Environmental Science Teachers, Postsecondary

25-1053.00
Median wage $94,980/yr6,690 employed (US)Rank #359 of 923 scored · top 39% by substitution

Teach courses in environmental science. 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
Exposure29
Augmentation61

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

25 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

12%

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

Why this score

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

Task automatabilityw 35%28

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

Technical feasibility todayw 20%30

panel mean rating 2.2/5 → substitution pressure 30/100

Cost vs. human wagew 15%40

panel mean rating 2.6/5 → substitution pressure 40/100

Adoption barriersw 20%inverted — strong barriers lower the score34

panel mean rating 3.7/5 (barrier strength) → substitution pressure 34/100

Sector adoption velocityw 10%26

panel mean rating 2.1/5 → substitution pressure 26/100

Task breakdown (25 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 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Postsecondary institutions are information-sector, highly digitized organizations; LMS and SIS adoption is near-universal and automated attendance/grade handling is standard, not emerging.
Sector adoption velocityclaude-sonnet-55/5Higher education has near-universal adoption of digital gradebooks and attendance systems; this is one of the most thoroughly digitized administrative tasks in academia.
Augmentation potentialclaude-haiku-4-5-202510014/5Beyond full automation, AI can assist instructors by flagging at-risk students based on attendance patterns, offering grade analytics dashboards, and automating roster reconciliation, enhancing human oversight while reducing manual load.
Augmentation potentialclaude-sonnet-54/5AI-enhanced systems help instructors flag at-risk students, auto-calculate grades, and generate reports, meaningfully boosting efficiency while the instructor retains oversight of final grades.
Task automatabilityclaude-haiku-4-5-202510015/5Student record maintenance (attendance, grades, etc.) is purely data entry and management work that can be fully automated via learning management systems and administrative software; no judgment or human interaction required and time savings far exceed 50%.
Task automatabilityclaude-sonnet-55/5Attendance and grade recordkeeping is a structured, rules-based data entry and tracking task that off-the-shelf LMS/SIS systems with AI features can fully automate today.
Adoption barriersclaude-haiku-4-5-202510012/5Most institutions mandate use of specific systems (Canvas, Banner, etc.), and IT/registrar departments control system access and data standards; however, automation within existing systems faces minimal legal barriers and integration is standard practice.
Adoption barriersclaude-sonnet-52/5Some institutional policy and FERPA-related data handling requirements exist, but no licensing requirement mandates a human personally maintain these records, so barriers are low to moderate.
Cost vs. human wageclaude-haiku-4-5-202510015/5Institutional LMS/SIS infrastructure is leveraged across thousands of students and records; the per-task marginal cost of automating one teacher's records is negligible compared to the instructor labor cost.
Cost vs. human wageclaude-sonnet-55/5Automated recordkeeping software costs a small fraction of the faculty time it would take to manually maintain these records, representing an order-of-magnitude cost reduction.
Technical feasibility todayclaude-haiku-4-5-202510015/5Learning management systems (Canvas, Blackboard, Banner, etc.) and student information systems are mature, deployed at scale in virtually all postsecondary institutions, and reliably handle attendance tracking, grade recording, and record-keeping.
Technical feasibility todayclaude-sonnet-55/5Mature learning management systems (Canvas, Blackboard, PowerSchool, etc.) already automate attendance tracking, gradebook calculations, and record retention at scale in production across universities.

Compile bibliographies of specialized materials for outside reading assignments.

79

CI 76–81 · exposure 75 · augmentation 88 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Academic institutions are adopting AI writing and research tools at middling pace—pilots are common in progressive departments, but widespread systematic use of AI-generated bibliographies remains limited due to quality control concerns and traditional pedagogical preferences.
Sector adoption velocityclaude-sonnet-53/5Higher education is moderately adopting AI research tools, but many faculty still compile bibliographies manually or with traditional library tools rather than AI systems.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists instructors by rapidly generating initial drafts, discovering relevant papers, and formatting citations, allowing faculty to focus on selecting, ordering, and annotating materials for pedagogical alignment rather than mechanical search and formatting.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up literature discovery and citation formatting, letting instructors quickly build and refine curated reading lists while retaining final judgment on relevance.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems (ChatGPT, Claude, specialized academic tools) can efficiently compile bibliographies from broad topics, format citations, and suggest relevant specialized materials with minimal human intervention, achieving substantial time savings over manual research and formatting.
Task automatabilityclaude-sonnet-54/5AI tools can search literature databases, generate topical reading lists, and format citations with minimal human input, meeting the time-saving threshold for most of this task.dim
Adoption barriersclaude-haiku-4-5-202510012/5Minimal legal or regulatory barriers exist; the main friction is faculty preference for human curation and institutional review processes, but nothing legally requires human compilation of reading lists.
Adoption barriersclaude-sonnet-51/5No licensing or regulatory requirement mandates a human compile reading lists; it's a low-stakes administrative/academic task.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-assisted bibliography compilation costs pennies per assignment via API calls or subscriptions, versus hours of human librarian or faculty time at $30–80/hour loaded cost, representing orders-of-magnitude savings.
Cost vs. human wageclaude-sonnet-55/5AI-assisted bibliography compilation costs pennies compared to faculty hourly wages spent manually searching and curating references.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple deployed products (Zotero with AI plugins, ChatGPT, dedicated academic research assistants, and citation managers) reliably generate formatted bibliographies at scale, though instructors typically review for topical accuracy and relevance to course goals.
Technical feasibility todayclaude-sonnet-54/5Deployed tools like reference managers, AI research assistants (e.g., Elicit, Semantic Scholar, ChatGPT with browsing) reliably compile bibliographies today, though occasional inaccuracies or hallucinated citations require verification.

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

74

CI 71–76 · exposure 75 · augmentation 100 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Higher education has begun piloting AI for content generation, but adoption remains uneven; many institutions are cautious about AI-generated pedagogy, slowing production-scale deployment.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI tools for content creation at a moderate pace, with growing faculty use of generative AI for prep but institutional caution, policy debates, and uneven digitization slow deeper adoption.
Augmentation potentialclaude-haiku-4-5-202510015/5AI drafting tools significantly augment instructor productivity by generating initial templates, brainstorming assignment ideas, and providing format consistency, while faculty maintain full creative and pedagogical control.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up drafting of syllabi, assignments, and handouts, letting instructors focus on refining content, ensuring accuracy, and tailoring materials to their course.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems (LLMs, content generators) can produce high-quality syllabi, assignments, and handouts at 50%+ time savings compared to writing from scratch, though review and customization remain necessary.
Task automatabilityclaude-sonnet-54/5LLMs can generate syllabi, homework problems, and handouts from a course outline or textbook with substantial time savings, though instructor review and customization are still needed for accuracy and alignment with learning objectives.
Adoption barriersclaude-haiku-4-5-202510013/5While content generation is not legally restricted, institutional policies, accreditation expectations, and faculty autonomy over curricula create moderate friction to full automation.
Adoption barriersclaude-sonnet-52/5No licensing or legal requirement mandates human-only authorship of syllabi or assignments, though institutional policies, accreditation standards, and academic integrity norms create some review friction.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI inference costs for generating course materials are orders of magnitude cheaper than faculty labor time; a few dollars of compute replaces hours of skilled preparation work.
Cost vs. human wageclaude-sonnet-55/5Generating draft course materials via AI costs cents to dollars in inference compared to hours of faculty or TA time, an order-of-magnitude cost advantage even after review time.
Technical feasibility todayclaude-haiku-4-5-202510014/5Mature AI writing tools are widely deployed and reliably generate pedagogical content; however, institutional variability and subject-specific nuance require human oversight, limiting full autonomous production.
Technical feasibility todayclaude-sonnet-54/5Widely deployed tools like ChatGPT, Claude, and specialized ed-tech products (e.g., Coursera, LMS-integrated AI assistants) are already used by instructors to draft syllabi and assignments reliably, though outputs typically need editing.

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

62

CI 51–72 · exposure 62 · augmentation 75 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Higher education has moved toward LMS-integrated assessment tools and automated grading, but adoption varies widely by institution type and discipline. Pilots are common and growing, but full production deployment of AI-driven exam compilation and grading without human review remains inconsistent across sectors.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI grading tools unevenly and cautiously, with pilots more common than full production deployment, reflecting the sector's generally slower digitization of assessment practices.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists instructors by auto-generating candidate exam questions, producing grading rubrics, and handling the mechanical scoring of large classes, freeing faculty time for review and feedback refinement. This augmentation is widely observed in practice and measurably raises instructor productivity on assessment tasks.
Augmentation potentialclaude-sonnet-54/5AI tools substantially help instructors draft exam questions, create rubrics, and pre-grade objective and some subjective responses, meaningfully speeding up the overall process while instructors retain final control.
Task automatabilityclaude-haiku-4-5-202510014/5AI can compile exam questions from learning materials, generate grading rubrics, and automatically score objective assessments and some subjective work with high accuracy. However, the human oversight and final judgment on grading decisions typically remain necessary, preventing a full end-to-end automation with >50% time savings across all exam types without significant setup.
Task automatabilityclaude-sonnet-53/5AI can generate exam questions and grade objective or even short-answer responses reasonably well, but compiling exams aligned to specific course content and grading nuanced scientific reasoning still needs faculty oversight, limiting full end-to-end automation.
Adoption barriersclaude-haiku-4-5-202510013/5Educational institutions often require faculty oversight and sign-off on grading, and there is organizational preference for human judgment on subjective assessment. However, no hard legal barrier prevents delegation to AI, and many schools already use automated grading systems, creating only moderate friction.
Adoption barriersclaude-sonnet-52/5No licensing requirement mandates human grading, though institutional academic integrity policies and grade appeal processes create moderate friction and typically require instructor sign-off on final grades.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI grading and exam compilation cost a fraction of instructor time spent on these routine tasks; integrated LMS solutions have low marginal cost per assessment. Full replacement cost (inference + oversight) is substantially cheaper than paying instructor labor for routine exam administration and grading.
Cost vs. human wageclaude-sonnet-54/5AI-based question generation and grading of multiple-choice or short-answer items is far cheaper per unit than faculty or TA time, though complex grading may still require human review, moderating the savings somewhat.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (learning management systems with AI-powered assessment tools, automated grading platforms) reliably handle objective question generation and scoring at scale in academic institutions today. Subjective answer grading remains less mature but increasingly functional in production systems.
Technical feasibility todayclaude-sonnet-53/5Products like AI-assisted quiz generators and automated grading tools (e.g., in LMS platforms) are deployed in production, but reliability for open-ended science exams with reasoning/derivations remains inconsistent.

Participate in student recruitment, registration, and placement activities.

59

CI 30–87 · exposure 58 · augmentation 63 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Higher education increasingly deploys CRM systems, automated advising chatbots, and algorithmic course matching; many institutions have active production pilots for recruitment and placement automation in professional services and information sectors.
Sector adoption velocityclaude-sonnet-52/5Higher education administrative processes are adopting AI slowly, with pilots for chatbots and CRM tools but limited integration into faculty-driven recruitment and placement tasks.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists human advisors by automating scheduling, flagging promising candidates, matching students to opportunities, and managing follow-ups, allowing advisors to focus on relationship-building and advising rather than administrative workflows.
Augmentation potentialclaude-sonnet-53/5AI can help draft recruitment materials, screen applications, and manage registration logistics, offering moderate productivity gains while faculty retain decision-making roles.
Task automatabilityclaude-haiku-4-5-202510015/5AI can fully automate outreach emails, follow-ups, scheduling, application screening, course registration workflows, and job matching/placement matching with near-equivalent quality to human coordinators, achieving >50% time savings through chatbots, CRM automation, and algorithmic matching.
Task automatabilityclaude-sonnet-52/5This task involves relational, decision-making, and administrative activities like advising and evaluating student fit that require human judgment and interpersonal interaction; AI can assist but not substitute end-to-end.
Adoption barriersclaude-haiku-4-5-202510012/5While universities may prefer human touch for some recruitment and prefer advisors review placements, there are no legal licensing requirements, liability barriers, or regulatory mandates that a human must perform recruitment, registration, or initial placement screening.
Adoption barriersclaude-sonnet-53/5No licensing requirement specifically for this activity, but institutional policies, FERPA-related privacy concerns, and expectations of faculty involvement create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510015/5Automated email, chatbot communication, application screening, and placement algorithms cost a fraction of full-time recruiter or advisor labor, easily an order of magnitude cheaper per interaction at scale.
Cost vs. human wageclaude-sonnet-52/5AI tools can reduce some administrative costs in outreach and scheduling, but the faculty time spent on personal interviews, recommendations, and placement judgment still requires paid human effort, keeping costs comparable.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple deployed products exist (CRM platforms, student recruitment chatbots, scheduling automation, placement matching algorithms) performing these tasks reliably in educational institutions, though integration and oversight still require human oversight in many contexts.
Technical feasibility todayclaude-sonnet-52/5Some CRM and chatbot tools assist with recruitment outreach and registration logistics, but no deployed product reliably handles the full scope including advising and placement decisions.

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

56

CI 54–59 · exposure 50 · augmentation 75 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Higher education has seen pilot and partial adoption of AI grading tools, particularly for objective assessments and administrative burden reduction, but widespread replacement of instructor grading on complex assignments remains rare. Adoption is faster in large institutions and slower in small liberal arts colleges.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI grading tools steadily but unevenly; many science departments remain cautious about automated grading of lab and analytical work.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments instructor productivity by pre-grading, flagging outliers, suggesting feedback, and organizing submissions, allowing instructors to focus high-touch review on complex work. This assistive model is increasingly common and demonstrably raises instructor efficiency while preserving human oversight.
Augmentation potentialclaude-sonnet-54/5AI significantly speeds up drafting feedback, flagging errors, and summarizing common mistakes across assignments, letting instructors focus on judgment-heavy evaluation.
Task automatabilityclaude-haiku-4-5-202510013/5AI can automate significant portions of grading—objective components (multiple choice, short answer matching rubrics) and formatting/organization checks are largely automatable. However, nuanced assessment of conceptual understanding, experimental design quality, and written argumentation typically requires human judgment, so full end-to-end automation at equal quality remains constrained.
Task automatabilityclaude-sonnet-53/5AI can draft grades and feedback for structured assignments and short papers, but nuanced evaluation of lab work, original research analysis, and course-specific rubrics still requires substantial human oversight to reach equal quality.4
Adoption barriersclaude-haiku-4-5-202510013/5Moderate barriers exist: institutions value human instructor judgment and feedback for pedagogical depth, accreditation bodies expect instructor verification of learning outcomes, and faculty have union/contractual protections. However, no legal requirement mandates humans alone, and many institutions already use automated systems for high-volume grading components.
Adoption barriersclaude-sonnet-52/5No licensing requirement mandates a human grade coursework, though academic integrity policies and institutional norms often require instructor accountability for final grades.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI grading tools are inexpensive (per-student subscription or one-time licensing), often under $1–5 per student per course, whereas instructor time for detailed grading of a full class easily costs $20–50+ per student. At-scale, AI is substantially cheaper for the automatable portion.
Cost vs. human wageclaude-sonnet-54/5AI grading assistance is inexpensive relative to instructor or TA time, especially for large sections, though some human review is needed, moderating the savings slightly.
Technical feasibility todayclaude-haiku-4-5-202510013/5Deployed products (learning management systems with AI-assisted grading, plagiarism detection, rubric-based auto-graders) perform narrow grading tasks reliably, but materials error rates increase on open-ended work, laboratory reports, and complex synthesis assignments. Production adoption is growing but uneven across institutions.
Technical feasibility todayclaude-sonnet-53/5AI-assisted grading tools (e.g., Gradescope, LLM-based feedback generators) are deployed in many courses, but reliability varies by discipline and instructors typically still review or override grades, especially for lab reports and open-ended science work.

Write letters of recommendation for students.

43

CI 32–54 · exposure 38 · augmentation 88 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Postsecondary education has adopted digital tools for administration but remains culturally and institutionally resistant to replacing personal judgment in recommendation writing. Adoption is minimal; most teachers still write these letters themselves, viewing them as a professional responsibility tied to their reputation.
Sector adoption velocityclaude-sonnet-53/5Academia is a knowledge-work sector with reasonably fast informal adoption of AI writing tools by individual faculty, though institutional policies on AI-assisted letters remain inconsistent and cautious.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist by organizing student data, suggesting structure, drafting initial sections, and prompting the teacher to recall specific achievements—accelerating the writing process while the teacher retains control over authenticity and accuracy. This assistance can substantially raise productivity without replacing the recommender.
Augmentation potentialclaude-sonnet-55/5AI is highly effective at drafting structure, phrasing, and tone from bullet points the professor supplies, substantially speeding up letter writing while the professor retains final control and signs off.
Task automatabilityclaude-haiku-4-5-202510012/5AI can draft template language and organize student information quickly, but writing a credible letter of recommendation requires genuine knowledge of the student's character, abilities, and achievements—often gathered through direct interaction over months or years—that cannot be meaningfully automated. Current systems cannot reliably substitute for the teacher's authentic judgment without human review and substantial revision.
Task automatabilityclaude-sonnet-53/5AI can draft a competent letter given details about the student's performance, but the task requires the teacher's genuine firsthand knowledge, judgment, and personal voice, so full automation without human review/input falls short of true equivalence.
Adoption barriersclaude-haiku-4-5-202510014/5Letters of recommendation carry legal and reputational weight; institutions and students expect authenticity from the named recommender. Many schools and programs have explicit policies that letters must be written by the named individual, and signing off on AI-generated content without personal knowledge could expose the writer to liability or ethical violation.
Adoption barriersclaude-sonnet-53/5No licensing barrier, but there's real liability/authenticity concern since recommendation letters carry the professor's signature and professional credibility, requiring genuine personal knowledge and endorsement.
Cost vs. human wageclaude-haiku-4-5-202510014/5An AI system can generate a draft letter in seconds at near-zero marginal cost, whereas a teacher's time investment in writing or revising a letter involves significant human labor. Even accounting for oversight and editing, the cost ratio favors AI substantially.
Cost vs. human wageclaude-sonnet-54/5Generating a draft letter costs pennies in compute versus the substantial time a professor would spend writing from scratch, though some human time is still needed for review and personalization.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools can generate letter drafts from prompts or structured input, no deployed product reliably produces standalone letters of recommendation that satisfy admissions committees or employers without heavy human editing. The task demands specificity, authenticity, and institutional trust that AI systems have not demonstrated in production use.
Technical feasibility todayclaude-sonnet-53/5Commercial writing assistants (ChatGPT, Grammarly, etc.) are widely used by faculty to draft recommendation letters, but reliability depends heavily on the input details provided and human editing is standard practice.

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

42

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI literature-monitoring tools is slow in academia; most postsecondary educators still rely on traditional journal subscriptions, email alerts, and in-person conference attendance. Academic conservatism and lack of institutional urgency limit velocity.
Sector adoption velocityclaude-sonnet-53/5Academia has moderate AI tool adoption for literature discovery and summarization (e.g., research assistants, citation tools), but conference attendance and collegial networks remain traditionally human-centric with slow AI integration.
Augmentation potentialclaude-haiku-4-5-202510014/5AI substantially augments this task by filtering and summarizing literature, identifying relevant papers, and alerting to emerging topics, allowing faculty to spend conference time and peer discussion more strategically. The human remains central, but AI multiplies their reach.
Augmentation potentialclaude-sonnet-54/5AI literature summarization, alerting, and search tools significantly speed up staying current with new research, meaningfully augmenting this task even though full automation isn't feasible.
Task automatabilityclaude-haiku-4-5-202510012/5AI can summarize current literature and aggregate conference abstracts, but cannot replicate the nuanced peer discussion, networking insight-gathering, and expert judgment required to truly 'keep abreast' of field developments. Reading literature is partially automatable; the synthesis and relationship-building components are not.
Task automatabilityclaude-sonnet-52/5AI can help surface and summarize literature, but the core task of genuinely staying current through synthesis, professional judgment, and networking cannot be fully offloaded to AI today with equal quality end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: professional norms and institutional expectations that faculty remain engaged in their field, conferencing is often tenure/promotion requirement, and peer relationships and serendipitous conference interactions carry organizational value that AI cannot fulfill.
Adoption barriersclaude-sonnet-52/5No licensing or legal requirement mandates human-only engagement with literature or colleagues, though academic norms around peer interaction and professional presence create some friction.
Cost vs. human wageclaude-haiku-4-5-202510014/5Literature monitoring via AI (paper summarization, alerting services) is substantially cheaper than the time cost of a postsecondary educator manually scanning journals and attending conferences, though conference attendance remains a human expense.
Cost vs. human wageclaude-sonnet-53/5AI-assisted literature review tools are cheap relative to a professor's time spent reading, but the task also includes conferences and personal networking that AI cannot substitute for at any cost.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI products can summarize research papers and curate relevant literature feeds with reasonable reliability, but no deployed system fully captures informal collegial discussion or replaces the value of conference participation. These assist but do not replace the core task.
Technical feasibility todayclaude-sonnet-53/5Products like literature summarization tools, research alert systems, and AI search assistants exist and are used, but they cover only the reading/discovery portion, not colleague discussion or conference participation.

Write grant proposals to procure external research funding.

39

CI 25–52 · exposure 38 · augmentation 75 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Academic adoption of AI for grant writing remains exploratory; most institutions have not integrated AI into formal proposal workflows, and many funders discourage or require disclosure of AI use. Adoption velocity is slow compared to information-sector automation patterns.
Sector adoption velocityclaude-sonnet-52/5Higher education and research administration are historically slow adopters of AI tools for high-stakes writing tasks, with cautious, uneven uptake compared to fast-moving sectors like finance or tech.
Augmentation potentialclaude-haiku-4-5-202510014/5AI excels at accelerating first-draft generation, literature synthesis, and formatting compliance—tasks that occupy significant faculty time. A researcher using AI to outline, draft sections, and organize citations can substantially increase proposal output quality and speed while retaining full strategic and narrative control.
Augmentation potentialclaude-sonnet-54/5AI is widely useful for drafting sections, improving clarity, formatting to funder guidelines, and brainstorming framing, meaningfully speeding up the writing process while the researcher retains final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can draft sections like literature reviews and budget justifications, grant writing requires unique institutional knowledge, researcher vision, and strategic narrative alignment with funder priorities that current systems cannot fully capture. Meaningful automation would require 50%+ time savings at equal quality, but human review and substantial revision remain necessary.
Task automatabilityclaude-sonnet-53/5AI can draft substantial portions of grant text (background, methods framing, budget justification language) given inputs, but strategic framing, novel research ideas, and tailoring to funder priorities still require significant human effort to reach equal quality.
Adoption barriersclaude-haiku-4-5-202510014/5Grant proposals must reflect the researcher's genuine vision and institutional commitment, creating inherent human-in-the-loop requirements. Funding agencies expect authentic researcher voice and institutional accountability, and legal liability for misrepresented claims rests with the institution and researcher, not the tool provider.
Adoption barriersclaude-sonnet-52/5No licensing requirement for grant writing itself, but institutional signatures, PI accountability, and funder expectations of original intellectual contribution create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI tools for writing assistance cost less than hiring a grants professional, but a faculty member's time to integrate AI output, verify compliance, and ensure strategic alignment remains substantial. The all-in cost per completed proposal still approaches or exceeds a portion of the researcher's loaded wage.
Cost vs. human wageclaude-sonnet-53/5AI drafting reduces some writing time cheaply, but the overall cost is still dominated by expert oversight, iteration, and institutional review, making net savings moderate rather than order-of-magnitude.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI writing assistants exist and can help with drafting, but no deployed product reliably produces submission-ready grant proposals end-to-end. Products show material weaknesses in understanding funding priorities, compliance requirements, and institutional context specificity that make them unsuitable for autonomous use.
Technical feasibility todayclaude-sonnet-53/5LLM tools (e.g., ChatGPT, specialized grant-writing assistants) are used in practice to draft proposal sections, but no product reliably produces a fundable, submission-ready proposal without heavy expert revision.

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

32

CI 25–39 · exposure 30 · augmentation 75 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education remains conservative in automating advising; adoption is slow and mainly experimental. Most institutions use AI as a chatbot supplement, not replacement; cultural and organizational barriers limit rapid rollout of autonomous advising.
Sector adoption velocityclaude-sonnet-52/5Higher education is a slower-adopting sector for AI advising; while some universities pilot chatbots, faculty-driven academic and career advising remains largely traditional.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can effectively assist advisors by drafting personalized curriculum suggestions, summarizing career pathways, flagging prerequisite conflicts, and retrieving institution-specific policies, meaningfully reducing preparation time while the advisor retains judgment and relationship.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist by aggregating career data, suggesting course pathways, drafting communications, and answering routine questions, freeing instructors for higher-value guidance.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can provide generic career information and curriculum overviews, advising students requires understanding individual learning goals, constraints, aptitudes, and negotiating institutional constraints. Current AI lacks the sustained context and judgment to replace the full advisory process, though it can draft templates or suggest options.
Task automatabilityclaude-sonnet-52/5Advising requires understanding an individual student's history, goals, and institutional nuances, plus building rapport; AI can support but not fully replace this personalized, relationship-based interaction today.rating
Adoption barriersclaude-haiku-4-5-202510014/5Institutional policy, accreditation, and student expectations often require face-to-face or authenticated human advisors for curriculum and career guidance; liability for poor guidance can fall on the institution and advisor. Substitution faces organizational and regulatory friction.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement for academic advising, but institutional policy, liability for poor guidance, and student preference for human mentorship create meaningful friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5Advising is a labor-intensive, high-touch function performed by relatively lower-cost educators. While AI inference is cheap, per-student oversight and integration into institutional advising workflows adds non-trivial cost; likely comparable to existing labor rather than substantially cheaper.
Cost vs. human wageclaude-sonnet-53/5AI advising tools are cheap to run, but human oversight, error correction, and complex case handling keep costs roughly comparable when quality is held constant.
Technical feasibility todayclaude-haiku-4-5-202510013/5Chatbots and educational platforms offer basic career guidance and curriculum information with reasonable accuracy, but deployed systems are narrowly scoped or require heavy human oversight. No production system fully replaces human advisors at institutional scale; most are assistive rather than autonomous.
Technical feasibility todayclaude-sonnet-52/5Chatbots and advising tools exist for basic FAQs and degree-planning, but no deployed product reliably handles nuanced career/curriculum advising for postsecondary students at scale.

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

31

CI 29–34 · exposure 25 · augmentation 75 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI in postsecondary curriculum planning remains limited; most institutions are in pilot phases with AI-assisted drafting rather than automated or delegated planning. Academic culture and governance move slowly on core educational functions.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools unevenly and cautiously for curriculum design, with pilots and individual faculty experimentation more common than institution-wide production deployment.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist faculty by generating first-draft content, suggesting learning outcomes alignment, surfacing pedagogical literature, and automating material formatting—all while the faculty member retains full curricular judgment. This augmentation substantially raises productivity on content preparation phases.
Augmentation potentialclaude-sonnet-54/5AI is already useful for brainstorming course structures, drafting materials, updating content with current research, and suggesting assessment methods, meaningfully speeding up an instructor's curriculum development work.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with generating course outlines, content summaries, and draft materials, but curriculum planning requires substantive pedagogical judgment, alignment with institutional standards, and evaluation of student learning outcomes. The creative and evaluative components critical to revision cycles remain heavily dependent on human expertise and context.
Task automatabilityclaude-sonnet-52/5AI can draft syllabi, suggest readings, and generate assessment items, but the final curricular judgment about pedagogical sequencing, learning objectives, and institutional/accreditation fit requires domain expertise and contextual decision-making beyond current AI capability to fully replace.
Adoption barriersclaude-haiku-4-5-202510014/5Postsecondary institutions maintain accreditation standards, faculty governance structures, and quality assurance processes that legally and organizationally require faculty expertise and sign-off on curriculum decisions. Collective bargaining agreements often protect curriculum design as a core faculty prerogative.
Adoption barriersclaude-sonnet-53/5No licensing requirement mandates a human perform this, but accreditation standards, departmental governance, and faculty ownership of curriculum content create moderate organizational friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510013/5AI content generation tools are inexpensive per iteration, but effective curriculum planning integration still requires faculty time for oversight and validation. The all-in cost (tool subscription + faculty review) is roughly comparable to faculty time spent on these tasks from scratch.
Cost vs. human wageclaude-sonnet-53/5AI drafting tools are cheap per-query, but the human oversight, subject-matter validation, and institutional review needed to finalize curricula narrows the cost advantage to roughly comparable once integration and correction time are included.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools exist for content generation and basic course design templates, no deployed product reliably handles the full curriculum evaluation and revision cycle with the nuance required in postsecondary education. Products exist for drafting but not for the integrated planning and evaluation at institutional quality standards.
Technical feasibility todayclaude-sonnet-52/5Products like ChatGPT and specialized ed-tech tools can generate lesson plans and course outlines, but no deployed system reliably performs holistic curriculum evaluation and revision at scale in postsecondary environmental science departments today.

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

28

CI 25–30 · exposure 25 · augmentation 63 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education procurement remains largely manual and relationship-driven; while some universities use e-procurement systems, full AI-driven material selection and acquisition remains rare, with institutions preferring human faculty input on educational materials.
Sector adoption velocityclaude-sonnet-52/5Higher education procurement processes are typically slow-moving and administratively bound, with limited AI adoption in academic purchasing workflows to date.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could meaningfully assist by searching catalogs, comparing suppliers, tracking inventory, and flagging budget overruns, allowing faculty to focus on the judgment-critical aspects of material selection; however, the core task remains anchored in human decision-making.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist by recommending textbooks, comparing lab equipment options, summarizing reviews, and drafting purchase orders, saving faculty research time.
Task automatabilityclaude-haiku-4-5-202510012/5Procurement involves significant judgment about institutional needs, vendor selection, budget constraints, and domain-specific requirements that currently require human oversight. AI could assist with inventory tracking or supplier research, but the core decision-making and authorization steps remain human-dependent.
Task automatabilityclaude-sonnet-52/5AI can suggest textbook options or compile supplier lists, but selecting appropriate materials requires judgment about curriculum fit, budget constraints, and physically procuring/ordering lab equipment, which AI cannot execute end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Institutional procurement typically requires authorization by licensed purchasing officers, departmental approval chains, and compliance with university policies and regulations; liability for incorrect material selections falls on the institution, creating legal and operational friction.
Adoption barriersclaude-sonnet-53/5Institutional purchasing rules, budget approval processes, and departmental authority create moderate friction, though no formal licensing requirement exists for material selection itself.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI integration for procurement (including vendor APIs, approval workflows, and human oversight) is comparable to or potentially more expensive than having staff manage selections directly, especially for mid-sized academic departments with established supplier relationships.
Cost vs. human wageclaude-sonnet-52/5AI assistance for research is cheap, but the overall task still requires human review, purchasing authority, and vendor coordination, limiting overall cost savings.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can search for suppliers and compare prices, no production system reliably handles the full procurement workflow end-to-end for educational institutions, which involve institutional policies, approval chains, and relationship management with vendors.
Technical feasibility todayclaude-sonnet-52/5No deployed product autonomously manages procurement decisions and purchasing for academic departments; existing tools only assist with research or recommendation subtasks.

Prepare and deliver lectures to undergraduate or graduate students on topics such as hazardous waste management, industrial safety, and environmental toxicology.

27

CI 25–29 · exposure 25 · augmentation 75 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education adoption of AI lecture delivery remains limited; institutions are using AI for content drafting and administrative support, but full-lecture replacement is not yet occurring at scale, with most pilots still experimental.
Sector adoption velocityclaude-sonnet-52/5Higher education adoption of AI for lecture delivery is still nascent, with instructors mainly using AI for prep and materials rather than full lecture automation.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can meaningfully assist instructors by generating lecture drafts, creating visual aids, synthesizing research on emerging topics like environmental toxicology, and preparing assessments, substantially raising instructor productivity while the human remains responsible for delivery and pedagogical decisions.
Augmentation potentialclaude-sonnet-54/5AI substantially helps in preparing lecture content, generating examples, summarizing recent research on toxicology or hazardous waste topics, and creating supplementary materials, boosting instructor productivity.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can generate lecture outlines and visual materials, delivering live lectures requires real-time interaction, student engagement monitoring, and responsive pedagogical adjustment that current systems cannot reliably perform end-to-end with 50% time savings at equal quality.
Task automatabilityclaude-sonnet-52/5AI can help draft lecture content and slides, but live delivery, adapting to student questions, and maintaining classroom engagement require human presence and judgment that current systems cannot fully replace.
Adoption barriersclaude-haiku-4-5-202510014/5Universities have accreditation standards, student expectations, and institutional policies that typically require a qualified human instructor to directly deliver and be accountable for instruction; human-contact and pedagogical judgment requirements are significant legal and organizational barriers.
Adoption barriersclaude-sonnet-54/5Accredited institutions require qualified, often credentialed faculty to teach and assess students, and there are strong norms and accreditation requirements favoring human instructors of record.
Cost vs. human wageclaude-haiku-4-5-202510012/5Generating lecture drafts has low marginal cost, but the full pipeline—content creation, visual design, integration with learning management systems, and oversight—plus the human instructor still being essential makes all-in costs comparable to or exceeding current faculty labor.
Cost vs. human wageclaude-sonnet-53/5AI tools can cheaply generate lecture materials, reducing prep time costs, but full lecture delivery still requires a paid instructor, keeping overall cost comparable to human-only delivery.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI can draft lecture content and slides, but no deployed product reliably delivers full lectures with student engagement, real-time question handling, and dynamic teaching adjustments at production scale in university settings.
Technical feasibility todayclaude-sonnet-52/5Products exist for content generation (slide decks, lecture notes) and some AI-narrated video lectures, but no deployed product reliably delivers full interactive lecture experiences at scale in postsecondary environmental science courses.

Initiate, facilitate, and moderate classroom discussions.

19

CI 6–32 · exposure 13 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education remains slow to adopt AI for core instructional tasks like live classroom facilitation, with most pilots focusing on asynchronous tools (grading aids, chatbots) rather than real-time discussion moderation; adoption is still experimental, not production-scaled.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI unevenly and cautiously for teaching-facing tasks, with classroom facilitation itself seeing minimal real deployment despite institutional experimentation with AI-related tools.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by suggesting follow-up questions, summarizing discussion threads, or flagging off-topic comments, providing useful supportive functions; however, the core pedagogical judgment and interpersonal skill of facilitation remain primarily human-driven.
Augmentation potentialclaude-sonnet-53/5AI can help instructors prepare discussion prompts, summarize themes, or generate follow-up questions, but doesn't materially change real-time facilitation itself.
Task automatabilityclaude-haiku-4-5-202510011/5Facilitating and moderating classroom discussions requires real-time responsiveness to student comments, dynamic redirection, social judgment, and the ability to build rapport—all contextual and interpersonal elements that current AI cannot reliably perform in a live classroom setting at equivalent quality.
Task automatabilityclaude-sonnet-52/5Leading discussions requires real-time reading of student engagement, adaptive follow-up questioning, and classroom management that current AI cannot reliably replicate end-to-end in a live setting.'
Adoption barriersclaude-haiku-4-5-202510015/5Postsecondary teaching involves direct student contact, institutional accreditation requirements, and professional expectations that faculty personally engage in classroom facilitation; liability and regulatory frameworks treat the instructor as accountable for learning outcomes and classroom management.
Adoption barriersclaude-sonnet-54/5Accredited institutions require a qualified instructor of record to lead courses, and human presence and authority in classroom dynamics is an implicit institutional and pedagogical requirement.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI infrastructure for discussion support is inexpensive relative to instructor wages, but the task itself—live facilitation—would require either human oversight or full replacement, making direct cost comparison difficult; however, the marginal cost of AI-assisted discussion tools is very low.
Cost vs. human wageclaude-sonnet-52/5Since AI cannot substitute for the live-facilitation role, any AI tool would be additive cost rather than a replacement, making cost comparison unfavorable relative to the instructor already present.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can generate discussion prompts and summarize conversation threads in limited, text-based settings, no deployed product reliably facilitates live classroom discourse with appropriate pedagogical judgment, tone-matching, and the ability to read room dynamics and adjust in real time.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously facilitates live in-person postsecondary classroom discussions; existing chatbot discussion tools are limited to online/asynchronous supplementary contexts.

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

18

CI 10–25 · exposure 17 · augmentation 63 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic research remains slow to adopt automation; researchers control their own workflows and institutions emphasize human scholarship. AI adoption in academia is limited to auxiliary tools rather than autonomous research production.
Sector adoption velocityclaude-sonnet-52/5Higher education and academic research are moderate adopters of AI tools for writing and analysis, but adoption for actually conducting and publishing original research is still nascent and cautious due to integrity concerns.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist with literature synthesis, draft writing, data visualization, and statistical computation, raising researcher productivity on specific subtasks while the researcher remains responsible for validation and intellectual direction.
Augmentation potentialclaude-sonnet-54/5AI substantially aids literature synthesis, statistical analysis, manuscript drafting, and editing, meaningfully boosting researcher productivity while the scientist retains control over design and conclusions.
Task automatabilityclaude-haiku-4-5-202510011/5Research discovery, hypothesis formation, experimental design, and knowledge synthesis require domain expertise, creativity, and judgment that current AI cannot replicate end-to-end. While AI can assist with literature review and data analysis, the core intellectual work of conducting original research remains fundamentally human.
Task automatabilityclaude-sonnet-52/5AI can assist with literature review, data analysis, and drafting, but original environmental science research requires fieldwork, experimental design, novel insight, and judgment that current AI cannot autonomously perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Academic publication requires human researcher authorship and institutional affiliation; peer review expects human accountability; funding agencies require human principal investigators. Regulatory and professional norms strongly protect this task for credentialed humans.
Adoption barriersclaude-sonnet-54/5Academic publishing requires named authorship, peer review, ethical accountability, and credentialed expertise; journals and institutions have strong norms against non-human-led research, creating substantial structural barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5The fully-loaded cost of a researcher (salary, benefits, equipment, overhead) far exceeds the cost of AI writing or analysis tools, and AI cannot replace the researcher's effort in designing and executing novel research.
Cost vs. human wageclaude-sonnet-52/5While AI can cheaply assist with parts of the workflow (summarization, coding, drafting), the human researcher's time for design, fieldwork, analysis, and validation remains the dominant cost, so overall savings are modest.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product conducts independent research in environmental science from problem formulation through publication. AI tools exist for components (writing, literature search, statistical analysis) but not for the integrative research process itself, which requires human oversight and validation at every stage.
Technical feasibility todayclaude-sonnet-52/5Products like AI writing assistants and literature-review tools are used in research support, but no deployed system independently conducts original scientific research and publishes credible findings without heavy human oversight.

Review papers or serve on editorial boards for scientific journals, and review grant proposals for various agencies.

17

CI 5–29 · exposure 17 · augmentation 63 · importance 3.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Scientific publishing and grant review remain highly conservative sectors with strong institutional and professional norms favoring human expert oversight; AI adoption is limited to preliminary screening and summary tools, not replacement of reviewers.
Sector adoption velocityclaude-sonnet-52/5Academic publishing and grant review are conservative, slow-moving institutions with cautious, uneven AI adoption due to concerns about confidentiality, bias, and accountability.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by summarizing papers, identifying methodological gaps, and organizing grant proposal content, reducing reviewer reading load; however, the core judgment remains inherently human-dependent.
Augmentation potentialclaude-sonnet-54/5AI can meaningfully assist reviewers by summarizing papers, checking statistical validity, identifying related literature, and drafting review comments, significantly speeding up the human's workflow.
Task automatabilityclaude-haiku-4-5-202510011/5Reviewing scientific papers and grant proposals requires nuanced judgment of methodological rigor, novelty, significance, and fit—domains where current AI lacks domain expertise validation and cannot reliably assess research quality at the level expected by journals and funding agencies.
Task automatabilityclaude-sonnet-52/5AI can summarize papers and flag methodological issues, but substantive peer review requires domain judgment, novelty assessment, and accountability that current AI cannot fully replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Editorial boards and grant review processes are governed by professional standards, journal policies, and funder regulations that legally and ethically require human expert judgment; liability and research integrity requirements make automated substitution infeasible.
Adoption barriersclaude-sonnet-54/5Peer review and grant review are gatekeeping functions requiring named, credentialed experts for accountability and conflict-of-interest management, creating strong institutional and reputational barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5Peer review and grant evaluation are typically performed by volunteer experts or lightly compensated reviewers; the cost of AI-generated review plus human oversight would likely exceed the minimal labor cost of expert review.
Cost vs. human wageclaude-sonnet-53/5AI tools can cheaply assist with initial screening and summarization, but the human reviewer's time for actual evaluative judgment remains the dominant cost, keeping overall savings moderate.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can summarize papers and flag structural issues, no deployed system reliably performs peer review or grant evaluation end-to-end; human experts remain the standard in all major journals and funding agencies, with AI used only as a screening aid at best.
Technical feasibility todayclaude-sonnet-52/5Some journals experiment with AI-assisted screening (plagiarism, statistics checks), but no deployed product independently performs full peer review or grant evaluation reliably in production.

Provide professional consulting services to government or industry.

13

CI 9–16 · exposure 5 · 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 in environmental science remains a high-touch, relationship-driven field with minimal AI automation in production. Adoption is limited to assistive tools for research and drafting, not autonomous consulting delivery.
Sector adoption velocityclaude-sonnet-52/5Higher education and specialized environmental consulting sectors have been slower to adopt AI agents for client-facing expert advisory work compared to fast-moving digital sectors.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by synthesizing research, generating draft reports, and analyzing environmental datasets, allowing consultants to cover more ground and iterate faster. However, the strategic and interpersonal core remains human-dependent.
Augmentation potentialclaude-sonnet-54/5AI can strongly assist with literature review, data analysis, report drafting, and scenario modeling, meaningfully boosting the productivity of the human expert delivering consulting services.
Task automatabilityclaude-haiku-4-5-202510011/5Consulting requires domain expertise synthesis, stakeholder negotiation, contextual judgment, and recommendation of novel solutions—tasks that current AI cannot perform end-to-end. AI cannot independently advise government or industry on complex environmental policy or technical strategy without human expert oversight.
Task automatabilityclaude-sonnet-51/5Consulting requires original expert judgment, synthesis of context-specific data, and accountability that current AI cannot autonomously provide end-to-end for government or industry clients.
Adoption barriersclaude-haiku-4-5-202510014/5Professional consulting carries high liability and legal risk; clients typically require a licensed expert to sign off on recommendations. Regulatory frameworks and client trust norms strongly favor human accountability and personal professional standing.
Adoption barriersclaude-sonnet-54/5Professional consulting often requires credentialed expertise, institutional affiliation, and accountability for advice given to government/industry, creating significant liability and trust barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI research and document generation tools can reduce preparation time, but the core consulting work—client relationship, strategic recommendations, liability—must be performed by expensive domain experts. All-in cost remains dominated by human expertise.
Cost vs. human wageclaude-sonnet-52/5While AI can cut research time, the bulk of consulting value is expert judgment, credibility, and liability-bearing recommendations, so cost savings versus a full human consultant are modest.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably delivers professional consulting services autonomously. AI tools may assist with research or draft reports, but the client-facing advisory and decision-making responsibility remains strictly human in production settings.
Technical feasibility todayclaude-sonnet-52/5AI tools can support research and drafting for consultants, but no deployed product independently delivers professional environmental science consulting services to clients.

Maintain regularly scheduled office hours to advise and assist students.

6

CI 0–11 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic institutions operate under tenure and employment frameworks that make displacement of faculty office hours culturally, legally, and contractually difficult; adoption of AI for core advising roles remains minimal and limited to supplementary chatbots.
Sector adoption velocityclaude-sonnet-52/5Higher education is a slow-adopting sector for replacing direct faculty-student interaction, though AI tutoring tools are being piloted alongside, not instead of, office hours.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist by drafting advising notes, summarizing student records, or scheduling, but the core task—real-time human presence and judgment—offers minimal scope for augmentation; any gains are peripheral to the essential function.
Augmentation potentialclaude-sonnet-53/5AI can help by answering routine student questions, drafting responses, or triaging inquiries before or after office hours, but the core advising interaction remains human-led.
Task automatabilityclaude-haiku-4-5-202510011/5Maintaining office hours to advise and assist students requires real-time interpersonal engagement, judgment about individual student needs, and the ability to build rapport—capabilities current AI systems cannot reliably replicate. The task is fundamentally human-centered and cannot be meaningfully automated to meet the 50% time-saving threshold.
Task automatabilityclaude-sonnet-51/5This task requires a human faculty member to be physically or virtually present and personally accountable for student advising; AI cannot substitute for the institutional requirement of a professor holding office hours.
Adoption barriersclaude-haiku-4-5-202510015/5Strong regulatory, organizational, and pedagogical barriers exist: accreditation standards expect faculty-student interaction, institutions legally employ faculty to fulfill advising obligations, and students/parents expect human mentorship and accountability that cannot be delegated to machines.
Adoption barriersclaude-sonnet-54/5Office hours are an institutional and often contractual/accreditation expectation tied to the instructor of record, creating strong organizational and role-based barriers to full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying and maintaining an AI system to simulate office hours, plus oversight and error correction, would exceed the salary cost of a faculty member already required to hold office hours as part of their contracted duties.
Cost vs. human wageclaude-sonnet-52/5The task is bundled into the professor's salary and role, so there's no separate 'AI cost' comparison; any AI assistance is additive rather than substitutive, keeping cost savings low.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably conducts live office hours that replace a human advisor; chatbots can provide generic information but cannot deliver the nuanced, personalized guidance and mentorship that office hours provide.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs scheduled faculty office hours as a substitute for the professor; chatbots can supplement but not replace this designated human availability.

Supervise students' laboratory and field work.

4

CI 0–9 · exposure 8 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions, especially postsecondary science departments, are laggards in AI adoption for core teaching functions, and there is no meaningful effort to automate student supervision given safety and accountability requirements.
Sector adoption velocityclaude-sonnet-51/5Higher education lab/field supervision remains a physically-anchored, low-digitization task with essentially no AI displacement occurring.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with minor tasks like documenting observations or analyzing post-session data, but it offers limited augmentation to the core supervision function, which depends on human judgment, presence, and responsive intervention.
Augmentation potentialclaude-sonnet-52/5AI can help with pre-lab safety checklists, data logging, or generating lab manuals, but offers minimal real-time assistance during actual physical supervision.
Task automatabilityclaude-haiku-4-5-202510012/5AI cannot meaningfully supervise students in laboratory or field settings, which requires real-time presence, safety oversight, and intervention capacity. While AI could assist with some documentation or post-hoc review, the core supervision function—ensuring student safety and guiding hands-on work—remains fundamentally human.
Task automatabilityclaude-sonnet-51/5Direct physical supervision of students conducting lab experiments or field work requires real-time presence, safety monitoring, and hands-on judgment that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Educational institutions have legal and fiduciary duties to supervise student safety in laboratories and field work; a qualified human instructor must be physically present and responsible. This creates hard barriers rooted in liability, institutional policy, and educational accreditation standards.
Adoption barriersclaude-sonnet-55/5Safety regulations, liability for student injury, and institutional policies require a qualified human instructor present during lab and field activities.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems cannot replace human supervision for this task, so cost comparison is moot; attempting to use AI as a substitute would create liability and safety risks that make it far more expensive than employing a human supervisor.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for physical supervision, so cost comparison favors the human by default since the AI alternative doesn't exist for the core task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs live supervision of student laboratory or field work in any production environment. The task requires embodied presence, situational awareness, and safety responsibility that current AI systems cannot deliver.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises students physically in labs or field settings; at most AI provides supplementary monitoring tools, not autonomous supervision.

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

4

CI 0–7 · exposure 0 · 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 structural changes to faculty roles, and supervisory responsibility remains a core function that institutions have not sought to automate or replace, even in early-adopter research contexts.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools for content generation and grading assistance but supervisory and mentorship roles remain largely untouched by automation.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist a supervising faculty member with administrative tasks (scheduling, progress tracking summaries, literature recommendations) but cannot replace the judgment and mentorship that define the role.
Augmentation potentialclaude-sonnet-53/5AI can help by drafting feedback, summarizing student progress, or suggesting research resources, but the core supervisory judgment and mentoring remain human-driven.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising teaching, internships, and research requires ongoing human judgment about student progress, mentorship, discipline, and research direction. Current AI cannot meaningfully replace the relational, evaluative, and advisory core of this supervisory role.
Task automatabilityclaude-sonnet-51/5Supervising students' teaching, internships, and research requires mentorship, judgment, and relationship-building that current AI cannot perform end-to-end.'
Adoption barriersclaude-haiku-4-5-202510015/5Supervision of graduate research and teaching requires a credentialed faculty member with legal and institutional responsibility. Accreditation bodies, universities, and funding agencies mandate human supervision of student researchers and teaching assistants.
Adoption barriersclaude-sonnet-54/5Institutional accreditation, degree-granting requirements, and academic norms mandate qualified faculty oversight of student research and teaching, creating strong structural barriers.
Cost vs. human wageclaude-haiku-4-5-202510011/5The all-in cost of AI systems (infrastructure, integration, human oversight to ensure quality mentorship and institutional compliance) would exceed the cost of a faculty member already present in the institution.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute delivering equivalent supervisory outcomes, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs academic supervision at scale. AI lacks the contextual understanding, relationship continuity, and institutional authority required to supervise students and evaluate their intellectual work.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises student research or teaching internships in place of a faculty member; this remains a human-only function in academia.

Collaborate with colleagues to address teaching and research issues.

4

CI 0–7 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education remains a laggard sector for AI adoption, and the interpersonal, deliberative nature of faculty collaboration has shown minimal displacement by automation technology.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools for research support and content generation but has not restructured collegial collaboration processes themselves.
Augmentation potentialclaude-haiku-4-5-202510012/5AI tools could modestly assist with organizing research literature or synthesizing data for discussion, but the creative problem-solving and relationship-building at the heart of faculty collaboration remains primarily human-driven and offers limited augmentation opportunity.
Augmentation potentialclaude-sonnet-53/5AI can help draft joint proposals, summarize research literature, or organize meeting notes, aiding collaboration without replacing the human interaction itself.
Task automatabilityclaude-haiku-4-5-202510011/5Collaboration on teaching and research issues fundamentally requires human judgment, relationship-building, and contextual understanding of institutional and pedagogical nuances that current AI cannot replicate end-to-end. While AI can assist with drafting or organizing information, it cannot substitute for the interpersonal negotiation and decision-making central to this task.
Task automatabilityclaude-sonnet-51/5Collaboration among colleagues on teaching and research strategy requires relationship-building, negotiation, and shared judgment that current AI cannot perform end-to-end.img AI has no agentic role as a colleague in these interactions.'
Adoption barriersclaude-haiku-4-5-202510015/5Collaboration with colleagues is an inherently human-contact and judgment-intensive activity embedded in academic culture and governance; institutional norms, intellectual responsibility, and collegial trust create strong barriers to any form of substitution.
Adoption barriersclaude-sonnet-54/5Academic collaboration is embedded in institutional governance, tenure/promotion processes, and departmental norms requiring human participation and accountability.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI cannot perform this task at all, so the cost comparison is moot—a human colleague must remain involved, making any AI cost-additive rather than cost-replacing.
Cost vs. human wageclaude-sonnet-51/5There is no AI system performing this collaborative role, so no cost comparison favors AI; the task is inherently interpersonal.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs genuine collaboration on academic and research issues; this requires sustained dialogue, mutual accountability, and organizational context that exceeds current AI capabilities in production environments.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for a faculty member collaborating with peers on curriculum or research direction; this remains a human social and intellectual process.

Act as advisers to student organizations.

3

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions are among the slowest sectors to adopt AI automation, and advising—which depends on trust and human judgment—is unlikely to be displaced by automation in the foreseeable future.
Sector adoption velocityclaude-sonnet-51/5Higher education advising roles show minimal AI displacement; this is a low-digitization, relationship-driven function.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist with administrative scheduling or documentation tasks related to student organizations, but offers minimal assistance with the core advising, mentoring, and judgment work itself.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, drafting communications, or brainstorming event ideas, but offers limited support for the core mentoring and oversight aspects of the role.
Task automatabilityclaude-haiku-4-5-202510011/5Acting as an adviser to student organizations requires real-time mentoring, relationship-building, pastoral care, and nuanced judgment about student needs and organizational dynamics. These are fundamentally human-centric tasks that AI cannot meaningfully perform end-to-end today.
Task automatabilityclaude-sonnet-51/5Advising student organizations requires mentorship, relationship-building, event guidance, and institutional judgment that AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Educational institutions require credentialed faculty advisers for student organizations, and there are legal, fiduciary, and reputational barriers to delegating advising duties away from qualified humans. Regulatory and organizational requirements strongly protect this human role.
Adoption barriersclaude-sonnet-54/5Institutions typically require a designated faculty/staff member to formally advise and be accountable for student organizations, creating an organizational and often policy-based requirement for a human.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of an AI system attempting to provide advising services would exceed the cost of a faculty member allocating existing time to this role, given the low automation potential and high oversight requirements.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute providing this output, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably advises student organizations in the way humans do. This task requires contextual understanding of individual students, institutional culture, and adaptive guidance that current AI systems cannot replicate in production.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product serves as an actual faculty advisor to a student organization; this remains a human relational role.

Participate in campus and community events.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 2.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This task involves human discretion and institutional presence in community contexts, which are inherently resistant to AI substitution and not a focus of automation initiatives in higher education.
Sector adoption velocityclaude-sonnet-51/5Higher education and in-person community engagement are low-digitization, slow-adopting contexts for this kind of physical/social task.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist minimally (e.g., helping prepare event materials or scheduling), but the core act of participating in and engaging at events remains entirely human-driven with limited augmentation potential.
Augmentation potentialclaude-sonnet-52/5AI can help with scheduling, event planning materials, or follow-up communications, but offers minimal assistance to the act of participation itself.
Task automatabilityclaude-haiku-4-5-202510011/5Participation in campus and community events requires real-time physical presence, social interaction, relationship-building, and spontaneous engagement—none of which current AI can perform end-to-end. AI cannot substitute for a human instructor's presence and networking role at these events.
Task automatabilityclaude-sonnet-51/5Participating in physical campus and community events requires human presence, social interaction, and representation of the institution, which AI cannot perform end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Institutional representation, faculty leadership, and community trust require a licensed human faculty member. Events often require accredited personnel to lead or vouch for the institution's mission and credibility.
Adoption barriersclaude-sonnet-54/5Institutional norms, expectations of faculty visibility, and community relationship-building strongly favor human presence, though not a formal licensing requirement.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI has no meaningful cost advantage here because the task cannot be automated; a human faculty member must attend, making direct cost comparison irrelevant.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute for physical/social attendance, so cost comparison favors the human by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can autonomously attend, participate in, and represent an educational institution at community events. This task fundamentally requires human agency and embodied presence.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product attends or participates in in-person events on behalf of a person; this remains outside current product capabilities.

Perform administrative duties, such as serving as department head.

0

CI 0–0 · exposure 0 · augmentation 38 · 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 have shown minimal adoption of AI for administrative leadership roles. Educational governance remains highly centralized around human administrators, with little sector-wide movement toward automation of department headship duties.
Sector adoption velocityclaude-sonnet-51/5Higher education administrative leadership roles show essentially no movement toward AI-driven substitution; adoption in this specific function is negligible.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can provide modest assistance with email management, scheduling, or data compilation, but the strategic and interpersonal nature of department headship limits meaningful augmentation to narrow administrative support tasks rather than core leadership functions.
Augmentation potentialclaude-sonnet-53/5AI tools can help draft memos, summarize meeting notes, schedule tasks, or analyze budget data, providing moderate assistance to a department head's administrative workload.
Task automatabilityclaude-haiku-4-5-202510011/5Administrative duties involving department headship require complex human judgment, stakeholder negotiation, personnel decisions, and institutional knowledge that current AI cannot perform end-to-end. These responsibilities demand accountability, decision-making authority, and interpersonal leadership that exceed what AI systems can reliably do today.
Task automatabilityclaude-sonnet-51/5Serving as department head involves leadership, personnel decisions, conflict resolution, budget negotiation, and institutional politics that require human judgment, relationships, and accountability that current AI cannot replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Hard regulatory and institutional barriers exist: a human department head must legally hold the position, sign off on personnel and budget decisions, and bear fiduciary responsibility. Accreditation, employment law, and institutional governance explicitly require human accountability in these roles.
Adoption barriersclaude-sonnet-55/5Department head roles typically require formal faculty appointment, institutional governance approval, and accountability structures that legally and organizationally require a qualified human.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of attempting to automate department headship duties (if feasible) would far exceed human labor costs, since a department head's salary is typically $80k–$120k+ annually, and the specialized expertise and accountability required cannot be replaced by current AI at any price point.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that can substitute for this role, so cost comparison favors the human entirely; AI cannot deliver the output at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably performs the role of department head. While AI can assist with scheduling or document drafting, the core duties—hiring decisions, budget allocation, policy enforcement, faculty evaluation, and institutional representation—remain exclusively human functions in all current academic deployments.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs the role of a department head; this remains a human leadership and administrative position with no automation substitute in production.

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

0

CI 0–0 · exposure 0 · augmentation 38 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic governance is highly resistant to automation. Committees are structural elements of institutional decision-making tied to faculty roles, tenure protections, and accreditation standards; no meaningful adoption signal exists.
Sector adoption velocityclaude-sonnet-51/5Higher education governance structures are slow-moving and highly resistant to structural automation of representative roles like committee membership.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with preparing meeting materials, summarizing previous discussions, or drafting proposals, but the deliberative and decision-making core of committee work remains human-centered and cannot be meaningfully augmented by AI participation.
Augmentation potentialclaude-sonnet-53/5AI can help summarize meeting materials, draft policy language, or analyze data to inform committee discussions, meaningfully aiding preparation even though it can't replace the deliberative role.
Task automatabilityclaude-haiku-4-5-202510011/5Committee work requires deliberation, consensus-building, negotiation, and judgment on institutional priorities and policies. AI cannot participate in discussions, vote, represent constituencies, or take responsibility for governance decisions—core requirements of committee service.
Task automatabilityclaude-sonnet-51/5Committee service involves in-person deliberation, negotiation, institutional politics, and judgment calls that require human presence and accountability; AI cannot substitute for the actual role of a committee member.
Adoption barriersclaude-haiku-4-5-202510015/5Institutional committees require human faculty members with authority, accountability, and legal standing to make binding decisions on policy and academic matters. Regulatory and fiduciary requirements mandate human participation and signature.
Adoption barriersclaude-sonnet-55/5Committee membership typically requires formal faculty appointment, governance rules, and institutional bylaws mandating human representation, making substitution essentially barred.
Cost vs. human wageclaude-haiku-4-5-202510011/5Committee participation is a governance responsibility performed by salaried faculty. The cost structure is fundamentally different from task outsourcing; there is no meaningful per-task cost comparison to AI.
Cost vs. human wageclaude-sonnet-51/5There is no AI product replacing this role, so cost comparison favors humans by default since AI cannot perform the core function.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can independently serve on committees; humans must attend meetings, engage in dialogue, and sign institutional decisions. AI systems lack legal standing and accountability.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the function of serving as a voting/deliberating committee member representing faculty interests; this remains entirely human-performed.

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.