted feedback rather than transcribing scores, substantially raising their output.
Augmentation potentialclaude-sonnet-54/5AI substantially helps instructors draft exam questions, create rubrics, and pre-grade responses, meaningfully speeding up the workflow while the instructor retains final judgment.
Task automatabilityclaude-haiku-4-5-202510014/5AI can compile exams from question banks, administer them via platforms, and auto-grade objective questions (multiple choice, true/false) at scale with minimal human oversight. However, grading subjective responses (essays, lab reports) still requires significant human judgment, preventing full end-to-end automation. Current systems can likely achieve 50%+ time savings on the compilation and objective grading components.
Task automatabilityclaude-sonnet-53/5AI can generate exam questions and grade objective or short-answer responses well, but compiling exams aligned to specific course content and grading nuanced biology explanations still needs instructor oversight, so only partial time savings are realized end-to-end.
Adoption barriersclaude-haiku-4-5-202510012/5Few regulatory or licensing barriers prevent automation of exam administration and grading. Most institutions can adopt learning-management-system tools without special authorization. The main frictions are organizational (faculty resistance to reducing hands-on grading) and modest concerns about assessment validity, but no hard legal requirement for human sign-off exists.
Adoption barriersclaude-sonnet-53/5No licensing requirement mandates a human grade exams, but academic integrity standards, institutional grading policies, and accreditation expectations create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-driven exam systems cost only server/licensing fees per student, typically $1–10 per exam. A faculty member grading 100 exams at $50/hour loaded wage would cost ~$200–400 in labor. The AI cost is negligible by comparison, especially for objective grading.
Cost vs. human wageclaude-sonnet-53/5AI tools reduce time spent on drafting and initial grading but still require faculty verification and calibration, so overall cost savings are moderate rather than order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple deployed products reliably perform exam administration and objective auto-grading in production (Canvas, Blackboard, Examsoft, etc.). These systems handle at-scale grading of standardized questions with high accuracy. Subjective-answer grading remains lower-fidelity but assistive tools (rubric-guided feedback systems) are increasingly present in production learning management systems.
Technical feasibility todayclaude-sonnet-53/5Products like AI-assisted quiz generators and automated grading tools (e.g., Gradescope, LLM-based graders) are deployed in higher ed, but accuracy on open-ended science answers remains imperfect and requires human review.

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

61

CI 48–74 · exposure 62 · augmentation 88 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Higher education is rapidly adopting AI grading assistants in LMS platforms (Canvas, Blackboard integrations) and standalone tools; pilot and early-production use is widespread among postsecondary institutions.
Sector adoption velocityclaude-sonnet-52/5Higher education is a relatively slow-adopting sector for grading automation due to academic culture, integrity concerns, and heterogeneous course content, with pilots more common than widespread production use.
Augmentation potentialclaude-haiku-4-5-202510015/5AI dramatically assists faculty by pre-grading assignments, flagging outliers, and providing detailed feedback summaries, enabling instructors to focus on high-stakes or complex evaluations while significantly reducing manual grading time.
Augmentation potentialclaude-sonnet-54/5AI substantially assists instructors by drafting feedback, flagging errors, and speeding up grading of routine assignments, while the instructor retains final judgment and accountability.
Task automatabilityclaude-haiku-4-5-202510014/5AI systems can now evaluate and grade objective components (factual accuracy, calculation correctness, structure) and provide preliminary grades on written work with high consistency. However, nuanced judgment on experimental design reasoning and originality in lab work still benefits from human review, preventing a full 5.
Task automatabilityclaude-sonnet-53/5AI can grade objective assessments and provide first-pass feedback on written assignments, but nuanced evaluation of lab work, scientific reasoning, and academic integrity still requires human judgment, so only partial time savings are achievable at equal quality.
Adoption barriersclaude-haiku-4-5-202510013/5Academic institutions have oversight requirements (faculty verification, appeals processes) and cultural preference for human grading authority, but no strict legal barrier prevents delegating grading to AI. Institutional friction and accreditation norms create friction without hard prohibition.
Adoption barriersclaude-sonnet-53/5No formal licensing barrier exists for grading, but institutional academic integrity concerns, accreditation standards, and instructor accountability for final grades create meaningful friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510015/5Per-task inference cost for grading is negligible (pennies per assignment), while faculty labor costs run $50–100+ per hour; the ratio is at least 100:1 in favor of AI once integrated.
Cost vs. human wageclaude-sonnet-53/5AI grading tools reduce time spent per assignment but still require faculty oversight and calibration, so total cost savings are moderate rather than order-of-magnitude given the need for accuracy checks in a specialized science domain.
Technical feasibility todayclaude-haiku-4-5-202510014/5Mature products (LLM-based grading tools, learning management system integrations with AI backends) reliably grade assignments and papers in production at many institutions. Error rates on standardized rubrics are low, though subjective elements still require human spot-check.
Technical feasibility todayclaude-sonnet-53/5Products like Gradescope and LLM-based grading assistants are used in production for essays and structured problem sets, but reliability drops for open-ended lab reports and complex scientific content, requiring instructor review.

Assist students who need extra help with their coursework outside of class.

44

CI 37–50 · exposure 34 · augmentation 75 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Postsecondary education remains a laggard sector in AI adoption despite recent hype; most institutions still rely on traditional office hours and human peer tutors. While some schools pilot AI writing assistants, systemic replacement of instructor-led tutoring remains rare and experimentally contested.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI tutoring tools and chatbots at a moderate pace, with pilots and some formal integration but many students still relying on office hours and TAs.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can assist instructors by generating practice problems, drafting explanations of complex concepts, or flagging common student misconceptions from logs—allowing instructors to focus on one-on-one mentoring. AI-powered study tools and instant feedback loops meaningfully amplify instructor capacity and student autonomy.
Augmentation potentialclaude-sonnet-54/5AI tools substantially help students review material, get practice explanations, and prepare questions before or after seeking human help, meaningfully extending available support hours.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can provide basic tutoring and answer factual biology questions, the task requires diagnosing individual student knowledge gaps, providing personalized explanations, and emotional support—elements that demand human judgment and adaptability. Current AI tutoring systems cannot reliably replicate the nuanced, interactive scaffolding that effective remedial instruction requires.
Task automatabilityclaude-sonnet-52/5AI tutoring tools can explain biology concepts and answer questions, but genuine student support involves diagnosing individual misunderstandings, motivation, and personalized mentorship that current systems only partially replicate.
Adoption barriersclaude-haiku-4-5-202510014/5Faculty have contractual obligations (office hours, student support) and institutions have accreditation expectations that students receive instructor engagement. Liability concerns around incorrect science instruction, student expectation for human availability, and institutional inertia create strong friction against pure AI substitution for academic support.
Adoption barriersclaude-sonnet-52/5No licensing requirement blocks AI-assisted tutoring, though institutional policies, academic integrity concerns, and student preference for human mentorship create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510014/5The inference cost of AI tutoring is very low (pennies per session), whereas a postsecondary instructor's loaded wage for office hours or tutoring is substantial (typically $50–100+ per hour). Integration and oversight costs are minimal for scalable AI systems.
Cost vs. human wageclaude-sonnet-54/5AI tutoring costs pennies per session versus a professor's or TA's hourly wage, making it far cheaper for basic Q&A support, though oversight and curriculum alignment add some cost.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI tutoring products (e.g., ChatGPT, specialized ed-tech platforms) exist and can deliver content explanations and practice problems, but they lack robust assessment of misconceptions, struggle with complex conceptual questions in biology, and cannot replicate the rapport-building and motivation that human instructors provide. Deployment is limited and often used as supplement, not replacement.
Technical feasibility todayclaude-sonnet-53/5AI chatbots and tutoring platforms (e.g., ChatGPT, Khan Academy's Khanmigo) are deployed and used by students for extra help, though reliability varies and they don't replace office-hours interaction with an instructor.

Review papers for publication in journals.

41

CI 25–57 · exposure 45 · augmentation 63 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Academic publishing is digitizing and adopting AI screening tools, but adoption remains concentrated in large publishers and selective journals; most university-based review remains manual. Pilots are common; production deployment of AI as a core reviewer decision-maker is rare, reflecting professional and reputational hesitance in the sector.
Sector adoption velocityclaude-sonnet-52/5Academic publishing has been slow to adopt AI as a substitute for reviewers, with cautious pilot programs and explicit bans on AI review at many major journals due to confidentiality and accountability concerns. Adoption is currently limited to peripheral support tools rather than substitution.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants significantly augment human reviewers by rapidly summarizing papers, cross-referencing citations, identifying methodological red flags, and flagging statistical inconsistencies—allowing experienced reviewers to focus on novelty, significance, and fit. This productivity boost is widely demonstrable and preserves human expertise in the loop.
Augmentation potentialclaude-sonnet-53/5AI can help reviewers by summarizing papers, checking statistical methods, flagging plagiarism, or identifying missing citations, improving efficiency of the review process. However, the core evaluative judgment remains human-driven, so augmentation is moderate rather than transformative.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems can automatically extract key methodological, statistical, and structural elements from papers, flag common issues (plagiarism, missing details, basic inconsistencies), and generate detailed technical feedback on ~70–80% of typical review concerns. However, expert judgment on novelty, significance, and fit-for-journal decisions requires human oversight, preventing full end-to-end automation at the ≥50% time-saving threshold without substantive review modification.
Task automatabilityclaude-sonnet-52/5Reviewing scientific papers requires deep domain expertise, judgment about novelty and validity, and contextual understanding of a field's literature that current AI cannot reliably replicate end-to-end. AI can assist with checks like statistics, plagiarism, or formatting but cannot independently render a substantive, trustworthy peer review.
Adoption barriersclaude-haiku-4-5-202510014/5Peer review legitimacy and journal reputation depend on human expert accountability; most high-impact journals have explicit policies requiring human reviewers and editorial decisions. Liability concerns (flawed recommendations affecting publication), professional norms, and accreditation standards create strong organizational and contractual barriers to full AI substitution of reviewer judgment.
Adoption barriersclaude-sonnet-54/5Peer review is deeply embedded in scholarly norms requiring recognized expert judgment, with editorial policies at most journals mandating human reviewers for legitimacy and accountability. Using AI as the primary reviewer would raise ethical and confidentiality concerns given unpublished manuscript content.
Cost vs. human wageclaude-haiku-4-5-202510014/5An automated pre-screening pass using commercial LLMs or specialized academic AI typically costs $5–50 per paper, whereas a postdoctoral or faculty reviewer's time (3–4 hours at ~$50–100/hour loaded) runs $150–400+ per review. AI-assisted workflows deliver at least 3–5× cost advantage on the frontline screening phase.
Cost vs. human wageclaude-sonnet-52/5While AI inference is cheap, the oversight and verification needed to trust an AI-generated review of complex scientific work adds substantial cost, and unreliable output could damage journal credibility. Human reviewers, though unpaid typically, represent a sunk institutional cost that AI does not straightforwardly displace at lower cost given quality risks.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI-powered manuscript screening tools (e.g., Elicit, Scholarcy, AI-assisted review platforms) are deployed in some academic workflows and reliably perform components like summary generation, reference checking, and style/structure analysis. However, production deployment for core editorial/peer-review decisions remains patchy; most journals still rely on human reviewers, and AI systems have material error rates on nuanced scientific judgment.
Technical feasibility todayclaude-sonnet-52/5Some journals experiment with AI-assisted screening or plagiarism/statistics checks, but no deployed product independently performs full peer review reliably in production. Current tools remain adjuncts to human reviewers rather than substitutes.

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

36

CI 25–46 · exposure 30 · augmentation 75 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5While academic institutions use some AI-assisted literature tools, adoption remains slow and partial; most faculty still rely primarily on manual reading, email alerts, and in-person conferences, with AI serving only a supplementary role in most postsecondary environments.
Sector adoption velocityclaude-sonnet-53/5Academia has moderate AI adoption for literature review and search tools, with growing but not yet deep integration into daily scholarly workflows.'
Augmentation potentialclaude-haiku-4-5-202510014/5AI excels at augmenting this task by rapidly summarizing papers, identifying trends across large literatures, and filtering noise—allowing academics to spend more time on selective reading and deeper synthesis while AI handles routine scanning and categorization.
Augmentation potentialclaude-sonnet-54/5AI substantially aids literature discovery, summarization, and trend-spotting, meaningfully boosting a professor's ability to stay current while they still attend conferences and engage with peers themselves.'
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with summarizing literature and flagging relevant papers, but the task fundamentally requires human judgment about which developments matter, selective engagement with peers at conferences, and building professional relationships—activities that resist full automation and wouldn't achieve 50% time savings end-to-end.
Task automatabilityclaude-sonnet-52/5AI can help summarize and surface relevant literature, but genuinely keeping abreast requires synthesis, judgment, and social interaction (colleague discussions, conferences) that cannot be fully automated end-to-end today.'
Adoption barriersclaude-haiku-4-5-202510014/5Academic norms, professional licensing (tenure systems), peer review structures, and institutional expectations strongly favor human engagement with the field; there is also inherent value placed on human judgment about research significance that creates resistance to AI replacement.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but tenure and academic norms place value on personal engagement, conference presence, and networking, creating moderate organizational and cultural friction against full substitution.'
Cost vs. human wageclaude-haiku-4-5-202510012/5While AI tools for literature monitoring are inexpensive, the task includes irreplaceable human activities (conference attendance, relationship-building) that cannot be substituted; the cost advantage of AI assistance is marginal relative to the human time still required.
Cost vs. human wageclaude-sonnet-53/5AI literature tools are cheap relative to a professor's time spent reading, but the task also includes non-automatable social components, so overall cost savings are only partial.'
Technical feasibility todayclaude-haiku-4-5-202510013/5Current AI tools (literature summarizers, semantic search, alerting systems) exist and are used in academic settings, but they often miss nuanced significance and fail to replicate the serendipitous learning and networking that define conference participation and colleague conversations.
Technical feasibility todayclaude-sonnet-52/5Tools like literature summarizers and research alert systems exist and are used by academics, but no deployed product autonomously replicates the full scope of staying current including networking and conference participation.'

Write grant proposals to procure external research funding.

31

CI 25–36 · exposure 25 · augmentation 63 · importance 3.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Academia lags in AI adoption; grant writing remains highly credentialed and personalized work. While some universities pilot AI writing assistants, systematic adoption in grant proposal generation is minimal and cautious due to quality concerns and institutional governance.
Sector adoption velocityclaude-sonnet-53/5Academia has moderate AI adoption for writing assistance, with growing use of tools like ChatGPT for drafting, though many institutions have policies constraining full use in grant submissions.
Augmentation potentialclaude-haiku-4-5-202510013/5Current AI tools can assist by drafting sections (background, methods overview), organizing citations, and catching formatting errors, meaningfully reducing baseline writing burden. However, augmentation is limited to scaffolding; the core intellectual work of proposal strategy and justification remains substantially human.
Augmentation potentialclaude-sonnet-54/5AI substantially helps with literature summaries, drafting narrative sections, formatting, and editing, meaningfully speeding up the writing process while the researcher retains control over scientific content and strategy.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can draft text and organize sections of grant proposals, the task requires novel scientific justification, strategic positioning within funder priorities, and integration of an individual researcher's unique contributions. Current AI tools cannot reliably generate the distinctive intellectual content and strategic framing needed for competitive proposals, making end-to-end automation far below the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5AI can draft sections and boilerplate but writing a competitive grant requires original research framing, novel hypotheses, and strategic positioning that current AI cannot reliably generate at equal quality end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Grant authorship and institutional sign-off carry legal and reputational stakes; institutions typically require human faculty ownership and accountability for proposal content. Funder policies and institutional review boards impose oversight requirements that embed human judgment into the process.
Adoption barriersclaude-sonnet-53/5No licensing requirement, but funding agencies expect PI-authored intellectual content, institutional sign-off, and accountability for accuracy, creating moderate organizational and reputational barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI tools (ChatGPT, specialized writing assistants) cost pennies per proposal draft, but a faculty member's time to substantially revise, verify claims, and ensure intellectual integrity remains the dominant cost. The human effort required to make AI output acceptable remains comparable to writing from scratch.
Cost vs. human wageclaude-sonnet-53/5AI drafting assistance is cheap relative to faculty time, but the necessary human expert review, data generation, and strategic revision keep overall cost savings moderate rather than order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably writes competitive grant proposals end-to-end. AI writing assistants exist, but they produce generic drafts requiring extensive expert revision. Grant success depends on domain expertise, strategic fit assessment, and funder-specific knowledge that deployed systems cannot consistently deliver.
Technical feasibility todayclaude-sonnet-52/5Products like ChatGPT and specialized grant-writing tools exist and are used for drafting and editing, but no deployed system reliably produces fundable full proposals without heavy expert revision.

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

29

CI 23–35 · exposure 25 · augmentation 50 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic institutions, especially in laboratory procurement, remain traditional and slow-moving. Procurement processes are often paper-based or legacy systems; adoption of AI-driven selection remains minimal and experimental even in 2024.
Sector adoption velocityclaude-sonnet-52/5Higher education administrative/procurement functions are adopting AI slowly compared to core information-work sectors, with most institutions still using manual or legacy procurement systems.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could usefully assist by generating supplier comparisons, flagging equipment that meets safety standards, or summarizing product reviews. However, the core task of evaluating pedagogical fit and making institutional selections still fundamentally requires human expertise and judgment.
Augmentation potentialclaude-sonnet-53/5AI can meaningfully help by suggesting textbook options, comparing supplier catalogs, and drafting purchase requests, improving efficiency while the instructor makes final decisions.
Task automatabilityclaude-haiku-4-5-202510012/5While AI could assist with identifying and cataloging materials or generating procurement lists, the task requires human judgment about suitability, budget constraints, and institutional needs. The final selection and actual procurement involve negotiation, vendor relationship management, and approval workflows that resist full automation.
Task automatabilityclaude-sonnet-52/5AI can help research and shortlist textbooks or lab supplies, but the final selection requires judgment about curriculum fit, budget approval, and physical procurement steps that AI cannot execute end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Universities typically require faculty or authorized procurement staff to select materials based on accreditation standards, safety certifications, and institutional policies. Signature authority and liability for equipment selection rest with credentialed humans, creating a hard requirement that limits automation.
Adoption barriersclaude-sonnet-52/5No licensing requirement, but institutional purchasing procedures, budget approval chains, and vendor relationships create organizational friction beyond pure AI capability.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI integration into procurement workflows would require substantial initial setup and custom configuration for institutional requirements. The human cost of a faculty member or procurement specialist reviewing selections and managing relationships remains lower than the bundled cost of AI infrastructure and oversight per acquisition cycle.
Cost vs. human wageclaude-sonnet-52/5AI assistance for research is cheap, but the actual procurement, vendor negotiation, and physical logistics still require human labor and administrative systems, limiting overall savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5Some procurement systems have light AI features for cataloging and recommendation, but no deployed product reliably end-to-end selects, justifies, and obtains specialized laboratory materials for academic institutions. Most universities still rely on human domain experts working with purchasing departments.
Technical feasibility todayclaude-sonnet-52/5No deployed product autonomously handles ordering and procuring physical lab equipment and textbooks; existing tools only assist with recommendations or comparison shopping.

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

28

CI 23–34 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education adoption of AI for core advising functions remains limited; advising is still viewed as a distinctly human faculty responsibility. Most institutions use AI only for peripheral tasks (scheduling, information lookup) rather than substantive curriculum or career guidance.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools unevenly and cautiously, especially for advising functions tied to student outcomes and institutional accreditation concerns.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist advisors by retrieving curriculum requirements, summarizing student transcripts, suggesting relevant career pathways, and managing administrative tasks, thereby improving productivity for the faculty advisor who retains primary responsibility and judgment.
Augmentation potentialclaude-sonnet-54/5AI can help faculty prepare curriculum information, summarize career pathways, and draft guidance material, meaningfully speeding up preparation even though the human interaction remains central.
Task automatabilityclaude-haiku-4-5-202510012/5Advising on academic and vocational curricula and career issues requires understanding individual student circumstances, aspirations, and institutional pathways. While AI can provide information retrieval and suggestions, the personalized judgment, mentorship, and relationship-building aspects resist full automation at the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5Advising requires personalized judgment about a student's specific academic history, goals, and institutional context; AI can supply general information but not authoritative, accountable advising end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Postsecondary institutions typically embed advising responsibility in faculty roles, and there is strong institutional expectation that faculty, not automated systems, provide mentorship and career guidance. Regulations, accreditation standards, and organizational culture all favor human advisors; liability concerns arise when students follow poor automated advice.
Adoption barriersclaude-sonnet-53/5No licensing requirement for advising itself, but institutional policy, liability for poor guidance, and student preference for personal relationships with faculty create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5The loaded cost of a tenured postsecondary faculty advisor (with benefits and institutional overhead) is substantial, but automating meaningful advising would require sophisticated AI oversight and integration. The cost advantage of an AI system would not be decisive compared to human faculty already employed for teaching.
Cost vs. human wageclaude-sonnet-53/5AI tools can cheaply handle routine informational queries, but a full advising interaction with follow-up and judgment still requires paid faculty/advisor time, making costs roughly comparable when quality is held constant.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs holistic academic and career advising at scale in higher education today. Chatbots can answer procedural questions, but genuine advising—evaluating student fit for programs, discussing career trajectories, and navigating complex institutional constraints—remains human-centric in practice.
Technical feasibility todayclaude-sonnet-52/5Chatbots and advising-support tools exist at some universities, but they mostly handle FAQs and scheduling rather than substantive academic/vocational/career counseling reliably in production.

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

28

CI 25–30 · exposure 25 · augmentation 75 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Postsecondary institutions show slow adoption of AI for curriculum planning; adoption remains mostly experimental (pilots, drafting aids) rather than production replacement. Faculty resistance to algorithmic curriculum design and governance structures preserve human control.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools slowly and unevenly for curriculum design, with most use confined to individual instructor experimentation rather than institutional-scale deployment.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist faculty by drafting syllabi, suggesting learning activities, analyzing content alignment, and generating assessment materials, substantially raising productivity while the instructor retains full authority over pedagogical decisions and institutional compliance.
Augmentation potentialclaude-sonnet-54/5AI is genuinely useful for brainstorming course topics, generating drafts of syllabi, creating assessment questions, and summarizing new research to incorporate into curriculum, meaningfully speeding up an instructor's planning process.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with generating draft course materials and analyzing pedagogical approaches, but curriculum planning requires pedagogical judgment, learning outcome alignment, institutional requirements, and iterative refinement that humans must oversee. Full end-to-end automation achieving 50% time savings at equal quality is not demonstrated.
Task automatabilityclaude-sonnet-52/5AI can draft syllabi and suggest content updates, but planning and revising curricula requires judgment about pedagogy, accreditation standards, and departmental context that current AI cannot reliably handle end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Curriculum decisions rest with faculty expertise and institutional authority; accreditation bodies, department governance, and professional standards require human faculty ownership of instructional design. Institutions rarely delegate core pedagogical decisions to AI without human sign-off.
Adoption barriersclaude-sonnet-53/5No licensing requirement per se, but curriculum decisions typically require faculty expertise, departmental/accreditation approval, and academic freedom norms that create institutional friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5While AI tools are cheap per use, curriculum design demands substantial human expert review, iteration, and accountability. The total cost of oversight and revision approaches or exceeds the human professor's time, negating cost advantage.
Cost vs. human wageclaude-sonnet-52/5AI drafting assistance is cheap, but the human oversight, subject-matter validation, and institutional approval processes still dominate the cost, keeping overall savings modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI tools (ChatGPT, Claude) can generate course outlines and materials, but no deployed product reliably plans, evaluates, and revises complete curricula to institutional standards. Existing systems lack deep understanding of accreditation requirements, student cohort needs, and disciplinary progression.
Technical feasibility todayclaude-sonnet-52/5Products like ChatGPT and specialized ed-tech tools can generate course outlines and materials, but no deployed product autonomously plans and revises full postsecondary biology curricula in production.

Participate in student recruitment, registration, and placement activities.

28

CI 25–30 · exposure 25 · augmentation 50 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education is relatively slow to automate student-facing functions; while some institutions use chatbots for registration, most recruitment and placement remain human-driven. Adoption of AI for these tasks remains in the pilot phase in most institutions rather than deep production deployment.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI slowly for interpersonal advising and recruitment tasks, with most AI use limited to marketing and basic chat support rather than faculty-level participation.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can augment this task by automating routine scheduling, flagging candidate profiles for review, and providing background research on placement opportunities. However, augmentation is modest because the task fundamentally depends on human judgment and relationship-building with students.
Augmentation potentialclaude-sonnet-53/5AI can help faculty by automating outreach emails, summarizing applicant data, or scheduling, providing moderate productivity gains while humans retain the relational and evaluative aspects of recruitment and placement.
Task automatabilityclaude-haiku-4-5-202510012/5This task requires relationship-building, interpersonal persuasion, and nuanced judgment about student fit and placement outcomes. While AI could assist with administrative screening and scheduling, the core recruitment and placement decisions depend on human judgment and authentic engagement that current AI systems cannot reliably automate at 50% time savings.
Task automatabilityclaude-sonnet-52/5This task involves interpersonal recruitment, advising, and administrative coordination that current AI can partially support (e.g., drafting materials, scheduling) but cannot fully replace given the relational and judgment-based nature of student placement decisions.rebalancing.rebalance{},
Adoption barriersclaude-haiku-4-5-202510014/5Institutional and professional norms strongly favor human involvement in student recruitment, advising, and placement—it is often explicitly part of faculty responsibility and affects student experience and institutional reputation. Educational accreditation and institutional policies typically require human discretion in placement decisions.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement compels faculty involvement, but institutional norms, personalized advising expectations, and accreditation/administrative structures create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for administrative support (chatbots, scheduling automation) are relatively cheap, but they only handle peripheral tasks. The core recruitment and placement work still requires invested faculty or staff time, making the all-in cost per placement comparable to or higher than human-only approaches.
Cost vs. human wageclaude-sonnet-52/5AI tools can reduce some administrative overhead but faculty involvement in recruitment/placement still requires substantial human time for interviews, advising, and judgment calls, keeping costs comparable to human-only processes.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs end-to-end student recruitment, registration, and placement. Chatbots can handle basic inquiries and registration logistics, but identifying suitable placements and recruiting candidates require human expertise and relationship-building that exceeds current AI capabilities in production.
Technical feasibility todayclaude-sonnet-52/5Some CRM and chatbot tools assist with recruitment outreach and registration logistics in higher ed, but faculty-level participation in placement decisions is not handled by deployed AI products today.

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

26

CI 25–28 · exposure 25 · augmentation 75 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Academic and research institutions have adopted AI tools as assistants (writing aids, data analysis) slowly and cautiously, with skepticism toward AI in research design and experimental work. Disciplinary norms, tenure systems, and risk-averse funding bodies have resulted in minimal replacement of human research activity; adoption remains exploratory rather than transformative.
Sector adoption velocityclaude-sonnet-53/5Academic research is adopting AI tools (literature search, writing assistance, data analysis) at a moderate pace, but core experimental and analytical processes in biology remain slower to fully integrate AI agents.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly augments research productivity by accelerating literature reviews, automating routine statistical analyses, drafting manuscript sections, and suggesting experimental designs—tasks that occupy substantial research time. Biological science teachers using AI assistants for these sub-tasks report material gains in efficiency while maintaining human oversight and decision-making.
Augmentation potentialclaude-sonnet-54/5AI significantly boosts productivity in literature review, hypothesis generation support, drafting manuscripts, and statistical analysis, while the researcher retains control over experimental design and interpretation.
Task automatabilityclaude-haiku-4-5-202510012/5Research design, hypothesis formation, and experimental execution require deep domain expertise and human judgment that current AI cannot reliably replace. While AI can assist with literature review, data analysis, and manuscript drafting, the creative and novel contribution that defines publishable research remains fundamentally dependent on human researchers; only narrow, routine analytical sub-tasks approach the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5AI can assist with literature review, drafting, and data analysis, but original hypothesis generation, wet-lab experimentation, and rigorous scientific judgment remain largely human-driven and cannot be fully automated at equal quality today.
Adoption barriersclaude-haiku-4-5-202510014/5Significant barriers protect this task: peer review processes require human expertise and accountability, ethical review boards (IRBs) mandate human responsibility for research design, and institutional/funding requirements typically demand that named researchers (humans) take authorship and liability. Publication standards inherently require human verification and creative contribution.
Adoption barriersclaude-sonnet-54/5Publication requires human authorship, accountability, and peer review; academic norms, authorship ethics, and institutional/tenure requirements create strong barriers to AI substituting the human researcher.
Cost vs. human wageclaude-haiku-4-5-202510012/5The loaded salary of a postseondary biological scientist is high ($70k–$120k+ annually), and the cost of AI-assisted tools (language models, analysis services) remains modest per task. However, the human researcher's salary is the baseline for comparison, and current AI cannot reduce the overall research cost because humans remain the bottleneck in novelty and rigor.
Cost vs. human wageclaude-sonnet-52/5AI can cut costs for literature review and manuscript drafting, but the core research (experiments, data collection, peer-reviewed validation) still requires expensive human labor and equipment, keeping overall costs comparable to human-led research.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably conducts end-to-end research in biological sciences or manages the full publication workflow autonomously. AI tools exist for writing and data analysis, but they function as assistants requiring extensive human oversight; they cannot independently conceive, execute, or validate novel research at publication quality.
Technical feasibility todayclaude-sonnet-52/5Tools like AI-assisted literature search, writing aids, and data analysis are deployed, but no product independently conducts biological research and publishes credible findings without substantial human oversight and experimentation.

Prepare and deliver lectures to undergraduate or graduate students on topics such as molecular biology, marine biology, and botany.

26

CI 23–30 · exposure 25 · augmentation 75 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education institutions have been slow to adopt AI-driven instruction at scale; lectures remain a core institutional practice controlled by faculty governance. Pilot use of AI lecture aids exists but full displacement remains rare.
Sector adoption velocityclaude-sonnet-52/5Higher education adoption of AI for teaching is happening but remains cautious and pilot-stage, with institutional and cultural resistance to replacing human lecturers.
Augmentation potentialclaude-haiku-4-5-202510014/5AI already assists faculty by drafting lecture outlines, generating practice problems, creating visual aids, and explaining complex topics in multiple formats. These tools materially raise preparation efficiency while instructors retain full pedagogical control and student contact.
Augmentation potentialclaude-sonnet-54/5AI substantially aids lecture preparation—outlining content, generating visuals, summarizing research, and answering student queries—significantly boosting instructor productivity while the professor still delivers the lecture.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can draft lecture content and generate visual materials, delivering effective lectures requires real-time audience engagement, adaptive pacing, and answering spontaneous questions that current systems cannot reliably handle end-to-end. Preparation is automatable but delivery remains fundamentally human.
Task automatabilityclaude-sonnet-52/5AI can help draft lecture content and slides, but live delivery, adapting to student questions, and embodying instructor presence in a classroom remain largely human tasks not meeting the 50% time-saving bar end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Universities have strong accreditation and pedagogical expectations that instruction be delivered by qualified faculty; student outcomes and institutional reputation depend on perceived quality of direct instruction. Institutional and regulatory inertia strongly protect this role.
Adoption barriersclaude-sonnet-53/5No formal licensing barrier prevents AI-assisted content creation, but institutional norms, accreditation expectations, and student/faculty preference for human instructors create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of developing and maintaining a fully AI-driven lecture system, including content curation, error-checking, and persistent student interaction, likely exceeds the loaded cost of a tenured or adjunct instructor per course, especially when factoring in liability and student support.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply generate content drafts, but full lecture preparation plus delivery still requires substantial human oversight and in-person presence, keeping overall costs comparable to or only modestly below human cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs full lecture delivery (preparation + live instruction). AI can generate lecture notes and slides with material errors, but production systems do not yet substitute for the instructor role in higher education at scale.
Technical feasibility todayclaude-sonnet-52/5Tools like ChatGPT and slide generators assist with lecture prep, but no deployed product reliably delivers full postsecondary lectures autonomously in production settings.

Initiate, facilitate, and moderate classroom discussions.

23

CI 20–25 · exposure 16 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Postsecondary education is a traditionally change-resistant sector; while some institutions pilot hybrid or asynchronous discussions, live classroom moderation by humans remains the institutional norm with slow movement toward AI substitution.
Sector adoption velocityclaude-sonnet-52/5Higher education is adopting AI for content creation and grading support but adoption of AI as a live classroom discussion moderator is essentially nonexistent, reflecting the sector's slower pace on interactive teaching automation.
Augmentation potentialclaude-haiku-4-5-202510014/5AI assistants can meaningfully augment instructors by preparing discussion notes, suggesting follow-up questions, analyzing participation patterns post-hoc, and drafting summaries, substantially raising instructor productivity without removing the human from the facilitation role.
Augmentation potentialclaude-sonnet-53/5AI can help instructors prepare discussion questions, summarize readings, and suggest talking points, providing moderate productivity gains in preparing for and following up on discussions.
Task automatabilityclaude-haiku-4-5-202510012/5AI can generate discussion prompts and provide topical summaries, but cannot authentically replicate the real-time judgment, emotional attunement, and adaptive facilitation required to moderate a classroom discussion where participation dynamics and educational goals shift moment-to-moment.
Task automatabilityclaude-sonnet-52/5AI chatbots can prompt discussion questions but live, real-time facilitation of a dynamic classroom of students requires reading social cues, adapting to unexpected tangents, and managing group dynamics that current AI cannot reliably replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Educational institutions have strong norms and often contractual requirements for instructor-led classroom engagement; accreditation bodies and student expectations reinforce direct human facilitation as a core component of postsecondary education.
Adoption barriersclaude-sonnet-53/5No formal licensing barrier prevents AI involvement, but the expectation of live human presence, classroom management, and pedagogical judgment creates strong organizational and institutional friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5Oversight of AI-moderated discussions would require human review anyway, making the all-in cost (inference + integration + human monitoring) comparable to or higher than direct human instruction.
Cost vs. human wageclaude-sonnet-52/5Since AI cannot substitute for the core in-person facilitation, any AI use is supplementary; the human instructor's cost remains largely unchanged, so cost savings from AI are minimal for this specific task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably moderates live classroom discussions end-to-end. Chatbots can simulate discussion but lack the situated awareness, authority, and pedagogical judgment that live facilitation demands.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously runs live postsecondary classroom discussions; existing AI tools are limited to generating discussion prompts or asynchronous chat-based Q&A, not facilitating in-person group dialogue.

Maintain or repair lab equipment.

18

CI 10–26 · exposure 8 · augmentation 38 · importance 3.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI maintenance diagnostics in postsecondary labs is limited; most institutions rely on vendor service contracts or trained technical staff. Equipment repair remains relatively low-tech and fragmented across institutions, with slow digital transformation compared to information-intensive sectors.
Sector adoption velocityclaude-sonnet-51/5Physical maintenance tasks in academic lab settings show minimal AI adoption; this is a low-digitization, hands-on task category.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by providing diagnostic guidance, procedural documentation, parts lists, and vendor contact automation, reducing troubleshooting time and enabling technicians to work more efficiently. However, it does not fundamentally transform the physical task, leaving the human operator as the primary decision-maker and executor.
Augmentation potentialclaude-sonnet-52/5AI chatbots can offer troubleshooting tips or manuals lookup, providing minor assistance, but cannot meaningfully transform the physical repair process.
Task automatabilityclaude-haiku-4-5-202510012/5Diagnosing and repairing lab equipment requires hands-on physical manipulation, contextual troubleshooting, and real-time environmental assessment that current AI cannot perform autonomously. While AI could assist in diagnostic lookup or documentation, the core repair work—replacing parts, recalibrating instruments, testing functionality—remains fundamentally manual and context-dependent.
Task automatabilityclaude-sonnet-51/5Physical repair and maintenance of laboratory equipment (microscopes, centrifuges, incubators, etc.) requires hands-on manipulation and diagnosis that current AI cannot perform end-to-end without robotics.'
Adoption barriersclaude-haiku-4-5-202510013/5Lab equipment repair in educational settings often requires institutional knowledge, safety certification, and vendor-specific training; however, there are no hard legal bars to AI-assisted diagnostics or documentation. Safety concerns and liability for equipment damage create practical friction rather than formal regulatory barriers.
Adoption barriersclaude-sonnet-52/5No licensing requirement for lab equipment repair specifically, but physical presence and hands-on skill are inherent requirements that block software-only automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI support tools (diagnostic software, documentation) cost significantly less than a technician's labor per repair instance, but actual repair execution still requires skilled human technicians, making the total cost comparable to or exceeding human-only workflows when integration overhead is included.
Cost vs. human wageclaude-sonnet-51/5AI has no capability to substitute for the physical labor involved, so the human technician remains the only viable and thus cheaper option in practice.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system today reliably diagnoses and repairs lab equipment end-to-end. While computer vision can identify equipment and LLMs can provide procedural guidance, no production system physically executes repairs or validates equipment function across the diversity of biological lab instruments.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously maintains or repairs postsecondary science lab equipment; at best AI could provide troubleshooting text guidance, not actual repair.

Maintain regularly scheduled office hours to advise and assist students.

8

CI 0–16 · exposure 0 · augmentation 50 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education remains a traditionally human-centered sector with strong institutional and cultural resistance to removing faculty–student contact; adoption of AI for core advising functions remains negligible in production.
Sector adoption velocityclaude-sonnet-52/5Higher education is adopting AI chatbots for advising support and FAQs, but actual replacement of faculty office hours remains rare and mostly experimental.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist faculty by helping draft advising notes, suggest relevant resources, or organize student records, thereby reducing administrative load and allowing faculty to focus more on the relational aspects of office hours.
Augmentation potentialclaude-sonnet-53/5AI can help draft responses to common student questions, provide supplementary explanations, or triage inquiries, but the core advising interaction still relies on the instructor.
Task automatabilityclaude-haiku-4-5-202510011/5This task fundamentally requires synchronous human interaction, relationship-building, and nuanced advising tailored to individual student circumstances—capacities that current AI systems cannot reliably replicate in a scheduled office-hours format. While AI could help draft responses or suggest resources, it cannot serve as the required human advisor.
Task automatabilityclaude-sonnet-51/5This task fundamentally requires a human presence and relationship-based mentorship, availability, and personalized academic/career guidance that current AI cannot substitute for at equal quality.dinger
Adoption barriersclaude-haiku-4-5-202510015/5Educational institutions have explicit policies and accreditation standards requiring faculty-student contact and advising; there are also implicit regulatory and contractual expectations that faculty provide direct mentorship, creating hard barriers to automation.
Adoption barriersclaude-sonnet-53/5No licensing requirement for office hours specifically, but institutional norms, accreditation expectations, and student/faculty relationship expectations create meaningful friction against substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5Replacing faculty office hours with AI would incur significant system costs while also requiring human oversight; the all-in cost of a reliable AI system plus monitoring would exceed the opportunity cost of the faculty member's time.
Cost vs. human wageclaude-sonnet-52/5AI chat tools are cheap per interaction, but they cannot replace the value of the human advising role, so the effective cost comparison for equal-quality output favors the human.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can substitute for a faculty member maintaining regular office hours; the task is inherently a human institutional commitment involving personalized guidance, mentorship, and the authority to advise on academic matters.
Technical feasibility todayclaude-sonnet-51/5No deployed product replaces an instructor's office hours; chatbots exist for FAQs but not for the personal advising and relationship-building this task implies.

Provide professional consulting services to government or industry.

8

CI 5–11 · exposure 0 · augmentation 63 · importance 2.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Professional consulting remains deeply rooted in human expertise and client relationships; adoption of AI for independent consulting delivery is minimal. While some firms experiment with AI-assisted analysis, end-to-end consulting substitution has not occurred in measurable ways.
Sector adoption velocityclaude-sonnet-52/5Academic and consulting sectors are adopting AI tools for research support and writing, but full consulting engagements remain slow to shift to AI-driven delivery.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist consultants by synthesizing literature, modeling scenarios, organizing data, and drafting reports, raising productivity on analytical portions of the work. However, the human consultant remains essential for strategic judgment, client interaction, and accountability.
Augmentation potentialclaude-sonnet-54/5AI can significantly assist with literature review, data analysis, report drafting, and background research, meaningfully boosting the productivity of a human consultant.
Task automatabilityclaude-haiku-4-5-202510011/5Consulting services require deep domain expertise, client relationship management, strategic judgment, and stakeholder communication—all heavily dependent on human credibility and contextual understanding. Current AI systems cannot reliably replace the expert judgment, negotiation, and accountability required.
Task automatabilityclaude-sonnet-51/5Expert consulting on biological science matters requires deep domain expertise, judgment, credibility, and accountability that current AI cannot autonomously provide end-to-end to clients as a substitute for a human expert consultant.
Adoption barriersclaude-haiku-4-5-202510014/5Government and industry clients require consultants to be accountable subject-matter experts, often with credentials, liability insurance, and contractual responsibility for advice given. Professional licensing, client trust, and legal liability create substantial adoption barriers against full automation.
Adoption barriersclaude-sonnet-54/5Consulting engagements often require credentialed expertise, professional reputation, accountability for advice given, and sometimes formal certification or liability coverage, creating strong barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5A consulting engagement commands high loaded hourly rates ($150–500+/hour for postdoctoral-level expertise), reflecting liability, accountability, and specialized knowledge. Current AI inference is cheap, but integrating it into a credible consulting workflow adds minimal value and does not reduce overall cost per engagement delivered.
Cost vs. human wageclaude-sonnet-52/5While AI can cut research/drafting time, the deliverable (expert judgment, credibility, liability) still requires a paid human expert, so overall cost savings are modest, not order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs independent professional consulting in biology or related fields. While AI can assist with research synthesis or analysis, actual consulting delivery—advising government or industry clients on complex decisions—remains fundamentally human-performed.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently performs professional scientific consulting services to government or industry; this remains a human-expert-delivered service, sometimes AI-assisted.

Supervise students' laboratory work.

6

CI 0–13 · exposure 5 · augmentation 38 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions are slow to adopt unproven laboratory automation systems, and the legal and safety stakes make experimental AI substitution infeasible. No measurable production deployment of AI lab supervision exists in higher education.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI for content delivery and grading but physical lab supervision has seen little to no automation adoption.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist instructors with grading pre-lab quizzes, documenting results, or flagging unusual sensor readings, but core supervision—safety monitoring, feedback on technique, and real-time intervention—remains fundamentally human-centric work.
Augmentation potentialclaude-sonnet-53/5AI can assist with pre-lab instructions, safety checklists, lab report review, and answering procedural questions, but cannot replace in-person physical oversight.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising laboratory work requires real-time physical presence, intervention, and safety judgment that current AI cannot provide. AI systems cannot be physically present in labs, monitor hands-on techniques, or intervene to prevent accidents or equipment misuse.
Task automatabilityclaude-sonnet-51/5Supervising hands-on lab work requires physical presence, real-time safety monitoring, and hands-on correction of technique that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Safety regulations, institutional liability, accreditation standards, and duty of care legally require a qualified human instructor to physically supervise postsecondary laboratory work. Automation would likely violate regulations and expose institutions to unacceptable liability.
Adoption barriersclaude-sonnet-54/5Safety regulations, liability for lab accidents, and institutional requirements for qualified personnel to oversee student lab work create strong barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI supervision systems would require expensive camera infrastructure, computer vision models, and continuous monitoring, making the total cost exceed the loaded wage of part-time or adjunct teaching lab instructors who typically perform this work.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing the physical supervisory role, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI could assist with some documentation or post-hoc analysis of lab work, no deployed product reliably supervises active laboratory activity. Remote monitoring systems exist but cannot replace human oversight of safety, technique, and real-time troubleshooting that postsecondary lab supervision demands.
Technical feasibility todayclaude-sonnet-51/5No deployed product provides in-person supervision of students handling lab equipment, chemicals, or biological specimens; this remains research-stage at best for narrow simulation contexts.

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

6

CI 0–13 · exposure 5 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic institutions move slowly on automation of core instructional and supervisory roles due to professional norms, accreditation requirements, and faculty governance. Even with AI tools available, adoption remains limited to supplementary administrative tasks rather than replacing supervision itself.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools slowly for core mentorship and supervisory functions, though pilots for administrative aspects exist.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist faculty supervision through automated scheduling, progress dashboards, literature summaries, and initial draft feedback on student work, reducing administrative overhead. However, augmentation is limited to clerical and routine analytic tasks; the core mentoring and decision-making remain solidly human.
Augmentation potentialclaude-sonnet-53/5AI can help track student progress, draft feedback, suggest research resources, or grade drafts, but the core supervisory relationship remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Supervision requires judgment about student progress, mentoring relationships, research direction, and individual development—complex human interactions where AI cannot replace the core pedagogical and advisory function. Current AI systems cannot autonomously oversee research integrity, provide meaningful feedback tailored to individual growth, or make evaluative decisions about research quality and student readiness.
Task automatabilityclaude-sonnet-51/5Supervising students' research and teaching requires mentorship, judgment about individual progress, and relationship-based guidance that current AI cannot autonomously perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Educational accreditation, institutional policy, and ethical standards require that qualified faculty directly supervise research and teaching work. Many jurisdictions and funding agencies legally mandate faculty oversight of research integrity and student evaluation, creating hard regulatory and contractual barriers to automation.
Adoption barriersclaude-sonnet-54/5Institutional accreditation, faculty credentialing requirements, and formal advisor-of-record responsibilities create strong structural barriers to AI substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5Supervision is inherently a high-touch, relationship-intensive role where the faculty member's expertise and presence are the primary value. AI tools for administrative support are cheap but cannot replace the core supervisory work, making the all-in cost of AI substitution higher than maintaining human supervision.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI replacement for this supervisory role, so cost comparison favors the human faculty member entirely.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist with scheduling, progress tracking, and documentation, no deployed system reliably performs end-to-end supervision—which requires real-time responsiveness to student needs, ethical oversight of research conduct, and dynamic mentoring. Products exist for administrative elements (grading support, scheduling) but not for the supervisory relationship itself.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises student research or internships; this remains a human academic mentorship role with no production substitute.

Collaborate with colleagues to address teaching and research issues.

6

CI 5–7 · exposure 0 · augmentation 38 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic institutions are among the slowest sectors to automate professional judgment tasks, and there is no observable adoption of AI as a substitute for faculty collaboration on teaching and research issues.
Sector adoption velocityclaude-sonnet-52/5Higher education is a slow-adopting sector for AI in core collegial and governance functions, though some tools are used for scheduling or document sharing.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can offer modest assistance in organizing information or summarizing research for discussion, but the actual collaborative deliberation and relationship-building that defines this task remains minimally augmented by current systems.
Augmentation potentialclaude-sonnet-53/5AI can help draft shared documents, summarize meeting notes, or organize research collaboration logistics, offering moderate assistance despite the core interpersonal task remaining human.
Task automatabilityclaude-haiku-4-5-202510011/5Collaboration on teaching and research issues requires nuanced human judgment, relationship building, and contextual understanding of disciplinary challenges that current AI cannot replicate end-to-end. While AI can assist with information synthesis, the core collaborative negotiation and decision-making remains firmly human.
Task automatabilityclaude-sonnet-51/5Collaboration on teaching and research issues requires interpersonal negotiation, shared judgment, and relationship-building among colleagues that current AI cannot substitute for end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: academic institutions expect faculty to participate directly in collegial governance, decisions carry implicit professional and fiduciary responsibilities, and faculty culture prioritizes human-to-human collegial exchange as irreplaceable.
Adoption barriersclaude-sonnet-54/5Faculty governance, tenure processes, and departmental norms require actual human participation and accountability in collaborative academic decisions.
Cost vs. human wageclaude-haiku-4-5-202510011/5Meaningful engagement in collaborative problem-solving by AI would require significant human oversight and integration costs that would exceed the value of the contribution compared to direct human collaboration.
Cost vs. human wageclaude-sonnet-51/5There is no AI product substituting for this task, so cost comparison favors the human entirely.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can autonomously conduct meaningful collaboration with colleagues on complex institutional and research matters; such systems exist only in research prototypes at best.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs collegial collaboration on academic issues; AI tools at best facilitate communication but do not replace the collaborative act itself.

Perform administrative duties, such as serving as department head.

6

CI 0–11 · exposure 0 · augmentation 38 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic institutions operate with deeply embedded governance structures and human-led hierarchies. Adoption of AI for department leadership functions is negligible; universities continue to appoint qualified humans to these roles with minimal automation.
Sector adoption velocityclaude-sonnet-52/5Higher education administration adopts AI slowly for such leadership functions, though AI tools for scheduling and communications are increasingly used.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with specific administrative tasks like scheduling course listings, analyzing enrollment data, or drafting routine communications, but the leadership, judgment, and accountability dimensions of a department head role remain largely human-driven with limited augmentation potential.
Augmentation potentialclaude-sonnet-53/5AI can help draft memos, summarize meetings, manage schedules, and analyze budget data, meaningfully assisting a department head's administrative workload.
Task automatabilityclaude-haiku-4-5-202510011/5Serving as department head involves strategic decision-making, personnel management, budgeting oversight, and institutional advocacy that require human judgment and organizational authority. Current AI cannot perform these duties end-to-end or achieve meaningful time savings at equal quality.
Task automatabilityclaude-sonnet-51/5Departmental leadership involves interpersonal management, negotiation, personnel decisions, and institutional politics that current AI cannot execute end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Department head responsibilities are legally and organizationally vested in a named human with fiduciary and hiring authority. Institutional bylaws, accreditation standards, and employment law require a human leader to sign off on hiring, budgets, and curricular decisions.
Adoption barriersclaude-sonnet-54/5Department head roles require institutional appointment, tenure status, faculty governance approval, and accountability that cannot be delegated to software.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI has no cost advantage here; the task is fundamentally human-centric administrative and leadership work that cannot be substituted by inference. The cost of any partial AI assistance (e.g., data organization) is negligible relative to the salaried department head role.
Cost vs. human wageclaude-sonnet-52/5Since AI cannot substitute for the role itself, any cost comparison is moot except for minor administrative subtasks where AI may be cheap but only marginally offsets overall cost.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI system reliably performs department head duties in production. While AI can assist with scheduling or data summarization, the core responsibilities—hiring decisions, conflict resolution, budget advocacy, curriculum direction—remain outside the scope of current autonomous systems.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the role of department head; AI is at most used for scheduling or drafting support within such a role.

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

4

CI 0–9 · exposure 5 · augmentation 50 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Committee service is a core institutional governance function in higher education, and there is zero adoption of AI autonomous participation because legal accountability and faculty authority cannot be delegated. Adoption is structurally absent.
Sector adoption velocityclaude-sonnet-51/5Higher education governance and committee work show minimal AI adoption; this is a slow-moving, relationship-based institutional process.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist human committee members by analyzing policy proposals, preparing comparative data, or summarizing prior decisions, meaningfully reducing preparation burden. However, the core deliberative and voting functions remain with humans.
Augmentation potentialclaude-sonnet-53/5AI can help draft meeting summaries, policy language, or background research to support a committee member's contributions, offering moderate assistance.
Task automatabilityclaude-haiku-4-5-202510011/5Committee work requires judgment on institutional policy, stakeholder negotiation, and consensus-building across competing interests—tasks fundamentally dependent on human authority and accountability. Current AI cannot autonomously participate in decision-making that carries legal or fiduciary responsibility.
Task automatabilityclaude-sonnet-51/5Committee service requires interpersonal deliberation, political judgment, and institutional relationship-building that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Academic institutions legally require human faculty participation in governance and policy decisions; university bylaws and accreditation standards mandate that committees include faculty members with institutional authority. These are hard, regulatory barriers.
Adoption barriersclaude-sonnet-55/5Committee membership requires institutional standing, faculty status, and often elected or appointed authority tied to governance structures, making substitution essentially impossible.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools can reduce preparation time and administrative overhead, but the economic gain is modest since committee service is typically part of faculty duty rather than a standalone billable task. Savings are limited to documentation and analysis, not the service itself.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI alternative delivering equivalent output, so the cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can draft meeting agendas, summarize documents, or prepare analysis for human committee members, no deployed system autonomously serves on committees or makes institutional decisions. Some tools assist with preparation, but the core task—deliberating and voting—remains human-exclusive.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for a faculty member's participation in governance committees; this remains entirely human-driven.

Provide students course-related experiences, such as field trips, outside the classroom.

4

CI 0–7 · exposure 0 · augmentation 50 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Field trip organization remains a fundamentally human, hands-on activity in educational institutions with limited digitization of core duties. Adoption of AI for logistics assistance is minimal; the core task itself is not a target for automation in practice.
Sector adoption velocityclaude-sonnet-52/5Higher education administrative and pedagogical adoption of AI is growing, but for physical field experiences specifically there is little to no AI penetration.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist moderately by helping identify field sites, managing logistics (booking, itineraries, cost estimates), and suggesting learning objectives aligned with venues. However, the core pedagogical and supervisory elements remain human-directed, limiting augmentation scope.
Augmentation potentialclaude-sonnet-53/5AI can help with planning logistics, drafting itineraries, safety checklists, or generating pre/post-trip educational materials, meaningfully aiding preparation even though it can't replace the in-person experience.
Task automatabilityclaude-haiku-4-5-202510011/5This task is fundamentally about organizing and facilitating real-world educational experiences for students outside the classroom. AI cannot autonomously arrange transportation, coordinate field sites, supervise students, or provide the physical presence and pedagogical guidance required; no current system can replace these human functions.
Task automatabilityclaude-sonnet-51/5Organizing and leading physical field trips requires real-world logistics, transportation, permissions, and in-person supervision that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Strong legal and institutional barriers protect this task: instructors have fiduciary duty for student safety, duty of care requirements, institutional liability, and educational regulations mandate human supervision. No automation can legally replace the instructor's required presence and responsibility.
Adoption barriersclaude-sonnet-54/5Liability, safety, and institutional supervision requirements (e.g., faculty responsibility for student safety off-campus) create strong barriers to any non-human substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of AI tools to assist with logistics and planning is minimal, but the task itself—requiring human time for coordination, travel, and student supervision—means the all-in cost of automation remains higher than direct human labor for delivering the same educational outcome.
Cost vs. human wageclaude-sonnet-51/5AI cannot substitute for the human labor of chaperoning and conducting field experiences, so there is no meaningful cost comparison—the human cost is unavoidable.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs the coordination, logistics, and on-site facilitation of field trips. While AI can assist with planning (e.g., suggesting venues), it cannot execute the task end-to-end or substitute for the human instructor's presence and judgment.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product plans, supervises, or conducts field trips; this remains an inherently physical, human-led activity.

Act as advisers to student organizations.

4

CI 0–7 · exposure 0 · augmentation 25 · importance 2.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education has shown minimal adoption of AI for student advising roles; institutions continue to rely on faculty and dedicated staff, reflecting the low digitization and human-contact requirement of this work.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI for administrative and instructional support but advising/mentorship roles for student organizations see minimal AI penetration.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with administrative documentation or scheduling coordination for student organizations, but meaningful advising—addressing leadership conflicts, guiding decision-making, building mentor relationships—remains primarily a human function with only limited support possibilities.
Augmentation potentialclaude-sonnet-52/5AI can help with scheduling, drafting communications, or budget tracking for the organization, but it only marginally assists the core advisory relationship.
Task automatabilityclaude-haiku-4-5-202510011/5Advising student organizations requires understanding group dynamics, mentoring relationships, conflict resolution, and personalized guidance—tasks fundamentally dependent on human judgment, emotional intelligence, and sustained interpersonal engagement that current AI cannot replicate end-to-end.
Task automatabilityclaude-sonnet-51/5Advising student organizations requires ongoing relationship-building, mentorship, event oversight, and judgment calls that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Universities require faculty advisers to sign off on student organization decisions, policies, and governance; there is both a legal and institutional requirement that a human advisor be accountable and present.
Adoption barriersclaude-sonnet-54/5Many institutions require a designated faculty/staff advisor for liability, safety, and institutional policy reasons, creating a formal human-presence requirement.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI system to handle student advising (integration, training, oversight by faculty) would exceed the incremental cost of having faculty spend time on this task as part of their role.
Cost vs. human wageclaude-sonnet-51/5There is no AI service replacing this role, so cost comparison favors the human by default since AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs the advisory role for student organizations reliably today; this requires live interaction, accountability, and trust relationships that are beyond current AI capabilities.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product acts as a substitute faculty advisor for student clubs; this remains an in-person, relational role.

Participate in campus and community events, such as giving presentations to the public.

3

CI 0–5 · exposure 0 · augmentation 50 · importance 2.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions have shown minimal adoption of AI for live public-facing presentations; this task remains firmly a human-performed function in postsecondary settings.
Sector adoption velocityclaude-sonnet-51/5Higher education community outreach is a low-digitization, in-person activity with minimal AI adoption pressure.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with preparation—drafting talking points, generating visuals, or organizing slides—but the actual presentation delivery remains human-centered, so augmentation is limited to pre-event support.
Augmentation potentialclaude-sonnet-53/5AI can help prepare slides, talking points, or promotional materials for the presentation, offering moderate productivity support ahead of the event.
Task automatabilityclaude-haiku-4-5-202510011/5Public presentations require real-time audience engagement, dynamic interaction, and adaptive communication—capabilities that current AI systems cannot reliably perform end-to-end. While AI can draft slides or notes, the live performance, audience reading, and embodied presence remain distinctly human.
Task automatabilityclaude-sonnet-51/5Physical presence, live audience interaction, and personal representation of the institution cannot be performed end-to-end by AI systems today.
Adoption barriersclaude-haiku-4-5-202510015/5Campus and community events typically require an authorized human representative (the faculty member) to speak on behalf of the institution; institutional policy, liability concerns, and community expectations all mandate human presence and accountability.
Adoption barriersclaude-sonnet-54/5Institutional representation, community trust-building, and personal accountability create strong organizational and reputational barriers to substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI system capable of generating and delivering public presentations autonomously, plus the overhead of integration and monitoring, far exceeds the cost of a faculty member giving presentations as part of their regular role.
Cost vs. human wageclaude-sonnet-51/5There is no AI equivalent output for this in-person representational task, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably substitutes for a human presenter giving campus and community event presentations. Text-to-speech and slide generation tools exist, but they cannot replicate the credibility, authority, and interactive presence required in educational public engagement.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for a person physically attending and representing an institution at community events.

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.