Psychology Teachers, Postsecondary

25-1066.00
Median wage $80,340/yr41,530 employed (US)Rank #389 of 923 scored · top 42% by substitution

Teach courses in psychology, such as child, clinical, and developmental psychology, and psychological counseling. Includes both teachers primarily engaged in teaching and those who do a combination of teaching and research.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution30
Exposure27
Augmentation60

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

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

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

Tasks on the substitution scale

29 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

14%

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%27

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

Cost vs. human wagew 15%34

panel mean rating 2.4/5 → substitution pressure 34/100

Adoption barriersw 20%inverted — strong barriers lower the score36

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

Sector adoption velocityw 10%27

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

Task breakdown (29 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 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510015/5Nearly all postsecondary institutions have adopted LMS platforms and digital grade-recording systems; this is near-universal in higher education, with decades of deployment history and deep organizational integration.
Sector adoption velocityclaude-sonnet-55/5Higher education has near-universal adoption of digital LMS and gradebook systems for these administrative functions already.
Augmentation potentialclaude-haiku-4-5-202510014/5AI-assisted tools already flag anomalous attendance patterns, highlight at-risk students, and summarize grade distributions, augmenting instructor oversight and decision-making while instructors retain control over grade entry and interpretation.
Augmentation potentialclaude-sonnet-54/5Automated systems substantially reduce faculty time spent on record-keeping, letting instructors focus on teaching while software manages data entry and reporting.
Task automatabilityclaude-haiku-4-5-202510015/5Record-keeping of attendance, grades, and student data is highly structured and routine; modern learning management systems (Canvas, Blackboard, etc.) and administrative tools already automate most or all of this task end-to-end with >50% time savings and equal accuracy.
Task automatabilityclaude-sonnet-55/5Recording attendance and grades is a structured data-entry task easily handled by learning management systems and gradebook software with automation and integration capabilities.
Adoption barriersclaude-haiku-4-5-202510012/5While institutions may require human sign-off on final grades or compliance audits, the actual recording and computation of attendance and grade data faces minimal legal or regulatory barriers; most friction is organizational inertia rather than hard licensing requirements.
Adoption barriersclaude-sonnet-52/5Some institutional policy requires instructor sign-off on final grades, but the record-keeping mechanics themselves face little regulatory or licensing barrier.
Cost vs. human wageclaude-haiku-4-5-202510015/5LMS software and integration costs are amortized across many instructors and students; per-task cost of automated record-keeping is orders of magnitude cheaper than paying a human to manually track and enter attendance and grades.
Cost vs. human wageclaude-sonnet-55/5Automated record-keeping software costs a fraction of the faculty or administrative time it replaces, especially at institutional scale.
Technical feasibility todayclaude-haiku-4-5-202510015/5Mature, widely deployed LMS platforms and student information systems perform this reliably at scale in educational institutions worldwide; these are standard production systems used daily by millions of instructors.
Technical feasibility todayclaude-sonnet-55/5LMS platforms like Canvas, Blackboard, and Moodle already automate attendance tracking and gradebook calculations in production at scale across universities.

Compile bibliographies of specialized materials for outside reading assignments.

79

CI 76–81 · exposure 75 · augmentation 100 · importance 2.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Universities and faculty are beginning to adopt AI-assisted bibliography tools, but adoption remains inconsistent and often limited to drafting support rather than full automation due to scholarly preferences for human oversight.
Sector adoption velocityclaude-sonnet-53/5Higher education is moderately adopting AI research tools, with growing use of AI literature search assistants, though many faculty still compile lists manually or via traditional library tools.
Augmentation potentialclaude-haiku-4-5-202510015/5AI substantially augments faculty productivity by rapidly generating candidate bibliographies, discovering niche materials, and auto-formatting citations, allowing instructors to focus on critical evaluation and pedagogical alignment rather than tedious compilation.
Augmentation potentialclaude-sonnet-55/5AI significantly speeds up literature discovery and organization, letting instructors focus on curating quality and pedagogical fit rather than manual searching.
Task automatabilityclaude-haiku-4-5-202510014/5AI can automatically search academic databases, identify relevant papers and books, cross-reference citation formats, and generate formatted bibliographies with minimal human oversight. Current systems reliably perform literature discovery and citation compilation at 50%+ time savings.
Task automatabilityclaude-sonnet-54/5AI can search literature, identify relevant sources by topic/level, and compile formatted bibliographies quickly, requiring only light instructor review for relevance and accuracy.attribute this to advanced literature search and citation tools.
Adoption barriersclaude-haiku-4-5-202510012/5No licensing requirement mandates human compilation; however, academic institutions may prefer human curation for quality assurance and subject-matter judgment, and instructors may require verification that assignments reflect current scholarship.
Adoption barriersclaude-sonnet-51/5No licensing, regulatory, or liability barriers prevent using AI to compile reading lists; it's a low-stakes administrative/academic support task.
Cost vs. human wageclaude-haiku-4-5-202510015/5An AI query plus light human verification costs negligible amounts per bibliography compared to a faculty member spending hours manually searching databases, evaluating sources, and formatting citations.
Cost vs. human wageclaude-sonnet-55/5Automated bibliography compilation via AI tools costs a small fraction of the faculty time required to manually search and curate reading lists.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (ChatGPT, Claude, Perplexity, academic database APIs) demonstrably generate bibliographies from topic prompts and compile reading lists in production. Some limitations remain around verifying obscure sources and ensuring complete coverage, but the core task is reliably performed.
Technical feasibility todayclaude-sonnet-54/5Deployed tools like AI-powered reference managers, citation generators, and research assistants (e.g., Elicit, Consensus, Scite) reliably compile topical bibliographies today, though subject-specific curation still benefits from human vetting.

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

74

CI 71–76 · exposure 70 · 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 institutions are in early-to-mid adoption: pilots and voluntary use are common, but systematic replacement of instructor material-writing workflows is still emerging. Adoption varies widely by institution and discipline, not yet at the depth seen in professional services or tech sectors.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting generative AI for course prep at a moderate pace with growing faculty use, but institutional caution and inconsistent policies slow broader deployment relative to fast-moving sectors.
Augmentation potentialclaude-haiku-4-5-202510015/5AI strongly augments instructor productivity: it rapidly generates drafts that instructors customize, refine, and align with course goals. Instructors stay in control while offloading repetitive writing, significantly raising output quality and speed.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up drafting of syllabi, assignments, and handouts while instructors retain control over final content, pedagogy, and course-specific judgment.
Task automatabilityclaude-haiku-4-5-202510014/5Current AI systems can generate syllabi, homework assignments, and handouts with substantial time savings. LLMs can draft complete course materials from learning objectives, and with minimal human review, achieve the ≥50% time-saving threshold. However, some customization and institutional integration typically require human oversight.
Task automatabilityclaude-sonnet-54/5LLMs can generate syllabi, homework assignments, and handouts from a course description with substantial time savings, though instructor review/customization is typically still needed for accuracy and alignment with learning objectives.
Adoption barriersclaude-haiku-4-5-202510012/5Few legal or regulatory barriers exist to AI assistance in course material preparation. No licensing requirement mandates human authorship. The main friction is institutional inertia, faculty preference to retain creative control, and accreditation familiarity rather than hard legal constraints.
Adoption barriersclaude-sonnet-52/5No licensing requirement mandates human-only authorship, though some institutional policies on academic integrity and instructor ownership of course content create mild friction.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI inference cost per complete syllabus or assignment set is negligible (under $1), while instructor time saved is worth $50–100+. The all-in cost (API calls, minor oversight) is orders of magnitude cheaper than human labor for generating equivalent materials.
Cost vs. human wageclaude-sonnet-55/5Generating draft syllabi and assignments via AI costs cents to dollars versus hours of faculty time, an order-of-magnitude cost advantage even with review overhead.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed products (ChatGPT, Claude, specialized educational tools) reliably generate course materials at scale. Universities and instructors actively use these systems in production for syllabus and assignment drafting. Minor limitations exist around institution-specific formatting and compliance, but core material generation works well.
Technical feasibility todayclaude-sonnet-53/5Products like ChatGPT, Copilot, and specialized ed-tech tools are widely used by instructors to draft these materials, but reliability varies and outputs still require academic vetting for accuracy and appropriateness.

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

72

CI 56–87 · exposure 70 · augmentation 88 · 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 has rapidly adopted LMS-integrated automated grading and assessment tools over the past 5–10 years, particularly post-pandemic; major institutions routinely use these systems in production for large enrollment courses.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI grading and exam tools unevenly; many institutions remain cautious due to academic integrity and accreditation concerns, limiting deep production-scale adoption.
Augmentation potentialclaude-haiku-4-5-202510015/5AI assists faculty by suggesting exam questions, auto-grading portions of exams, flagging outliers or potential cheating, and providing analytics on student performance, substantially raising instructor productivity while retaining human oversight of assessment design and difficult judgments.
Augmentation potentialclaude-sonnet-54/5AI substantially helps instructors draft question banks, create rubrics, and provide first-pass grading or feedback, meaningfully speeding up the overall task while the instructor retains final oversight.
Task automatabilityclaude-haiku-4-5-202510015/5AI systems can now compile exams from question banks, generate exam content via LLMs, administer online assessments at scale, and grade objective/multiple-choice and short-answer items automatically, achieving well over 50% time savings with equal quality compared to manual grading of large cohorts.
Task automatabilityclaude-sonnet-53/5AI can draft exam questions and grade objective/short-answer content well, but compiling exams aligned to specific course objectives and grading nuanced essay responses at faculty-level quality still requires human judgment for a full end-to-end solution.
Adoption barriersclaude-haiku-4-5-202510012/5While some institutions require faculty sign-off on grading and there is modest resistance to full automation of assessment, no legal licensing or regulatory barrier prevents use of AI for exam compilation, administration, and grading of objective items; adoption is voluntary and organizational friction is low.
Adoption barriersclaude-sonnet-52/5No licensing requirement mandates human grading, though academic integrity policies and institutional norms often require instructor accountability for final grades and assessment design.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-driven exam administration and grading cost pennies per student per assessment once systems are in place, versus instructor hours at $50–100+ per hour; the cost differential is well over an order of magnitude at scale.
Cost vs. human wageclaude-sonnet-54/5AI tools for generating and grading multiple-choice or short-answer exams are inexpensive compared to instructor or TA time, though essay grading still requires costlier human oversight.
Technical feasibility todayclaude-haiku-4-5-202510014/5Learning management systems (Canvas, Blackboard, Moodle) with integrated AI grading tools and third-party services like Gradescope already perform these tasks in production at many institutions; some subjective grading still requires human review, but objective assessment is fully deployed.
Technical feasibility todayclaude-sonnet-53/5Products like AI-assisted quiz generators and automated essay scoring exist and are used in some LMS platforms, but they are not universally reliable for postsecondary-level psychology content grading without instructor review.

Select and obtain materials and supplies, such as textbooks.

69

CI 60–79 · exposure 70 · augmentation 75 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510014/5Higher education institutions increasingly deploy AI-assisted procurement and inventory systems; many use automated requisition systems integrated with budget management. Adoption is common in larger universities and well-resourced departments.
Sector adoption velocityclaude-sonnet-52/5Higher education, especially administrative/curricular tasks, adopts AI tools slowly and unevenly, with pilots more common than full production use for material selection.
Augmentation potentialclaude-haiku-4-5-202510014/5AI significantly assists by generating curated lists of available textbooks, flagging price changes, comparing editions, and recommending open-access alternatives—allowing instructors to make informed choices faster than manual research.
Augmentation potentialclaude-sonnet-54/5AI can efficiently generate lists of relevant textbooks, summarize reviews, compare pricing and editions, greatly assisting instructors while they retain final decision-making.
Task automatabilityclaude-haiku-4-5-202510014/5AI can automate most of this task today: identifying required textbooks based on course syllabi, comparing vendor options, checking inventory, and generating purchase orders. However, final approval and handling of institutional procurement rules still require human oversight in most settings.
Task automatabilityclaude-sonnet-54/5AI can research, compare, and recommend textbooks and materials, and even automate ordering through integrated systems, saving significant time versus manual review of catalogs and syllabi alignment.
Adoption barriersclaude-haiku-4-5-202510013/5Institutional procurement processes often require human sign-off on budgets and vendor selection due to policy, contracts, and financial controls. Ordering systems may require department authorization, creating moderate friction against full automation.
Adoption barriersclaude-sonnet-52/5No licensing requirement dictates who selects materials, though academic freedom norms and departmental approval processes create some institutional friction.
Cost vs. human wageclaude-haiku-4-5-202510015/5AI-driven procurement automation (integrated with institutional systems or commercial APIs) costs a fraction of the human labor required to manually identify, source, and order materials—potentially 10+ times cheaper per transaction.
Cost vs. human wageclaude-sonnet-54/5Using AI to search, compare, and shortlist textbooks/materials is far cheaper than the faculty time spent on manual review, though final purchasing still requires some administrative overhead.
Technical feasibility todayclaude-haiku-4-5-202510014/5Deployed procurement and inventory management systems with AI capabilities exist and perform reliably in higher education and enterprise contexts. Tools can search catalogs, compare prices, and flag items for order, though integration varies by institution.
Technical feasibility todayclaude-sonnet-53/5Tools like AI-assisted curriculum planning and procurement platforms exist, but faculty still typically make final selections manually with limited AI-native product deployment specifically for this niche task.

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

51

CI 34–67 · exposure 50 · augmentation 88 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Higher education is mid-range in AI adoption velocity. Universities are piloting AI-assisted curriculum tools and content generation, but widespread production deployment of fully AI-designed curricula remains limited. Traditional governance structures and faculty hiring models slow deep displacement, though experimentation is accelerating.
Sector adoption velocityclaude-sonnet-52/5Higher education is a comparatively slow-adopting sector for AI-driven curriculum design, with pilots and individual faculty experimentation more common than institutional-scale deployment.
Augmentation potentialclaude-haiku-4-5-202510015/5AI substantially augments faculty curriculum work: generating learning objectives, drafting modules, analyzing student feedback data, and suggesting revisions enables faculty to focus on pedagogical refinement and institutional fit. Faculty remain in the loop for all major decisions while seeing significant productivity gains.
Augmentation potentialclaude-sonnet-54/5AI tools are increasingly useful for brainstorming course content, generating drafts of materials, summarizing literature, and suggesting assessment methods, meaningfully speeding up an instructor's planning and revision work.
Task automatabilityclaude-haiku-4-5-202510014/5AI can draft curricula, generate course materials, evaluate student outcomes via data analysis, and suggest instructional revisions with high fidelity. The task involves structured content creation and analysis where AI systems can produce professional-quality output at significant time savings; however, pedagogical judgment and institutional constraints may require human refinement, preventing a full 5.
Task automatabilityclaude-sonnet-52/5AI can draft syllabi and suggest readings but cannot autonomously plan, evaluate, and revise an entire course curriculum with institutional context, learning outcomes alignment, and accreditation requirements without substantial human oversight.
Adoption barriersclaude-haiku-4-5-202510013/5Academic freedom and institutional quality assurance create moderate friction: faculty unions, accreditation bodies, and institutional tradition favor human-led curriculum decisions. However, no legal mandate requires a human to own curriculum design end-to-end, and shared governance rather than hard licensing requirements apply.
Adoption barriersclaude-sonnet-53/5No licensing requirement mandates a human, but accreditation standards, academic freedom norms, and departmental review processes create meaningful institutional friction against fully automating curriculum decisions.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI inference and content generation costs for curriculum work are orders of magnitude lower than faculty labor costs for the same output volume. Even with integration overhead and human oversight, the cost ratio strongly favors AI, though faculty still perform final judgment and approval.
Cost vs. human wageclaude-sonnet-53/5AI can cheaply generate draft materials, but the human time needed to review, contextualize, and validate curricular decisions keeps overall cost savings moderate rather than dramatic.
Technical feasibility todayclaude-haiku-4-5-202510014/5Multiple deployed products (learning management systems with AI assistants, curriculum design tools, content generation platforms) reliably perform substantial portions of curriculum planning and content creation in higher-ed contexts. Production use is established, though human review remains standard practice rather than exception.
Technical feasibility todayclaude-sonnet-52/5Some AI tools (course design assistants, LMS-integrated content generators) exist but are used as drafting aids rather than reliably performing full curriculum evaluation and revision in production at scale.

Develop and use multimedia course materials and other current technology, such as online courses.

48

CI 37–59 · exposure 42 · augmentation 88 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510013/5Universities are increasingly piloting AI tools for content generation and administrative course prep, but adoption remains mixed and often experimental rather than production-level displacement of the development task.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI tools for course design at a moderate pace—many pilots and growing use of AI authoring tools, but widespread institutional-scale deployment is still uneven across departments.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools demonstrably assist instructors in rapidly drafting lecture notes, generating quiz banks, producing multimedia assets, and organizing course platforms, meaningfully raising productivity while the instructor remains in control of pedagogical decisions and student interaction.
Augmentation potentialclaude-sonnet-55/5AI substantially speeds up drafting slides, quizzes, video scripts, and interactive materials, letting instructors focus on refinement and pedagogical judgment, making it a strong augmentation case.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in generating draft content, multimedia, and course structure, developing *and using* a cohesive online course requires sustained pedagogical judgment, student interaction, and curriculum alignment that AI cannot fully replace end-to-end. Current systems can accelerate parts (drafting lectures, generating graphics) but not deliver the integrated, evolving learning experience with >50% time savings at equal quality.
Task automatabilityclaude-sonnet-53/5AI can generate slides, scripts, quizzes, and draft online course structures, but selecting pedagogically appropriate content, aligning with learning objectives, and finalizing usable multimedia still requires substantial human curation.9 The task is not fully automatable end-to-end at equal quality without instructor oversight.
Adoption barriersclaude-haiku-4-5-202510013/5Institutional accreditation bodies and universities typically require faculty accountability and oversight of course design and delivery, creating friction against full automation. However, no strict legal barrier prevents AI-assisted or AI-generated course content.
Adoption barriersclaude-sonnet-52/5No licensing or legal requirement mandates human-only creation of course materials; institutional norms and academic freedom create mild friction but no hard barrier to AI-assisted content development.
Cost vs. human wageclaude-haiku-4-5-202510012/5Building multimedia materials with AI assistance still requires significant instructor time for curation, validation, and pedagogical redesign. The integrated cost of AI tools, oversight, and rework remains comparable to or higher than hiring a course designer, especially at scale.
Cost vs. human wageclaude-sonnet-54/5Generating draft multimedia content and course materials via AI tools is far cheaper than a faculty member or instructional designer building it manually from scratch, though final review and customization still cost time.
Technical feasibility todayclaude-haiku-4-5-202510013/5AI products exist that generate course materials (text, visuals, quizzes) and assist in online platform setup, but they still require substantial human oversight and editing. Few organizations deploy AI to autonomously create and manage full courses; deployment remains partial and pilot-heavy.
Technical feasibility todayclaude-sonnet-53/5Products like AI slide generators, video creation tools, and LMS-integrated authoring assistants exist and are used by educators, but they produce inconsistent quality and require significant editing before classroom deployment.

Write letters of recommendation for students.

43

CI 30–56 · exposure 38 · augmentation 75 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions remain conservative on AI-authored letters of recommendation due to credibility, legal, and ethical concerns. Adoption remains at pilot or exploratory stage; most postsecondary faculty continue to write letters personally with minimal AI assistance.
Sector adoption velocityclaude-sonnet-54/5Higher education faculty, especially in white-collar academic writing tasks, have rapidly adopted generative AI tools for drafting correspondence and letters since 2023.
Augmentation potentialclaude-haiku-4-5-202510013/5AI could assist by organizing student records, suggesting structure, or drafting neutral opening passages, modestly raising a professor's drafting speed. However, the core task—synthesizing personal knowledge and institutional judgment—remains human-dependent; augmentation is useful but incremental.
Augmentation potentialclaude-sonnet-55/5AI is highly effective as a drafting and editing assistant for recommendation letters, letting the teacher focus on providing accurate personal details while AI handles structure, phrasing, and tone.
Task automatabilityclaude-haiku-4-5-202510012/5Generating reference letters requires detailed knowledge of individual students' academic performance, character, and specific accomplishments—information AI lacks access to. While AI could draft template language, producing credible, substantive letters with required personal detail and institutional weight would require significant human input and fact-checking, falling short of the 50% time-saving threshold at equal quality.
Task automatabilityclaude-sonnet-53/5AI can draft a competent, well-structured letter given input about the student's performance and achievements, but the teacher must supply substantive personal knowledge and verify accuracy, limiting full end-to-end automation.
Adoption barriersclaude-haiku-4-5-202510014/5Recommendation letters carry legal and reputational weight—institutions and students rely on them for admissions, hiring, and funding decisions. A professor's signature and personal credibility are legally and professionally required; institutions would likely prohibit fully AI-authored letters, and liability concerns around false or fabricated claims create strong adoption barriers.
Adoption barriersclaude-sonnet-53/5There's no legal requirement for a human to write the letter, but strong norms around authenticity, personal knowledge, and academic integrity create real friction, and a letter perceived as AI-generated without genuine input could damage credibility and violate institutional trust norms.
Cost vs. human wageclaude-haiku-4-5-202510014/5AI-assisted drafting (if used for initial templates or structure) costs near-zero per attempt, while a professor's time is substantial; however, human review and fact-insertion remain necessary, limiting the cost advantage compared to a professor writing from scratch.
Cost vs. human wageclaude-sonnet-54/5Drafting with an LLM costs pennies versus the substantial time a professor spends writing from scratch, though the human still must review, personalize, and sign off, somewhat reducing the savings ratio.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably writes institutional letters of recommendation at the quality and personalization required for actual academic or employment contexts. AI systems can generate generic templates, but real recommendation letters demand verifiable claims about individuals and carry reputational stakes; current products are research-stage or narrow assistants only.
Technical feasibility todayclaude-sonnet-53/5General-purpose LLMs are widely used today to draft recommendation letters, but no specialized deployed product reliably handles this without significant human input on specifics and personal anecdotes.

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

41

CI 30–51 · exposure 34 · augmentation 75 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Academic institutions move slowly on changing professional development practices. While individual scholars may use AI summaries as a supplement, systemic adoption of AI-driven literature monitoring in postsecondary settings remains limited and cautious.
Sector adoption velocityclaude-sonnet-53/5Higher education and academic research are moderately adopting AI-assisted literature review and summarization tools, though full integration into scholarly practice is still uneven.
Augmentation potentialclaude-haiku-4-5-202510014/5AI excels at assisting this task: summarizing papers, flagging relevant articles, organizing conference programs, and identifying cross-disciplinary trends. Professors using these tools can dramatically reduce time spent on literature review while maintaining critical judgment and collegial engagement.
Augmentation potentialclaude-sonnet-54/5AI significantly speeds up literature discovery, summarization, and synthesis, meaningfully augmenting a professor's ability to keep current even though human engagement (conferences, colleague talks) remains essential.
Task automatabilityclaude-haiku-4-5-202510012/5AI can help summarize current literature and organize conference abstracts, but the social components (talking with colleagues, building professional networks) and judgment about which developments matter require human engagement. The task cannot be fully automated to meet the 50% time-saving threshold.
Task automatabilityclaude-sonnet-52/5AI can help surface and summarize literature but the core task—continuous professional engagement, judgment about relevance, and networking at conferences—requires sustained human involvement and cannot be fully offloaded.
Adoption barriersclaude-haiku-4-5-202510014/5Strong professional and organizational norms require academics to directly engage with the field through reading, collegial exchange, and conference participation. Institutional culture and accreditation expectations create friction against full delegation to automated systems.
Adoption barriersclaude-sonnet-51/5No licensing or regulatory barrier prevents using AI to assist with staying current; it's a personal professional development activity with no sign-off requirement.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI literature review and summarization tools are relatively cheap, but they still require human oversight and don't eliminate the need for conference attendance and colleague interaction. The human wage for staying current is modest relative to AI capability costs when integration and review are included.
Cost vs. human wageclaude-sonnet-53/5AI tools for literature search/summarization are cheap relative to time spent reading, but since only part of the task is addressed, overall cost savings versus the full human task are moderate.
Technical feasibility todayclaude-haiku-4-5-202510013/5Literature summarization and conference tracking tools exist (e.g., arXiv summaries, semantic search), but current products do not reliably curate field developments or replace genuine professional conversations. Material gaps remain in understanding nuanced disciplinary context.
Technical feasibility todayclaude-sonnet-53/5AI literature-summarization tools (e.g., research assistants, citation managers with AI features) are deployed and used by academics today, but they cover only the reading/search portion, not colleague discussion or conference participation.

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

39

CI 25–54 · exposure 38 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Educational institutions, especially postsecondary, are adopting AI slowly for high-stakes assessment. Pilots exist but production deployment of autonomous grading is limited, and faculty remain hesitant to cede grading authority. Adoption velocity in this sector lags that of information-intensive commercial domains.
Sector adoption velocityclaude-sonnet-53/5Higher education is adopting AI grading tools moderately, with pilots and increasing use of AI-assisted feedback tools, though full grading automation is not yet widespread in psychology courses.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist with preliminary grading (flagging suspected plagiarism, scoring objective components, organizing submissions) and giving instructors faster feedback on class-wide patterns, thereby raising productivity on routine administrative aspects while the instructor retains final judgment.
Augmentation potentialclaude-sonnet-54/5AI tools significantly speed up feedback generation, plagiarism detection, and rubric-based scoring, meaningfully augmenting instructor productivity while they remain the grade authority.
Task automatabilityclaude-haiku-4-5-202510012/5Current AI can assist with some grading elements (e.g., factual answer checking, basic rubric application) but cannot reliably evaluate the full range of student work—particularly open-ended responses, conceptual understanding, and nuanced written analysis—at quality parity with expert instructors. Meaningful end-to-end automation with 50% time saving at equal quality is not consistently achievable today.
Task automatabilityclaude-sonnet-53/5AI can grade objective assignments and provide draft feedback on essays, but nuanced grading of original research papers, lab work, and argumentation quality still requires human judgment for a large share of the task.'
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: institutions have established grading policies and academic integrity standards that require instructor judgment; faculty have contractual and professional responsibility for assessment; and there is significant organizational friction around delegating grades to automated systems. Liability and institutional trust in AI grading remain unresolved.
Adoption barriersclaude-sonnet-53/5No licensing requirement blocks AI-assisted grading, but institutional academic integrity policies, accreditation standards, and instructor accountability for final grades create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI grading tools require significant overhead (integration, prompt engineering, human review of outputs, error correction), and the loaded cost of adjunct or teaching assistant labor remains competitive. The cost advantage, if any, is modest and task-dependent.
Cost vs. human wageclaude-sonnet-54/5Once integrated, AI grading assistance is far cheaper per assignment than faculty/TA time, though setup, rubric calibration, and oversight impose some cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5Products like automated essay scoring and assignment checkers exist but have known limitations in accuracy and context sensitivity; they are deployed in narrow, structured settings (e.g., multiple-choice, formulaic responses) rather than the full spectrum of psychology coursework. Production adoption remains light and typically requires heavy human oversight.
Technical feasibility todayclaude-sonnet-53/5AI grading tools (e.g., Gradescope, Turnitin AI feedback) are deployed in production for rubric-based assignments, but reliable grading of open-ended psychology papers with disciplinary nuance is not yet mature at scale.

Write grant proposals to procure external research funding.

32

CI 25–39 · exposure 25 · augmentation 75 · importance 3.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI writing assistance in academic grant writing is slow; most institutions and faculty still rely on human-driven processes with minimal AI integration. The high stakes of funding decisions and conservative academic culture limit rapid deployment.
Sector adoption velocityclaude-sonnet-52/5Academia adopts AI writing tools unevenly and cautiously, with many institutions and funders scrutinizing AI-generated content in proposals, slowing broad production adoption.
Augmentation potentialclaude-haiku-4-5-202510014/5AI tools effectively augment grant writing by drafting initial sections, suggesting improvements to clarity and structure, and generating background literature summaries, meaningfully accelerating the process while the faculty member retains intellectual control and decision-making.
Augmentation potentialclaude-sonnet-54/5AI is widely used to help brainstorm ideas, draft boilerplate sections, improve clarity, and format proposals, meaningfully speeding up the writing process while the researcher retains intellectual ownership.
Task automatabilityclaude-haiku-4-5-202510012/5Grant writing requires synthesizing complex research narratives, strategic framing of intellectual merit, and institutional knowledge about funding priorities that vary widely by agency. While AI can draft sections and improve clarity, the conceptual vision, novelty claims, and persuasive integration of preliminary data demand significant human judgment; current systems cannot reliably produce fundable proposals end-to-end.
Task automatabilityclaude-sonnet-52/5AI can draft sections of a grant proposal but crafting a compelling, fundable proposal requires deep domain expertise, novel research framing, and strategic alignment with funder priorities that current AI cannot fully replicate end-to-end at equal quality.6
Adoption barriersclaude-haiku-4-5-202510014/5Grant proposals must be signed and submitted by the faculty member as principal investigator, and institutional research offices typically require human review and certification. Funders expect authentic institutional commitment and scholarly voice, creating both legal and reputational barriers to full automation.
Adoption barriersclaude-sonnet-52/5No licensing requirement exists, but funders and institutions expect PI authorship, original intellectual contribution, and accountability, creating some friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5A faculty member's time writing or supervising grant proposals is high-wage cognitive labor; while AI reduces some drafting time, the oversight, revision, and strategic decision-making required mean the all-in cost remains close to or exceeds unaugmented human effort for high-quality outcomes.
Cost vs. human wageclaude-sonnet-53/5AI drafting assistance is cheap relative to a professor's or grant writer's time, but the human review, editing, and strategic input still needed keeps overall cost savings moderate rather than dramatic.
Technical feasibility todayclaude-haiku-4-5-202510012/5Tools like ChatGPT can assist with boilerplate sections and editing, but no deployed product reliably generates competitive grant proposals without extensive human revision. Success in grant writing depends on domain expertise, track record framing, and nuanced understanding of reviewer expectations that AI cannot yet replicate consistently.
Technical feasibility todayclaude-sonnet-52/5Products like ChatGPT and specialized grant-writing tools exist and are used for drafting, but no deployed product reliably produces complete, competitive proposals without heavy human revision and subject-matter input.

Review books and journal articles for potential publication.

31

CI 25–37 · exposure 33 · augmentation 63 · importance 2.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Academic publishing is a laggard sector in AI adoption; while some journals use AI for initial triage, the core peer-review function remains human-controlled, and resistance to algorithmic decision-making in scholarship remains high.
Sector adoption velocityclaude-sonnet-52/5Academic publishing is cautious and slow-moving due to integrity concerns, with pilots for AI-assisted screening but limited substantive AI role in judgment-based review at present.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist by summarizing papers, flagging potential issues, and organizing submissions, helping reviewers work more efficiently, though the critical evaluation remains with human faculty.
Augmentation potentialclaude-sonnet-54/5AI substantially helps reviewers by summarizing papers, checking references, flagging statistical errors, and improving review-writing efficiency, while the human retains final judgment.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist with initial screening of abstracts and flagging relevance, but evaluating scholarly merit, theoretical soundness, and contribution to the field requires deep domain expertise and human judgment that AI cannot reliably replicate end-to-end.
Task automatabilityclaude-sonnet-53/5AI can summarize, check for methodological flaws, and flag issues in academic writing, but genuine peer review requires domain judgment, novelty assessment, and contextual expertise that current AI cannot fully replicate end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Academic publishing has strong norms and institutional requirements that publication decisions be made by qualified human experts; liability and reputational risk make algorithmic-only decisions unacceptable to journals and academic institutions.
Adoption barriersclaude-sonnet-54/5Peer review is an institutional and reputational function tied to named expert reviewers and journal policies; using AI without disclosure violates most publisher ethics guidelines, creating strong professional and normative barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-powered manuscript screening tools are becoming cheaper, but full editorial review by AI integrated into academic workflows still costs significantly relative to the value of the decision, and human experts remain essential for oversight.
Cost vs. human wageclaude-sonnet-53/5AI tools for initial screening are cheap, but full review requiring subject-matter judgment still needs paid expert time, making all-in cost roughly comparable once oversight is factored in.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools can summarize and categorize academic papers, no deployed product reliably performs comprehensive peer review or editorial assessment at the quality standard required for publication decisions in academic venues.
Technical feasibility todayclaude-sonnet-52/5Some journals experiment with AI-assisted screening (plagiarism, statistics checks, reference verification) but no deployed product independently performs substantive scholarly peer review at scale in production.

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

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 has been slow to deploy AI advisors at scale; most institutions still rely on human advisors or basic rule-based systems. Adoption remains in the pilot phase with few production implementations, reflecting both cultural resistance and risk aversion.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI slowly for personalized advising; pilots exist (chatbots for registration) but deep integration into career/curriculum advising is still nascent.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist advisors by surfacing relevant curricula, flagging prerequisite conflicts, or summarizing labor-market trends, materially reducing prep time. However, the core judgment task—matching student goals to pathways—remains human-centric, limiting the scope of augmentation.
Augmentation potentialclaude-sonnet-54/5AI can help faculty advisors quickly research program requirements, career pathways, and job market data, meaningfully speeding up the advising process.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can provide generic career information and curriculum overviews, advising students on academic and vocational paths requires understanding individual aptitudes, constraints, and goals—a personalized judgment task. AI cannot reliably match student profiles to outcomes at the depth and nuance human advisors provide, and integration into real student decision flows requires human validation.
Task automatabilityclaude-sonnet-52/5Advising requires understanding an individual student's history, goals, and emotional context, which current AI can support but not fully replace for nuanced, personalized guidance.
Adoption barriersclaude-haiku-4-5-202510014/5Educational advising has implicit pastoral and fiduciary responsibilities; institutions face liability if algorithmic advice harms student outcomes. There is also strong organizational and cultural friction—students, parents, and accreditors expect human contact and accountability in academic planning.
Adoption barriersclaude-sonnet-53/5No formal licensing requirement, but institutional norms, liability concerns about poor guidance, and student preference for human mentorship create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI advisory systems have modest operating costs, but require initial training data, curriculum databases, and oversight infrastructure. For a task currently performed by salaried faculty advisors, the all-in cost of a reliable AI system is unlikely to be cheaper than the marginal cost of advisor time.
Cost vs. human wageclaude-sonnet-52/5While AI advising tools are cheap per query, the oversight and integration needed to match a professor's nuanced judgment keeps costs comparable to human advising for meaningful cases.
Technical feasibility todayclaude-haiku-4-5-202510012/5Chatbots and career-matching tools exist but operate at low fidelity (generic suggestions, limited context awareness). No deployed product reliably replaces a postsecondary advisor's role in understanding a student's full situation, providing accountability, or handling complex multi-year planning.
Technical feasibility todayclaude-sonnet-52/5AI chatbots and advising tools exist in some universities but are typically limited to FAQ-style or scheduling support, not substantive academic/career mentorship.

Participate in student recruitment, registration, and placement activities.

28

CI 25–30 · exposure 25 · augmentation 50 · importance 3.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Educational institutions, especially postsecondary, adopt automation slowly; recruitment and placement remain labor-intensive and relationship-driven. Pilots exist but full displacement is rare; sector digitization lags finance and tech.
Sector adoption velocityclaude-sonnet-52/5Higher education is a moderate-to-slow adopter of AI for these functions; some institutions use AI-driven admissions chatbots, but faculty-level involvement in recruitment/placement remains largely traditional.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist faculty with candidate matching databases, scheduling, follow-up reminders, and analytics on recruitment pipelines, but humans remain essential for interviews, fit assessment, and placement counsel. Assistance is useful but not transformative.
Augmentation potentialclaude-sonnet-53/5AI can help draft recruitment materials, analyze applicant data, and manage administrative correspondence, giving moderate productivity gains while faculty retain decision-making roles.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with email outreach, data entry, and scheduling coordination, the task inherently requires human judgment in evaluating student fit, conducting interviews, and relationship-building that are central to recruitment and placement. Meaningful parts can be automated, but the core interpersonal and advisory elements resist full automation.
Task automatabilityclaude-sonnet-52/5This task combines interpersonal outreach, advising judgment, and institutional decision-making that AI can support but not fully replace; only fragments like scheduling or initial info-gathering could be automated with time savings.
Adoption barriersclaude-haiku-4-5-202510014/5Universities face significant organizational and regulatory barriers: faculty governance expects human advisors, FERPA and educational regulations constrain automated data handling, accreditation bodies value human-led placement, and institutional culture prioritizes personalized student support. Legal liability for placement advice also creates friction.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement, but institutional policy, personal relationships, and accreditation-related human oversight of admissions/placement decisions create moderate friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI infrastructure for recruitment (chatbots, scheduling tools, data processing) incurs setup and maintenance costs comparable to or exceeding the part-time or entry-level labor it might displace in a university context, especially with required human oversight.
Cost vs. human wageclaude-sonnet-52/5While AI can cheaply handle FAQs or scheduling, the faculty judgment components of recruitment and placement still require paid human time, keeping overall cost comparable to human-led processes.
Technical feasibility todayclaude-haiku-4-5-202510012/5Limited deployed systems handle recruitment and placement at scale; most institutions rely on human advisors and CRM tools with minimal AI decision-making. Some universities pilot chatbots for registration queries, but end-to-end recruitment and placement automation remains rare in production.
Technical feasibility todayclaude-sonnet-52/5Chatbots and CRM tools exist for admissions inquiries and basic registration logistics, but faculty involvement in recruitment/placement (e.g., interviewing candidates, advising on fit) isn't performed by deployed AI products today.

Prepare and deliver lectures to undergraduate or graduate students on topics such as abnormal psychology, cognitive processes, and work motivation.

25

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

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education has been slow to adopt AI for core instructional delivery outside of administrative support; most institutions still enforce in-person or synchronous instruction by credentialed faculty. Pilots with AI tutors exist but production-scale replacement is rare.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI unevenly; while some instructors use AI for content creation, actual lecture delivery automation remains rare and pilot-level.
Augmentation potentialclaude-haiku-4-5-202510014/5AI can significantly augment instructors by generating lecture outlines, synthesizing research, drafting exam questions, and providing real-time polling summaries—enabling faculty to focus on facilitation and deeper engagement. Current tools already assist in content preparation and student feedback analysis.
Augmentation potentialclaude-sonnet-54/5AI tools substantially help with drafting lecture outlines, generating examples, creating slides, and summarizing research, meaningfully boosting prep efficiency for instructors.
Task automatabilityclaude-haiku-4-5-202510012/5AI can assist in preparing lecture content (outlines, summaries, slide drafts) but cannot yet replicate the real-time interaction, Socratic questioning, and adaptive responsiveness that characterize effective live teaching. The human instructor remains essential for pedagogical judgment and classroom management.
Task automatabilityclaude-sonnet-52/5AI can help draft lecture content and slides, but live delivery, adapting to student questions, classroom management, and pedagogical presence require human execution not yet replaceable end-to-end.
Adoption barriersclaude-haiku-4-5-202510014/5Postsecondary teaching typically requires an advanced degree and institutional credentialing; many universities have contractual and governance requirements that a qualified human hold the instructor role and maintain direct student engagement. Regulatory and accreditation frameworks expect human instruction.
Adoption barriersclaude-sonnet-54/5Accreditation, university employment structures, and expectations of faculty-led instruction create strong institutional and credentialing barriers to full AI substitution for lecturing.
Cost vs. human wageclaude-haiku-4-5-202510012/5AI-assisted content preparation is inexpensive, but integrating it into a full course delivery system and ensuring quality oversight adds cost. A faculty member's loaded wage remains low relative to the marginal cost of AI tooling when quality control is factored in.
Cost vs. human wageclaude-sonnet-52/5Using AI for prep saves some time cheaply, but full lecture delivery still requires a paid instructor, so overall cost savings versus a professor's salary are modest given AI cannot substitute for delivery.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI tools exist for generating lecture notes and outlines, no deployed product reliably performs the full task of delivering engaging lectures with student interaction, assessment, and dynamic adjustment. Chatbots can answer questions but cannot substitute for a credentialed instructor's presence and authority.
Technical feasibility todayclaude-sonnet-52/5AI tools (e.g., slide generators, script drafting) support lecture prep, but no deployed product autonomously delivers postsecondary lectures reliably in real classrooms today.

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

19

CI 5–32 · exposure 13 · augmentation 63 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academic research in psychology remains a slow-adopter sector. While researchers use AI for writing assistance and literature search, the core research and publishing process has not shifted toward automated production, and cultural norms prioritize human expertise and accountability.
Sector adoption velocityclaude-sonnet-53/5Higher education and academic research have seen growing but uneven AI tool adoption (for writing, analysis, literature review) with pilots and guidelines still evolving rather than deep production-scale integration.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist by helping with literature synthesis, data analysis visualization, manuscript drafting feedback, and citation management, raising researcher productivity on writing and review tasks. However, the originality and judgment components of research remain substantially human-led.
Augmentation potentialclaude-sonnet-54/5AI substantially aids literature reviews, data analysis, drafting manuscripts, and formatting for publication, meaningfully boosting researcher productivity while the scientist retains control over design and conclusions.
Task automatabilityclaude-haiku-4-5-202510011/5Conducting original research and publishing findings requires domain expertise, novel insight generation, experimental design, and creative synthesis—capabilities that current AI cannot replicate at the level expected in peer-reviewed psychology. While AI can assist with literature review and drafting, the core task of generating new knowledge remains fundamentally human.
Task automatabilityclaude-sonnet-52/5AI can assist with literature review, data analysis, and drafting, but designing novel research, running studies (especially with human subjects), and generating genuinely original findings requires human judgment and cannot be end-to-end automated today.
Adoption barriersclaude-haiku-4-5-202510014/5Academic publishing requires human authorship, institutional affiliation, ethical oversight (IRB approval), and professional accountability. Journals expect human researchers to attest to methodology and findings, creating structural and normative barriers to full automation.
Adoption barriersclaude-sonnet-53/5No licensing requirement to publish, but academic norms, peer review, ethical oversight (IRB) for human-subjects research, and authorship/credit conventions create meaningful institutional friction against full automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of a human researcher (salary, benefits, overhead) is justified by their ability to generate novel, credible, publishable work. AI tools remain supportive only; they cannot replace the researcher's output, making the all-in cost of AI-only research prohibitively high relative to human researchers.
Cost vs. human wageclaude-sonnet-52/5AI can cheaply assist with parts of the process (literature synthesis, stats), but the overall research task still requires substantial paid human expert time for design, execution, and interpretation, keeping costs comparable to human-driven work.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably conducts original psychological research or produces publishable findings independently. Current systems cannot design studies, recruit participants, interpret nuanced human behavior, or make novel theoretical contributions at publication quality.
Technical feasibility todayclaude-sonnet-52/5AI writing and analysis tools are used in research workflows, but no deployed product independently conducts psychological research and produces publishable findings reliably; human-led research with AI assistance is the norm.

Initiate, facilitate, and moderate classroom discussions.

18

CI 5–30 · exposure 13 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Academia remains a laggard sector in AI automation; institutions are generally risk-averse about displacing human instruction in core pedagogical roles. Adoption of AI in classroom moderation is negligible; most universities prioritize human-led instruction despite technological feasibility pressure.
Sector adoption velocityclaude-sonnet-52/5Higher education has been slow to adopt AI for live pedagogical interaction, with most current use confined to asynchronous tools like grading or content generation rather than real-time classroom facilitation.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can assist instructors by suggesting discussion prompts, summarizing key points, or flagging engagement patterns, thereby raising productivity in preparation and reflection. However, the augmentation is partial and happens mostly outside the live discussion moment itself.
Augmentation potentialclaude-sonnet-53/5AI can help instructors prepare discussion questions, summarize readings, or suggest prompts beforehand, providing moderate support even though it cannot perform real-time facilitation itself.
Task automatabilityclaude-haiku-4-5-202510011/5Facilitating and moderating live classroom discussions requires real-time social presence, adapting to student contributions, reading room dynamics, and making judgment calls about whose voice to elevate—skills that current AI cannot perform end-to-end in a live educational setting. AI cannot substitute for the human moderator's ability to sense group dynamics and respond authentically.
Task automatabilityclaude-sonnet-52/5Leading a live, adaptive classroom discussion requires reading student engagement, improvising follow-up questions, and managing group dynamics in real time, which current AI cannot reliably do end-to-end.'
Adoption barriersclaude-haiku-4-5-202510014/5Educational accreditation standards, institutional policies, and student expectations strongly favor human instructor presence and judgment in classroom discussion. There is an implicit professional and legal requirement that a qualified human educator facilitate learning interactions, creating organizational and regulatory friction against automation.
Adoption barriersclaude-sonnet-53/5While no license specifically bars AI, universities value human presence and interpersonal mentorship in live instruction, and institutional norms strongly favor faculty-led discussion, creating moderate organizational and cultural friction.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying an AI system capable of reliably moderating live discussions, combined with human oversight and quality assurance, would exceed the cost of the instructor themselves actually performing the task. The instructional value of live moderation makes this difficult to economize.
Cost vs. human wageclaude-sonnet-52/5Since AI cannot substitute for live facilitation, any AI-based alternative would require substantial human oversight or an entirely different format, offering little to no cost savings for this specific task.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI systems can generate discussion prompts and provide suggested responses in asynchronous contexts, no deployed product reliably moderates live classroom discussions with the nuance and responsiveness required in higher education. Some tools assist with asynchronous forum moderation, but real-time facilitation remains beyond reliable automation.
Technical feasibility todayclaude-sonnet-51/5There are no deployed products that autonomously run and moderate live in-person classroom discussions for psychology courses; AI discussion tools are limited to online forums or chatbot Q&A, not live facilitation.

Recruit and hire new faculty.

16

CI 7–25 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Higher education has been slow to adopt AI in hiring, with most institutions still relying on traditional committee-based models. Even tech-forward universities deploy AI primarily in early screening stages; full adoption of automated faculty hiring remains limited and faces cultural resistance within academia.
Sector adoption velocityclaude-sonnet-52/5Higher education is a slow-adopting sector for AI in core administrative/governance functions like hiring, with pilots limited mostly to resume parsing tools rather than decision-making.
Augmentation potentialclaude-haiku-4-5-202510013/5AI tools can meaningfully assist faculty hiring committees by automating resume screening, coordinating schedules, and surfacing publication metrics, improving efficiency in preparatory work. However, the core evaluation of research impact and teaching fit remains primarily a human activity where AI plays a supporting rather than transformative role.
Augmentation potentialclaude-sonnet-53/5AI can help draft job postings, screen initial applications, or summarize CVs, providing moderate assistance while final decisions remain human-driven.
Task automatabilityclaude-haiku-4-5-202510012/5Recruiting and hiring faculty involves nuanced evaluation of research output, teaching philosophy, cultural fit, and interpersonal judgment that requires human oversight. While AI can assist with resume screening and scheduling, the core decision-making and candidate interviews demand experienced faculty evaluators, leaving only marginal automation potential.
Task automatabilityclaude-sonnet-51/5Hiring faculty involves judgment about candidate fit, research quality, interviews, and departmental consensus-building that cannot be executed end-to-end by AI today.
Adoption barriersclaude-haiku-4-5-202510014/5Faculty hiring is heavily regulated by institutional policies, labor law, and equal opportunity requirements, and final decisions typically require sign-off from department heads or hiring committees. University governance structures and the need for human judgment on academic merit create substantial legal and organizational friction against full automation.
Adoption barriersclaude-sonnet-54/5Faculty hiring is governed by university governance structures, tenure committees, EEO/labor regulations, and requires human authorization and accountability, creating strong institutional and legal barriers.
Cost vs. human wageclaude-haiku-4-5-202510012/5Faculty hiring involves significant institutional and legal complexity, requiring careful oversight and potential human expertise throughout. The total cost of AI-assisted recruiting (tools plus institutional integration plus human review) remains comparable to or higher than the cost of streamlined human-led processes, especially for specialized academic evaluation.
Cost vs. human wageclaude-sonnet-51/5The task requires committee deliberation, interviews, and institutional decision-making that AI cannot substitute for, so no meaningful cost savings accrue from AI replacing the process.
Technical feasibility todayclaude-haiku-4-5-202510012/5Current AI systems can support resume parsing and initial filtering with reasonable accuracy, but no mature product reliably performs end-to-end faculty hiring independently. Institutions still rely on human committees to assess research quality, conduct interviews, and make final hiring decisions with meaningful human judgment.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs faculty recruitment and hiring; at most AI tools assist with resume screening or scheduling, not the full task.

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

15

CI 5–25 · exposure 13 · augmentation 50 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education remains a laggard in AI adoption, and the core supervisory relationship is culturally protected; institutions have shown minimal interest in automating faculty mentoring roles, and there is no market signal of rapid displacement in this domain.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools slowly for core mentoring and evaluative functions, though some AI-assisted grading or scheduling tools are creeping into ancillary support tasks.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist faculty by automating scheduling, summarizing student progress notes, drafting feedback templates, and flagging anomalies in research data—improving administrative efficiency while the faculty member retains decision-making authority over mentoring and assessment.
Augmentation potentialclaude-sonnet-53/5AI can help with scheduling, tracking progress, summarizing research drafts, or flagging issues, providing moderate assistance while the human retains full supervisory responsibility.
Task automatabilityclaude-haiku-4-5-202510012/5Supervising research and internship work requires ongoing judgment about student progress, real-time feedback, and dynamic adjustment to individual circumstances. While AI can assist with scheduling and documentation, end-to-end supervision—including mentoring decisions, assessment of conceptual understanding, and interpersonal guidance—remains fundamentally human-dependent and cannot achieve 50% time savings at equal quality.
Task automatabilityclaude-sonnet-51/5Supervision involves relational mentoring, judgment calls on student performance, and accountability that current AI cannot perform end-to-end; no plausible 50% time savings at equal quality exists today.
Adoption barriersclaude-haiku-4-5-202510014/5Strong barriers exist: academic institutions typically require that faculty directly supervise students (institutional policy and accreditation standards), there is significant liability if supervision is delegated to automation, and educational relationships are predicated on human trust and accountability that cannot be outsourced.
Adoption barriersclaude-sonnet-54/5Institutional accreditation, degree-granting authority, and mentorship/evaluation responsibilities require a qualified faculty member; supervisory sign-off is tied to academic credentialing structures.
Cost vs. human wageclaude-haiku-4-5-202510012/5The cost of AI infrastructure, prompt engineering, and human oversight for supervision tasks approaches or exceeds the incremental cost of a faculty member's time, especially given the low-volume, high-stakes nature of academic mentoring.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot perform this supervisory role, there is no viable cost comparison—the human is the only option, making AI effectively infinitely costlier for the actual task.
Technical feasibility todayclaude-haiku-4-5-202510012/5No deployed product reliably performs full supervisory oversight of research and teaching activities. AI can support task tracking and may draft feedback, but production systems do not independently manage the mentoring, evaluation, and adaptive guidance core to this task.
Technical feasibility todayclaude-sonnet-51/5No deployed product supervises student teaching, internships, or research; this remains a purely human academic responsibility with no AI substitute in production.

Provide professional consulting services to government or industry.

13

CI 5–20 · exposure 8 · augmentation 63 · importance 2.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Consulting sectors are slow to automate judgment-heavy, relationship-driven services. Psychology consulting in particular depends on professional licensure, trust, and client relationships—factors that inhibit replacement of human consultants even as lower-touch advisory roles adopt AI.
Sector adoption velocityclaude-sonnet-52/5Higher education and consulting sectors show only moderate AI adoption for advisory work, with pilots for research support more common than full consulting substitution.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can meaningfully assist psychology consultants by generating literature reviews, drafting proposals, analyzing survey data, or synthesizing findings. However, the core consulting act—advising clients and taking responsibility for recommendations—remains human-led, making augmentation substantial but not transformative.
Augmentation potentialclaude-sonnet-54/5AI tools can significantly help consultants by synthesizing literature, drafting reports, analyzing data, and preparing presentations, meaningfully boosting productivity while the professional retains the client relationship.
Task automatabilityclaude-haiku-4-5-202510012/5Professional consulting requires deep domain expertise, client relationship management, and nuanced judgment about organizational/policy contexts. While AI can draft reports or conduct literature reviews, the core service—advising government or industry leaders on complex psychological matters—demands human credibility, accountability, and contextual understanding that current systems cannot reliably provide end-to-end.
Task automatabilityclaude-sonnet-51/5This task requires deep domain expertise, judgment, professional reputation, and interpersonal negotiation applied to unique client problems, which current AI cannot execute end-to-end without substantial human oversight and accountability.
Adoption barriersclaude-haiku-4-5-202510014/5Government and industry clients typically require a licensed or credentialed human consultant for legal liability, regulatory compliance, and professional accountability in sensitive domains (workplace psychology, policy advice, organizational assessment). Clients expect to hold a named expert accountable.
Adoption barriersclaude-sonnet-54/5Professional consulting often requires credentialed expertise, personal liability, and client trust in named individuals, creating strong barriers against full AI substitution even though no formal licensing mandate universally applies.
Cost vs. human wageclaude-haiku-4-5-202510011/5A qualified psychology consultant commands significant hourly fees ($150–400+), often justified by credentials, legal liability coverage, and exclusive expertise. Current AI systems cannot match this value proposition and require human oversight to avoid reputational and legal risk, making them more expensive as a replacement.
Cost vs. human wageclaude-sonnet-52/5While AI can cheaply draft supporting analyses, the actual consulting engagement—requiring expert judgment, liability acceptance, and client relationship management—still requires paying the human consultant, so cost savings are marginal at the task level.
Technical feasibility todayclaude-haiku-4-5-202510011/5No production systems demonstrably perform end-to-end professional consulting services to government or industry. AI tools exist for supporting research and writing, but substituting for a licensed psychologist's consulting authority and liability position is not a deployed capability in real organizations.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently performs professional psychology consulting to government or industry clients; this remains a human expert service requiring credentials and trust.

Collaborate with colleagues to address teaching and research issues.

9

CI 5–13 · exposure 0 · augmentation 38 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Despite high digitization in academia, actual adoption of AI agents in faculty collaboration remains minimal; institutions prioritize human peer review, committee structures, and collegial relationships that resist automation.
Sector adoption velocityclaude-sonnet-52/5Higher education is a relatively slow-adopting sector for AI in core collegial/governance functions, though AI tools are creeping into research and admin support.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could marginally assist by organizing literature, summarizing prior discussions, or suggesting structural solutions, but meaningful academic collaboration fundamentally requires human peers exercising independent judgment and accountability.
Augmentation potentialclaude-sonnet-53/5AI can assist by summarizing research, drafting collaborative documents, scheduling, or synthesizing literature to inform discussions, but the core collaborative interaction remains human-driven.
Task automatabilityclaude-haiku-4-5-202510011/5Collaboration on teaching and research issues requires nuanced interpersonal negotiation, consensus-building, and context-dependent judgment about institutional culture and individual expertise that current AI cannot perform end-to-end. While AI can draft memos or summarize discussions, it cannot meaningfully participate in the collaborative problem-solving that defines this task.
Task automatabilityclaude-sonnet-51/5This is an interpersonal, relational task involving real-time collaboration, negotiation, and shared decision-making among colleagues that cannot be replaced end-to-end by current AI systems.There is no way to substitute the human relationship-building and consensus-forming aspects.
Adoption barriersclaude-haiku-4-5-202510014/5Academic institutions have strong cultural norms around peer collaboration, professional autonomy, and accountability for research direction that create significant friction against AI substitution. Colleagues expect human judgment and shared responsibility for decisions.
Adoption barriersclaude-sonnet-53/5No licensing barrier exists, but strong organizational and cultural norms in academia require human faculty collaboration, peer governance, and shared responsibility for curriculum and research decisions.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task requires human expertise and accountability; any AI involvement would require substantial human oversight, making total cost comparable to or higher than direct human collaboration without meaningful time savings.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this task, so cost comparison favors the human by default since AI cannot replace the function at all.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can reliably conduct genuine peer collaboration on academic issues. Current systems lack the social understanding, institutional knowledge, and autonomous judgment required to be a functional colleague in resolving teaching and research problems.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs collegial collaboration on teaching and research issues; AI tools may support communication but cannot conduct the collaboration itself.

Maintain regularly scheduled office hours to advise and assist students.

6

CI 0–11 · exposure 0 · augmentation 38 · importance 3.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education adopts AI slowly for teaching-adjacent roles; student-facing advising remains an explicit institutional obligation requiring human presence, and there is minimal organizational pressure or evidence of adoption of AI-driven office-hour replacement.
Sector adoption velocityclaude-sonnet-52/5Higher education is adopting AI tools for content and administrative support, but replacing personal advising office hours with AI is not occurring in any meaningful way.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist marginally with scheduling, note-taking, or retrieving course information, but the core value of office hours—human connection, real-time responsiveness, and adaptive mentoring—is not substantively enhanced by current AI tools available to instructors.
Augmentation potentialclaude-sonnet-53/5AI can help prep materials, answer routine student questions beforehand, or summarize common issues, offering moderate assistance around the edges of office hours without transforming the core interpersonal task.
Task automatabilityclaude-haiku-4-5-202510011/5Office hours require real-time, adaptive interpersonal interaction—listening to individual concerns, providing personalized academic and emotional guidance, and building ongoing mentoring relationships. Current AI systems cannot replicate the empathetic presence, contextualized judgment, and rapport-building essential to this task.
Task automatabilityclaude-sonnet-51/5Office hours require live, relational presence with students for personalized academic and career advising, which is not something AI can substitute for in a way that meets the equal-quality bar.
Adoption barriersclaude-haiku-4-5-202510015/5Academic advising and student mentoring carry strong institutional and professional expectations that a qualified human faculty member must provide these services; student welfare concerns and accreditation standards impose hard barriers to full substitution.
Adoption barriersclaude-sonnet-54/5Institutional expectations, accreditation norms, and student support policies typically require faculty to be personally accessible; there's strong organizational and professional expectation for human contact, though not a strict legal licensing requirement.
Cost vs. human wageclaude-haiku-4-5-202510011/5Even if marginal automation of scheduling or documentation were possible, the core task—human presence and adaptive guidance—cannot be cost-effectively replaced; the human instructor's labor remains irreplaceable.
Cost vs. human wageclaude-sonnet-52/5While AI chat tools are cheap, they don't replace the function, so any partial AI substitute would be an add-on cost rather than a true replacement, keeping realistic cost comparisons unfavorable for full substitution.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably substitutes for a professor holding office hours; chatbots can answer FAQ-style questions but cannot conduct the nuanced, confidential, one-on-one advising and emotional support that defines this task.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the role of a professor holding office hours; chatbots can answer factual questions but do not replace scheduled human advising sessions.

Supervise students' laboratory work.

4

CI 0–7 · exposure 0 · augmentation 25 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Educational institutions have not adopted AI for active laboratory supervision because it violates safety and accreditation requirements; adoption remains near-zero.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools for administrative and instructional support but has been slow to change in-person supervisory and safety-critical lab activities.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist by pre-screening student work, generating feedback on lab reports, or monitoring recorded video for flagged anomalies, but the human supervisor remains essential and the actual live-supervision task is not meaningfully augmented.
Augmentation potentialclaude-sonnet-52/5AI can help with lab prep materials, data analysis guidance, or answering procedural questions, but offers limited direct assistance to the act of live supervision itself.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising laboratory work requires real-time observation, safety oversight, and immediate intervention with students—tasks that demand physical presence and nuanced judgment of student competence and lab conditions. Current AI cannot perform these in-person supervisory functions.
Task automatabilityclaude-sonnet-51/5Laboratory supervision requires real-time physical presence, safety oversight, and hands-on judgment about student technique that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Educational and laboratory safety regulations typically require a qualified human supervisor to be physically present and responsible for student safety, experimental integrity, and compliance—a hard legal barrier to full automation.
Adoption barriersclaude-sonnet-54/5Institutional liability, safety regulations for lab environments, and the need for a qualified instructor physically present create strong barriers to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5A faculty member or TA salary is already optimized for multiple responsibilities; AI systems cannot replace the legal and safety accountability required, making any comparison economically irrelevant to actual deployment.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function, so cost comparison favors the human by default since AI cannot deliver the output.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product today can supervise laboratory work end-to-end. While AI can assist with pre-lab instruction or post-lab analysis, the active supervision of students conducting experiments requires human presence and authority.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product supervises psychology students' laboratory work in production; this remains firmly a human physical-presence task.

Act as advisers to student organizations.

4

CI 0–7 · exposure 0 · augmentation 25 · importance 2.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Higher education, especially student advising and mentorship, is a laggard sector for automation. Faculty-student advising remains a core institutional function and cultural norm with no measurable move toward AI substitution.
Sector adoption velocityclaude-sonnet-52/5Higher education adopts AI tools for teaching support but advising roles for student organizations remain untouched by automation trends.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could potentially assist with scheduling logistics or providing information about organization policies, but the core advisory function—listening, motivating, problem-solving with groups—resists augmentation because the relationship and human judgment are central to the task.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, drafting communications, or budgeting for student organizations, but core advisory judgment and relationship work remain largely unassisted.
Task automatabilityclaude-haiku-4-5-202510011/5Advising student organizations requires understanding nuanced group dynamics, mentoring individual students, navigating interpersonal conflicts, and making contextual judgment calls that are deeply human and relationship-based. Current AI systems cannot reliably replicate the empathetic, responsive advisory role or build the trust relationships essential to this task.
Task automatabilityclaude-sonnet-51/5This task requires sustained interpersonal mentorship, relationship-building, and situational judgment with students that cannot be delegated to AI end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Universities have institutional and often legal expectations that faculty (licensed professionals) provide direct advising and pastoral care to students. Student organizations typically require human mentor accountability, and institutions face reputational and liability risks if they delegate this duty to automation.
Adoption barriersclaude-sonnet-54/5Institutions typically require a designated faculty/staff member for liability, oversight, and mentorship purposes, creating strong organizational and quasi-regulatory barriers to substitution.
Cost vs. human wageclaude-haiku-4-5-202510011/5The cost of deploying and overseeing an AI system to handle student advising would far exceed the marginal cost of having an actual faculty member provide guidance, particularly given the need for human oversight and liability concerns.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this function, so cost comparison favors the human by default since AI cannot deliver the output.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs the advisory role for student organizations in any production setting. This task requires ongoing relationship-building, real-time responsiveness to student needs, and institutional accountability that existing systems cannot provide.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the role of a student organization adviser; this is fundamentally a human relational and institutional role.

Provide clinical services to clients, such as assessing psychological problems and conducting psychotherapy.

3

CI 0–6 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption in clinical psychology remains negligible; regulatory, ethical, and liability barriers mean even pilot programs are rare, and the profession resists AI substitution for core clinical tasks.
Sector adoption velocityclaude-sonnet-52/5Healthcare and clinical psychology are historically slow to adopt AI for core clinical decision-making due to regulatory, ethical, and liability constraints, despite growth in adjacent digital mental health tools.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with note-taking, symptom screening forms, or session transcription, but provides minimal productivity gain for the core clinical work of assessment and psychotherapy, which remains human-dependent.
Augmentation potentialclaude-sonnet-53/5AI can assist with intake documentation, symptom tracking, treatment note drafting, and psychoeducational material, giving moderate productivity benefits while the clinician retains full responsibility for assessment and therapy.
Task automatabilityclaude-haiku-4-5-202510011/5Clinical psychological assessment and psychotherapy require genuine human judgment, therapeutic rapport, and real-time responsiveness to client distress. Current AI systems cannot conduct independent psychotherapy or diagnostic assessment meeting professional standards or legal requirements.
Task automatabilityclaude-sonnet-51/5Clinical assessment and psychotherapy require licensed judgment, relationship-building, and legal accountability that current AI cannot replicate end-to-end; no off-the-shelf system meets the 50% time-saving-at-equal-quality bar for actual clinical practice.
Adoption barriersclaude-haiku-4-5-202510015/5Psychology and clinical services are heavily regulated; only licensed psychologists/therapists can legally diagnose and treat mental health conditions, and liability for misdiagnosis or harm is severe and asymmetric.
Adoption barriersclaude-sonnet-55/5Clinical psychology practice is tightly regulated, requiring licensure, ethical codes, and legal accountability for diagnosis and treatment, making human sign-off legally mandatory.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems lack the liability coverage, licensing, and clinical validation required for clinical services, making them effectively more expensive than human clinicians when all compliance and oversight costs are included.
Cost vs. human wageclaude-sonnet-52/5While AI inference is cheap, the necessary human oversight, licensure, and liability coverage for clinical services keep effective costs comparable to or only marginally below a human clinician's loaded wage once compliance is factored in.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs independent clinical assessment or psychotherapy; these tasks require licensed practitioners and remain research-stage for AI. Chatbots may simulate conversation but cannot substitute for clinical care.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs clinical psychological assessment or psychotherapy autonomously in production; AI chatbot mental-health tools exist but are not substitutes for licensed clinical services and carry known safety/liability issues.

Participate in campus and community events.

3

CI 0–5 · exposure 0 · augmentation 25 · importance 2.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Education institutions, particularly postsecondary, remain low-adoption sectors for labor replacement, and faculty participation in community engagement is a non-delegable aspect of the position that organizations do not seek to automate.
Sector adoption velocityclaude-sonnet-51/5Higher education is a slow-adopting sector for AI in general, and this specific interpersonal/presence-based task sees essentially no automation push.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can provide limited assistance such as suggesting relevant events or drafting promotional materials, but offers minimal enhancement to the core act of attending and participating in person.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, event planning logistics, or drafting talking points, but offers minimal assistance to the core act of participating in person.
Task automatabilityclaude-haiku-4-5-202510011/5Participating in campus and community events requires physical presence, social interaction, and real-time relationship-building that current AI systems cannot perform. AI cannot substitute for the human presence and interpersonal engagement that defines meaningful participation.
Task automatabilityclaude-sonnet-51/5Participating in physical campus and community events requires personal presence, social interaction, and relationship-building that AI cannot perform on a person's behalf.'
Adoption barriersclaude-haiku-4-5-202510015/5Strong barriers exist: institutional expectations require faculty presence as representatives, community engagement is integral to faculty roles and often mandated by contract or institutional mission, and participation requires human credibility and relationship-building.
Adoption barriersclaude-sonnet-54/5Community and campus engagement inherently requires human presence, representation, and relationship-building, creating a strong structural barrier to automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5This task is not automatable, making cost comparison moot; the human faculty member must attend and participate directly as part of their role.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute to compare costs against, since the task requires human presence and social engagement.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product can physically attend or meaningfully participate in events. While AI can assist in planning or communication about events, it cannot substitute for actual human participation.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for a person's physical attendance and social participation in events; this is not a task AI products address.

Perform administrative duties, such as serving as department head.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 3.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5This is a human leadership role within university governance structures where AI automation is not being pursued or adopted in any sector.
Sector adoption velocityclaude-sonnet-52/5Higher education administration adopts AI slowly for governance and leadership functions, though it uses AI for support tasks like scheduling and reporting.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist a department head with administrative data compilation, scheduling optimization, or report drafting, but the core leadership, personnel decisions, and institutional representation cannot be augmented into productivity gains that preserve human control.
Augmentation potentialclaude-sonnet-53/5AI can meaningfully assist with scheduling, budget reports, correspondence drafting, and data summarization that support administrative duties, even though it cannot perform the leadership role itself.
Task automatabilityclaude-haiku-4-5-202510011/5Serving as department head requires human judgment, interpersonal negotiation, institutional decision-making, hiring/firing authority, and representation of the department—functions that cannot be meaningfully automated with current AI systems even with significant setup.
Task automatabilityclaude-sonnet-51/5Serving as department head involves relationship management, personnel decisions, strategic planning, and institutional politics that require human judgment and authority AI cannot exercise end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Legal and organizational barriers are absolute: department heads must be tenured faculty with institutional authority, legal liability, and fiduciary responsibility that cannot be transferred to an AI system.
Adoption barriersclaude-sonnet-55/5Department head roles require formal institutional appointment, faculty governance approval, and accountability structures that legally and organizationally require a human occupant.
Cost vs. human wageclaude-haiku-4-5-202510011/5Department head roles carry institutional liability, decision authority, and leadership responsibility that demand human salary. AI cannot fulfill this function, making cost comparison moot.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute performing this role, so no meaningful cost comparison exists; a human must be paid to hold the position regardless of AI tool costs.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs department headship end-to-end; this role fundamentally requires legal authority, accountability, and human presence that cannot be delegated to automation.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product performs departmental headship; at most AI tools assist with scheduling or document drafting as sub-components, not the administrative leadership role itself.

Supervise the clinical work of practicum students.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Psychology education and clinical training operate in heavily regulated, human-centered sectors where direct professional supervision is a legal and accreditation requirement, with minimal automation momentum.
Sector adoption velocityclaude-sonnet-51/5Clinical training and supervision in psychology programs show minimal AI adoption due to licensure, ethics, and accreditation requirements governing supervisory relationships.
Augmentation potentialclaude-haiku-4-5-202510012/5AI could assist with documentation, scheduling, or case note review, but the core task of real-time clinical supervision and judgment requires human expertise and cannot be meaningfully augmented by current systems while maintaining safety and accountability.
Augmentation potentialclaude-sonnet-52/5AI could help with scheduling, documentation review, or summarizing session notes, but offers limited assistance to the core supervisory judgment and mentorship functions.
Task automatabilityclaude-haiku-4-5-202510011/5Supervising clinical practicum students requires real-time observation, judgment of clinical competence, ethical guidance, and responsiveness to novel situations that demand human expertise and accountability. Current AI systems cannot conduct or oversee live clinical interactions in ways that meet professional and licensing standards.
Task automatabilityclaude-sonnet-51/5Clinical supervision requires live observation, ethical judgment, and licensed accountability for student trainee conduct with real clients, which AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Regulatory and licensing requirements are extremely high: supervising clinical practicum is a mandated responsibility of licensed professionals, and liability for student competence and client safety cannot be delegated to AI. Legal and ethical standards require human professional sign-off.
Adoption barriersclaude-sonnet-55/5Clinical supervision of practicum students is typically legally and professionally mandated to be performed by a licensed psychologist, with direct liability for client welfare and student certification.
Cost vs. human wageclaude-haiku-4-5-202510011/5Clinical supervision requires a licensed psychologist or qualified supervisor whose labor cost is high, and AI systems cannot provide equivalent oversight at any price point because the task demands professional accountability and judgment.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute providing equivalent output, so cost comparison favors the human supervisor by default.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product reliably supervises clinical practicum work; this requires licensed professionals to assess student competence, safety, and ethical decision-making in real clinical contexts. AI lacks the professional authority and legal standing to perform this gatekeeping function.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product provides clinical supervision of practicum students; this remains squarely a human, licensed-professional function.

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-4-5-202510011/5Committee governance in academia is deeply rooted in shared governance traditions and legal structures that strongly protect human decision-making authority. Adoption of AI substitution is not occurring.
Sector adoption velocityclaude-sonnet-51/5Academic governance structures are slow-changing and highly resistant to substituting human deliberative roles with AI systems.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with summarizing policy documents, drafting meeting agendas, or preparing analysis of institutional data before a committee meeting, but the deliberative and voting functions remain entirely human-dependent.
Augmentation potentialclaude-sonnet-53/5AI can help draft policy documents, summarize meeting materials, research precedents, and prepare committee members, providing moderate productivity support.
Task automatabilityclaude-haiku-4-5-202510011/5Committee service requires real-time deliberation, interpersonal negotiation, and institutional judgment that depend on human presence, accountability, and decision-making authority. AI cannot legally or meaningfully substitute for a professor's vote or voice in governance.
Task automatabilityclaude-sonnet-51/5Committee service requires human judgment, institutional politics, relationship navigation, and accountable decision-making that AI cannot perform or represent on someone's behalf.
Adoption barriersclaude-haiku-4-5-202510015/5Universities legally require human faculty members to serve on committees and make binding decisions on institutional governance. These tasks cannot be delegated to AI without fundamental changes to institutional authority and liability structures.
Adoption barriersclaude-sonnet-55/5Committee membership typically requires faculty status, institutional authorization, and formal governance roles that only accountable human employees can hold.
Cost vs. human wageclaude-haiku-4-5-202510011/5Committee service is a salaried duty of full-time faculty, not a discretionary task with measurable per-unit costs. AI has no cost advantage because the human presence is non-negotiable.
Cost vs. human wageclaude-sonnet-51/5There is no AI substitute product delivering this output, so cost comparison favors the human by default since AI cannot perform the core task.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product can autonomously serve on committees, participate in deliberations, or cast votes on institutional policies. These tasks require human legal authority and accountability.
Technical feasibility todayclaude-sonnet-51/5No deployed product substitutes for a human serving on and voting/deliberating in institutional committees; this is inherently a human governance role.

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.