Pediatricians, General
29-1221.00Median wage $210,040/yr39,390 employed (US)Rank #790 of 923 scored · top 86% by substitution
Diagnose, treat, and help prevent diseases and injuries in children. May refer patients to specialists for further diagnosis or treatment, as needed.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution15
Exposure15
Augmentation59
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
17 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
0%
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%13
panel mean rating 1.5/5 → substitution pressure 13/100
Technical feasibility todayw 20%19
panel mean rating 1.8/5 → substitution pressure 19/100
Cost vs. human wagew 15%17
panel mean rating 1.7/5 → substitution pressure 17/100
Adoption barriersw 20%inverted — strong barriers lower the score11
panel mean rating 4.6/5 (barrier strength) → substitution pressure 11/100
Sector adoption velocityw 10%22
panel mean rating 1.9/5 → substitution pressure 22/100
Task breakdown (17 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.
Collect, record, and maintain patient information, such as medical history, reports, or examination results.
40CI 30–50 · exposure 42 · augmentation 75 · importance 4.8/5 · click for rater detail
Collect, record, and maintain patient information, such as medical history, reports, or examination results.
40| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains a laggard in AI adoption for core clinical tasks due to regulatory friction, high error-cost asymmetry, and practitioner skepticism; while EHR vendors experiment with automation, deep production adoption of unsupervised documentation collection in pediatrics remains limited. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare adoption of AI documentation tools is growing quickly in some systems but remains uneven across practices, especially smaller pediatric clinics, placing it in the middle of adoption curves. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Scribing AI, voice-to-text, and document summarization tools demonstrably assist pediatricians by reducing typing burden and organizing information, allowing clinicians to focus on patient interaction and clinical judgment while maintaining human oversight of the record. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribes and structured-data extraction tools meaningfully reduce documentation burden and speed up chart maintenance, letting pediatricians spend more time on patient interaction while remaining in the loop for verification. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While EHR systems can auto-populate some structured fields and AI can extract data from documents, the task requires clinical judgment about what to record, verification of patient identity, and integration across multiple sources—critical steps that remain difficult for AI to perform end-to-end reliably without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can transcribe, summarize, and structure clinical notes and history via ambient scribes and EHR integration, but a physician must still verify accuracy and conduct the actual examination, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | HIPAA compliance, requirement for clinician sign-off on medical records, liability for documentation accuracy, and regulatory mandates that licensed providers must attest to the accuracy of patient information create substantial legal and institutional barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | Physicians must legally maintain and attest to medical records, and pediatric care involves consent/privacy protections (HIPAA, minors' records), creating moderate compliance and liability friction even though scribing itself isn't restricted. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Administrative labor to operate and oversee AI-assisted documentation systems (error correction, quality assurance, compliance review) often approaches or matches the cost of direct human data entry, especially for high-stakes pediatric records where errors carry liability. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI scribe/documentation tools cost a fraction of physician time saved, but licensing fees, integration, and required human review keep costs roughly comparable rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | EHR systems and document extraction tools exist in production, but they require significant human correction, verification, and clinical decision-making about what constitutes relevant information; no current system reliably handles the full task autonomously at a pediatric practice scale. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient AI scribes (e.g., Nuance DAX, Abridge) are deployed in real clinics and reliably draft notes, but error correction and human review remain standard, and adoption in pediatrics specifically is narrower than in adult primary care. |
Prepare government or organizational reports of birth, death, and disease statistics, workforce evaluations, or medical status of individuals.
31CI 25–37 · exposure 33 · augmentation 63 · importance 2.8/5 · click for rater detail
Prepare government or organizational reports of birth, death, and disease statistics, workforce evaluations, or medical status of individuals.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pediatric practices and public health agencies have adopted electronic health records and reporting systems slowly, with significant variation. Most report preparation remains manual or semi-automated with human oversight, reflecting low adoption velocity of autonomous reporting. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative and public health reporting systems are adopting AI slowly due to regulatory, privacy, and legacy IT constraints compared to sectors like finance or tech. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by automating data extraction, aggregation, and initial drafting of statistical summaries, reducing manual compilation work. However, clinical interpretation and regulatory compliance checking still require the pediatrician, limiting transformative potential. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can substantially speed up data aggregation, drafting, and formatting of these reports, letting physicians focus on verification and sign-off rather than manual compilation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While data compilation and basic statistical report generation can be partially automated, pediatricians must interpret clinical data, verify accuracy, and apply clinical judgment to produce credible reports. The task requires human validation and clinical context that current AI cannot reliably provide end-to-end. |
| Task automatability | claude-sonnet-5 | 3/5 | Drafting statistical summaries and structured reports from structured data (EHR fields, vital statistics) can largely be automated, but compiling and validating data across sources and ensuring accuracy for official records still requires human review. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Government health reports typically require physician sign-off and carry legal/liability implications. Many jurisdictions require licensed physicians to certify vital statistics and disease surveillance data, creating hard barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Government and organizational reports on birth/death/disease statistics often require physician certification, licensure, and legal accountability, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for data aggregation are inexpensive, but the oversight burden and required human verification by a pediatrician means total cost remains comparable to direct human report preparation. Licensing and compliance review cannot be fully offloaded. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-assisted drafting and data aggregation tools are cheaper than physician time for report compilation, but oversight, data verification, and liability review keep costs from being dramatically lower. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Data extraction and report templating tools exist, but no mature product reliably generates compliant government/organizational health reports without significant human review and manual correction. Clinical sensitivity and regulatory compliance requirements exceed current deployed AI reliability. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some EHR and public health systems auto-generate vital statistics reports, but comprehensive, reliable end-to-end automation of physician-authored reports in production is limited and often narrow in scope. |
Advise patients, parents or guardians, and community members concerning diet, activity, hygiene, and disease prevention.
27CI 25–29 · exposure 25 · augmentation 75 · importance 4.8/5 · click for rater detail
Advise patients, parents or guardians, and community members concerning diet, activity, hygiene, and disease prevention.
27| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains cautious on clinical decision support automation; pediatric practice is particularly conservative due to liability and trust factors. Adoption of AI for direct patient/parent counseling in clinical settings remains minimal despite digitization of other tasks. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatric primary care, has been slower than information/finance sectors to adopt AI for direct patient-facing advisory tasks due to regulatory and safety concerns. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can significantly assist by drafting personalized guidance summaries, retrieving evidence-based resources, and organizing talking points, allowing the pediatrician to focus on dialogue and relationship-building while raising overall efficiency and consistency. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by drafting patient education materials, summarizing guidelines, and answering routine questions, letting physicians focus on personalized counseling and complex cases. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate standard dietary and hygiene guidance, pediatric counseling requires nuanced judgment about individual family circumstances, cultural context, and child development stages. Current systems cannot reliably conduct the interactive, trust-building dialogue needed to ensure parents understand and will comply with advice, achieving less than 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | General health advice can be drafted or summarized by AI, but personalized counseling requires clinical judgment, context on patient history, and real-time interaction that current AI cannot fully replace end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pediatricians are licensed professionals; medical advice—especially to vulnerable populations like children—faces regulatory requirements, liability exposure for errors, and strong cultural/legal expectations that a qualified physician provide or directly oversee clinical guidance to parents. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Medical advice to patients typically requires licensed practitioner involvement due to liability, standard-of-care requirements, and regulatory expectations around pediatric care. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | While AI inference is cheap, the oversight and validation required to ensure safe, personalized pediatric guidance keeps total cost per task comparable to or exceeding a clinician's time, especially when accounting for liability risk. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-generated educational content is cheap to produce, but liability, oversight, and need for tailored clinical advice keep effective cost roughly comparable when accounting for required physician review. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs personalized pediatric health counseling independently. Chatbots can provide generic information, but they lack the contextual assessment, relationship-building, and real-time responsiveness to parental concerns that characterize actual clinical advisory practice. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and patient portals offer general health information, but no deployed product reliably substitutes for a pediatrician's personalized preventive counseling at scale in clinical practice. |
Examine patients or order, perform, and interpret diagnostic tests to obtain information on medical condition and determine diagnosis.
26CI 20–32 · exposure 30 · augmentation 75 · importance 4.9/5 · click for rater detail
Examine patients or order, perform, and interpret diagnostic tests to obtain information on medical condition and determine diagnosis.
26| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare has moderate AI adoption with many pilots in diagnostic imaging and lab interpretation, but actual displacement of physician diagnostic tasks in production is limited; cultural, regulatory, and liability concerns slow deep adoption of autonomous diagnostic systems. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatric primary care, is a historically slow-adopting sector for AI-driven clinical decision tools, with pilots more common than production-scale deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI systems that summarize lab results, flag abnormal imaging findings, and suggest differential diagnoses can meaningfully augment pediatrician productivity, helping them prioritize workload and reduce diagnostic error while the physician retains decision authority. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI decision-support and diagnostic-imaging tools meaningfully augment pediatricians by flagging abnormalities, suggesting differentials, and summarizing test results, improving speed and accuracy while the physician remains the decision-maker. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in interpreting some diagnostic tests (imaging, lab results), pediatricians must physically examine patients, establish rapport with children and families, and synthesize clinical judgment. Current AI cannot independently perform physical exams or achieve the 50% time-saving threshold for the full task. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical examination and much of diagnostic decision-making requires hands-on patient contact, real-time judgment, and integration of nuanced context that current AI cannot perform end-to-end; AI can assist test interpretation but not replace the full task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pediatricians are licensed physicians; legal and regulatory frameworks require a licensed medical doctor to examine patients, order diagnostic tests, and determine diagnoses. Malpractice liability, regulatory compliance, and patient safety requirements create strong hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Physical examination and diagnosis in pediatrics legally requires a licensed physician; strong liability, licensing, and human-contact requirements make full substitution essentially barred. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI diagnostic tools cost significant upfront investment plus ongoing oversight by expensive pediatricians; the all-in cost per diagnosis remains comparable to or higher than direct physician performance, especially when liability and integration costs are factored in. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools for specific diagnostic aids are cheap per query, but the overall task still requires a licensed physician's time for exam and judgment, so cost savings are partial, not order-of-magnitude for the whole task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed AI products exist for narrow diagnostic functions (radiology interpretation, lab analysis), but they require human oversight, work reliably only within specific modalities, and cannot replace the full diagnostic workflow that includes patient interaction and clinical decision-making. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Deployed products exist for narrow diagnostic support (e.g., imaging triage, symptom checkers) but no product performs full patient examination and diagnosis determination reliably in production. |
Plan, implement, or administer health programs or standards in hospitals, businesses, or communities for prevention or treatment of injury or illness.
25CI 25–25 · exposure 25 · augmentation 63 · importance 3.2/5 · click for rater detail
Plan, implement, or administer health programs or standards in hospitals, businesses, or communities for prevention or treatment of injury or illness.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of AI for strategic program planning remains slow and limited; organizations still prefer physician-led committees and expert consensus for major health initiatives, and regulatory conservatism further slows deployment compared to information or finance sectors. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administration is a sector with historically slower AI adoption for high-stakes programmatic decisions, though data-support tools are gradually being piloted. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist pediatricians in this role by synthesizing evidence, modeling scenarios, generating protocol templates, and organizing stakeholder input, raising efficiency in the evidence-gathering and planning phases while the clinician maintains decision authority. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist with epidemiological data analysis, drafting policy documents, benchmarking best practices, and summarizing research to support human-led program design. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with aspects like data analysis, evidence synthesis, and protocol drafting, the task requires significant human judgment, stakeholder engagement, and contextual decision-making that current AI cannot fully replicate. The core work of planning and administering health programs demands understanding organizational constraints, regulatory nuance, and community needs that fall short of the 50% time-saving threshold for full automation. |
| Task automatability | claude-sonnet-5 | 2/5 | Program planning and administration involves stakeholder coordination, contextual judgment, and organizational leadership that current AI cannot execute end-to-end, though it can assist with drafting and data analysis components. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Health program planning and administration in hospitals and clinical settings faces substantial regulatory barriers (accreditation standards, licensure requirements, liability), liability constraints (clinical judgment must remain with credentialed professionals), and organizational friction from resistance to algorithm-driven health decisions. A licensed physician typically must sign off on major program decisions. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Health program administration in clinical/public settings often requires licensed medical leadership, institutional accountability, and regulatory compliance, creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of deploying AI systems with sufficient oversight, validation, and human integration for health program administration likely approaches or exceeds the loaded cost of a pediatrician's time, especially given the need for clinical credibility and error-checking. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply generate reports or summaries, but the actual planning, negotiation, and administration still require costly human oversight and decision-making, keeping overall cost comparable to or only marginally better than human labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI product reliably performs end-to-end health program planning and administration. Current tools can support components (literature review, template generation) but lack the integration, accountability, and context-awareness needed for reliable production use in this complex, high-stakes domain. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | No deployed product autonomously plans or administers institutional health programs; AI tools exist for data analysis or literature synthesis but not for the full administrative/leadership function. |
Teach residents or medical students about pediatric topics.
21CI 16–25 · exposure 17 · augmentation 63 · importance 3.6/5 · click for rater detail
Teach residents or medical students about pediatric topics.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Medical education institutions are beginning to pilot AI for content generation and practice support, but live teaching by AI remains rare and unadopted; adoption is cautious and incremental, not fast or deep. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Medical education is a traditionally slow-adopting sector; while AI tools for case generation and quizzing are being piloted, deep production-level integration into resident teaching remains limited. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by generating reading materials, creating practice problems, or providing instant fact-checking during discussions, but the core mentoring and dynamic teaching remain human-centered with moderate productivity gains from AI tooling. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully augment teaching by generating case vignettes, quiz questions, summaries of guidelines, and personalized study aids, enhancing but not replacing the teaching physician. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Teaching residents and medical students requires real-time dialogue, Socratic questioning, assessment of understanding, and adaptive explanation tailored to learner gaps—tasks that demand human judgment, mentorship, and dynamic interaction that current AI cannot replicate end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can generate lecture content, quizzes, and case studies, but the live, interactive, mentorship-based teaching of residents involving clinical judgment modeling cannot be fully automated at equal quality today.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Medical education is heavily regulated (accreditation, board standards, duty to train the next generation), and residents/students expect and require direct mentorship from experienced physicians; institutional and legal frameworks strongly prefer human-led instruction and certification. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Medical education and accreditation standards (e.g., ACGME) require supervision and sign-off by licensed physician educators, creating strong institutional and regulatory barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integrating AI into teaching workflows (content creation, practice questions) may reduce preparation time, but the core teaching interaction still requires a licensed physician; the economic displacement is marginal, and oversight/quality-control costs add friction. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI content generation is cheap, the overall teaching task still requires substantial physician time for clinical demonstration, feedback, and supervision, keeping costs comparable to human-led instruction. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate lecture notes, answer questions, or create study materials, no deployed system reliably performs the full teaching task (live instruction, feedback, assessment, mentoring relationship building) in medical education settings at production scale. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Tools like ChatGPT or medical education platforms assist with content generation but no deployed product autonomously delivers pediatric resident teaching in production clinical training programs. |
Conduct research to study anatomy and develop or test medications, treatments, or procedures to prevent or control disease or injury.
19CI 14–25 · exposure 25 · augmentation 63 · importance 2.9/5 · click for rater detail
Conduct research to study anatomy and develop or test medications, treatments, or procedures to prevent or control disease or injury.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Research institutions remain conservative and human-centric in their governance; pediatric research particularly prioritizes safety and ethical oversight. Adoption of AI-led research without senior human investigators is negligible in practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Biomedical research is adopting AI tools for specific subtasks (drug discovery, imaging analysis) but broad production deployment of AI for end-to-end pediatric research remains in early/pilot stages given the field's caution and regulatory complexity. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist pediatric researchers with literature synthesis, statistical analysis, data visualization, and protocol optimization, enabling faster hypothesis testing and manuscript preparation while the researcher remains in control of direction and decision-making. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI significantly augments literature review, hypothesis generation, data analysis, and pattern recognition in medical research, meaningfully speeding up parts of the research process while researchers remain essential for experimental design and validation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires designing novel research protocols, managing human subjects, interpreting complex biological data, and making high-stakes methodological decisions that AI cannot perform end-to-end. While AI can assist with literature review and data analysis, the core research direction, hypothesis generation, and experimental design remain dependent on human scientific judgment. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can assist with literature review, data analysis, and hypothesis generation, but designing, conducting, and validating medical research including anatomical study and treatment testing requires hands-on experimentation, clinical judgment, and physical work that current AI cannot perform end-to-end.dimen |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Pediatric medical research is heavily regulated (IRB approval, FDA oversight, informed consent, liability for harm), requires licensed physicians and trained researchers to direct studies, and involves human subjects protection mandates that legally bind human judgment to all critical decisions. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Medical research involving human subjects, drug development, and treatment protocols is heavily regulated (IRB approval, FDA processes, clinical licensure), requiring credentialed physician-researchers to oversee and validate findings. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The infrastructure, regulatory oversight, human principal investigator labor, and specialized equipment required for pediatric research is extremely expensive. AI tools add cost rather than reducing it, since humans must direct, validate, and be responsible for all research outcomes. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can reduce costs for literature review and data analysis portions, but the overall research process still requires expensive lab work, clinical trials, and expert oversight, keeping total costs comparable to or only modestly below human-led research. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No AI system today can independently conduct medical research, recruit subjects, oversee clinical trials, or develop treatments. AI can support components (literature mining, statistical analysis) but deployed products do not reliably execute the full research task in production medical settings. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools (e.g., literature synthesis, drug discovery platforms like AlphaFold) exist and are used by researchers, but no deployed product autonomously conducts full research studies or clinical trials for pediatric treatments today. |
Explain procedures and discuss test results or prescribed treatments with patients and parents or guardians.
18CI 7–29 · exposure 17 · augmentation 63 · importance 4.8/5 · click for rater detail
Explain procedures and discuss test results or prescribed treatments with patients and parents or guardians.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of AI for core clinical communication is slow and cautious; regulators, liability concerns, and professional norms prioritize human clinicians. While some EHR vendors offer documentation drafting, actual delegated explanation to AI is rare and typically supplementary only. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatric primary care, has been slow to adopt AI for direct patient communication due to regulatory, trust, and safety concerns, with most use confined to administrative support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by drafting explanation templates, summarizing test results, preparing talking points, or generating written summaries for parents to take home. However, the live explanation and dialogue remain the clinician's responsibility, so augmentation is helpful but not transformative of the core interpersonal task. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can help physicians draft clear explanations, summarize test results, and prepare patient-friendly language, meaningfully speeding up preparation while the physician still delivers and personalizes the conversation. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Explaining procedures and discussing test results requires nuanced, empathetic communication adapted to individual patient/guardian comprehension levels, medical literacy, and emotional context. Current AI cannot reliably perform the full two-way dialogue, address follow-up questions, build trust, or handle the interpersonal judgment required for informed consent in a clinical setting. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can draft explanations of procedures/results, but real-time, personalized dialogue with worried parents involving trust, emotional nuance, and clinical judgment cannot be fully automated end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Regulatory and legal barriers are high: informed consent and informed assent require a licensed physician to explain risks/benefits and answer questions directly. Malpractice liability, duty of care, and child protection statutes mandate clinician-patient communication; an AI system cannot substitute or sign off without the physician present. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Communicating diagnoses and treatment plans is generally considered part of the physician's licensed scope of practice, with malpractice and informed-consent implications requiring a qualified human. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | A pediatrician's explanation is bundled into their consultation visit; automating only the explanatory component would require human oversight anyway (for liability and quality assurance), negating cost savings. The human clinician's judgment and presence remain legally and medically necessary. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI-generated explanations are cheap to produce, but the physician's time is still required for the actual conversation and liability, so overall cost savings are moderate rather than dramatic. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can generate explanation templates or draft educational materials, no deployed product reliably handles the full task end-to-end: real-time conversational explanation with appropriate adaptation, reading emotional cues, responding to patient questions, and documenting informed consent. Some telehealth and documentation tools assist but do not replace the clinician. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Chatbots and patient portals provide general explanations, but no deployed product reliably conducts full clinical result-discussions with parents in place of a physician at scale. |
Monitor patients' conditions and progress and reevaluate treatments as necessary.
18CI 11–25 · exposure 22 · augmentation 75 · importance 4.6/5 · click for rater detail
Monitor patients' conditions and progress and reevaluate treatments as necessary.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare remains a slow-adopting sector for AI autonomy, constrained by regulation, liability, and clinician skepticism. While EHR alerting and risk prediction tools exist, clinical practices have not widely deployed AI agents that independently reevaluate pediatric treatments; adoption remains at the pilot and optimization stage. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatrics, adopts AI slowly due to regulation, liability, and the need for human judgment despite growing use of clinical decision support tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI systems already assist pediatricians by summarizing patient progress, highlighting abnormal lab trends, suggesting differential diagnoses, and automating documentation—materially raising efficiency in the monitoring loop. These tools enhance the physician's ability to track and reassess patients faster while maintaining clinical oversight. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist by flagging trends, summarizing patient data, and suggesting differential considerations, helping physicians monitor and adjust care more efficiently. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in tracking vital signs, lab results, and flagging clinical anomalies, the task fundamentally requires human judgment to interpret subtle clinical changes, integrate patient history, and make treatment modifications—especially in pediatrics where developmental factors matter. End-to-end autonomous monitoring and reevaluation remains beyond current AI capability. |
| Task automatability | claude-sonnet-5 | 1/5 | Longitudinal clinical monitoring and treatment reevaluation require physical exams, judgment under uncertainty, and accountability that current AI cannot autonomously perform end-to-end for pediatric patients. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Pediatricians must legally and professionally sign off on treatment changes and clinical decisions. Medical regulations, malpractice liability, and the standard of care require a licensed physician to evaluate and authorize therapeutic modifications. These are hard barriers to autonomous replacement. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Licensed physician oversight and sign-off is legally required for treatment decisions in pediatric care, creating a hard regulatory and liability barrier. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Clinical monitoring support systems (integration, data management, oversight) cost thousands annually per patient, while a pediatrician visit or longitudinal follow-up costs on the order of hundreds per encounter. AI reduces some documentation burden but does not achieve cost parity with human-led monitoring on a per-task basis. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI can cheaply process data streams, but the human physician's judgment, liability, and exam remain necessary, so overall cost savings for the full task are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed EHR systems and clinical decision-support tools can monitor objective metrics and alert physicians to out-of-range values, but they operate in a narrow advisory scope. Real-world pediatric monitoring still depends on the physician to synthesize information, examine the patient, and adjust therapy—no mature product autonomously reevaluates treatment. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Products exist for flagging abnormal labs, vitals trends, or decision support alerts, but no deployed system independently monitors and reevaluates pediatric treatment plans reliably at scale. |
Refer patient to medical specialist or other practitioner when necessary.
16CI 6–25 · exposure 13 · augmentation 50 · importance 4.4/5 · click for rater detail
Refer patient to medical specialist or other practitioner when necessary.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pediatric practice remains highly traditional with slower digital adoption compared to finance or IT. AI-driven referral automation has not achieved meaningful production adoption in clinical settings; most adoption remains at pilot or advisory-tool stage rather than autonomous decision-making. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall adopts AI decision-support cautiously and slowly relative to other professional-services sectors, with clinical decision-making tasks especially lagging. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can augment by flagging patients with symptoms matching specialist profiles, suggesting relevant specialists, and automating scheduling logistics, meaningfully reducing time spent on research and coordination. However, the core clinical judgment remains with the pediatrician, making this a moderate augmentation use case. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help by surfacing relevant specialists, summarizing patient history, or flagging red-flag symptoms that warrant referral, aiding but not replacing the pediatrician's judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI systems can help identify when specialist referral may be needed by analyzing patient history and symptoms, but the clinical judgment, patient communication, and coordination required to decide *which* specialist and *when* to refer demand human expertise. Partial automation of information gathering and specialist identification is possible, but end-to-end autonomous referral decisions remain unreliable. |
| Task automatability | claude-sonnet-5 | 1/5 | Referral requires clinical judgment integrating patient history, exam findings, and evolving symptoms to decide if/when/where to refer, which is not something current AI can execute end-to-end reliably. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Pediatricians must exercise clinical judgment and accountability for referral decisions; liability and standard-of-care norms require a licensed physician to make and document the referral decision. Regulatory and professional barriers prevent autonomous AI substitution, and patient/family preference typically favors physician-directed specialist routing. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Referral decisions are a core physician responsibility carrying legal and licensing accountability; a licensed practitioner must make and sign off on this decision. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration of AI screening tools into referral workflows involves infrastructure, training, and oversight costs. The task itself—a brief but high-stakes clinical judgment—has relatively low absolute cost when performed by a pediatrician, making AI cost savings marginal relative to the physician's labor. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply suggest referral candidates, but the physician's licensed judgment and liability exposure remain necessary, so realized cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Clinical decision support tools exist to flag conditions warranting specialist input, but no deployed product reliably performs autonomous referral decisions at scale in production pediatric practice. Systems require substantial human review and cannot yet independently determine appropriateness and timing without error. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously makes specialist referral decisions in production; clinical decision support tools flag possibilities but do not replace physician referral judgment. |
Provide consulting services to other physicians.
16CI 11–20 · exposure 17 · augmentation 63 · importance 3.5/5 · click for rater detail
Provide consulting services to other physicians.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While healthcare is digitizing, physician consultation remains a high-trust, relationship-based service; adoption of AI-driven consulting (without human sign-off) is minimal. Most AI use in clinical settings remains supportive rather than autonomous in peer consultation contexts. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall adopts AI cautiously due to regulatory, liability, and safety concerns, with pilots more common than full production deployment in consulting contexts. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by organizing case summaries, suggesting diagnostic considerations, or highlighting relevant literature, helping a pediatrician prepare or refine a consultation. However, the consultation itself—judgment, communication, accountability—remains inherently human. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools like clinical decision support and literature synthesis significantly help physicians prepare for and inform consultations, improving speed and breadth of information review. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Consulting services require nuanced medical judgment, synthesis of complex patient cases, and peer-to-peer professional communication that demand human expertise and accountability. Current AI cannot reliably provide specialized medical advice that other physicians would accept as a substitute for human consultation. |
| Task automatability | claude-sonnet-5 | 2/5 | Physician-to-physician consulting requires synthesizing nuanced clinical judgment, liability, and contextual patient knowledge that current AI cannot fully replicate end-to-end.atterns While AI can support differential diagnosis suggestions, the actual consulting role including accountability and relationship remains human-driven. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Consulting services are legally and professionally bound to licensed physicians; liability for medical advice rests on the consulting physician's shoulders, and peer institutions typically expect human-to-human professional exchange. Regulatory requirements and professional standards create hard barriers to full substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medical consulting requires licensure, malpractice liability, and legal accountability, making it a hard-barrier task where only licensed physicians can formally consult and sign off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted tools are cheaper than consulting a human pediatrician, but consulting services require high-stakes expertise and human accountability, so the cost comparison is asymmetric—organizations still need to pay human physicians to validate AI suggestions. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI tools can cheaply generate suggestions, but the oversight, liability review, and integration into consulting workflows keep effective costs closer to comparable with physician time given the specialized expertise required. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist in summarizing case information or suggesting differential diagnoses, no deployed product reliably performs physician-to-physician consultation independently. Products that support clinical decision-making exist but require human physicians to validate and take responsibility for advice. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Clinical decision support tools exist and are used to assist physicians, but no deployed product independently provides physician-to-physician consulting services reliably at scale. |
Plan and execute medical care programs to aid in the mental and physical growth and development of children and adolescents.
5CI 3–7 · exposure 5 · augmentation 63 · importance 4.4/5 · click for rater detail
Plan and execute medical care programs to aid in the mental and physical growth and development of children and adolescents.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Pediatrics remains a primarily human-directed specialty with slow AI adoption. While electronic health records and some documentation support exists, actual care program planning and execution automation is minimal; most adoption remains in pilot phases or narrow administrative tasks. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatric primary care, has been slower than information/finance sectors to deploy AI at the point of care due to regulatory, safety, and trust constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist pediatricians by synthesizing growth charts, flagging developmental red flags, retrieving evidence-based guidelines, and automating documentation, thereby freeing cognitive load for clinical judgment. However, the augmentation is partial rather than transformative given that core planning remains physician-directed. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI tools can meaningfully assist with growth chart analysis, developmental screening flagging, documentation, and literature review, enhancing physician efficiency while the physician remains fully in control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires integrated clinical judgment, personalized assessment of individual child development trajectories, and decision-making that accounts for complex medical and psychosocial factors. Current AI cannot autonomously plan and execute comprehensive care programs that meet the threshold of 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This task involves longitudinal clinical judgment, physical examination, individualized treatment planning, and ongoing relationship-based care that current AI cannot execute end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Pediatricians must be licensed physicians who hold legal and ethical responsibility for patient care plans. Regulatory frameworks (state medical boards, malpractice liability, informed consent requirements) mandate that a qualified physician must author and sign off on care programs, creating insurmountable barriers to autonomous substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medical licensure, malpractice liability, and legal requirements mandate a licensed physician to diagnose, plan, and oversee treatment for minors, creating hard regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The integration, clinical oversight, and liability management costs for AI-assisted pediatric care planning remain high relative to physician salaries, particularly given the personalized nature of child development monitoring and the need for human supervision and accountability. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Given AI cannot perform the task independently, any comparison would require full human oversight, making AI substitution more costly than the human clinician for the complete task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with literature review, differential diagnosis support, and documentation, no deployed product can autonomously plan and execute full pediatric care programs. Existing clinical decision support tools operate in narrow domains and require substantial physician oversight and final decision-making. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously plans and executes comprehensive pediatric care programs; AI is at most a decision-support or documentation aid within physician-led workflows. |
Prescribe or administer treatment, therapy, medication, vaccination, and other specialized medical care to treat or prevent illness, disease, or injury in infants and children.
1CI 0–3 · exposure 0 · augmentation 50 · importance 5.0/5 · click for rater detail
Prescribe or administer treatment, therapy, medication, vaccination, and other specialized medical care to treat or prevent illness, disease, or injury in infants and children.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI to fully prescribe or administer treatment to pediatric patients is negligible because regulatory and legal requirements prohibit substitution; AI adoption in pediatrics remains limited to decision-support and documentation roles. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Pediatric clinical care adopts AI slowly for diagnostics/documentation, but core prescribing and hands-on treatment remain untouched by automation trends. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist pediatricians by suggesting evidence-based treatment options, flagging drug interactions, and summarizing patient data, improving efficiency in medication selection and decision-making. However, augmentation is constrained to advisory functions while the physician retains full clinical authority and responsibility. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with dosage calculations, treatment guideline lookup, and clinical decision support, improving efficiency while the physician retains full control and responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Prescribing and administering medical treatment to infants and children requires integrating complex clinical judgment, patient history, physical examination, and individual risk factors that current AI cannot reliably perform end-to-end. Medical decisions in pediatrics demand accountability and real-time adaptation to patient response that exceeds current AI capabilities. |
| Task automatability | claude-sonnet-5 | 1/5 | Prescribing and administering treatment requires physical presence, hands-on care, clinical judgment for a vulnerable population, and legal accountability that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task has near-complete regulatory and legal barriers: prescribing medication and administering medical care to children is restricted by law to licensed physicians. Medical boards, state licensing, DEA regulations for controlled substances, and liability frameworks all mandate human physician accountability. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Prescribing medication and administering care to minors requires licensure, controlled-substance authority, malpractice liability, and legal responsibility that only a credentialed physician can hold. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI currently cannot substitute for the physician's role; the cost comparison is moot because the task cannot be automated. Any AI support requires physician oversight, meaning the human cost remains largely intact. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot perform the physical administration or bear liability, so there is no substitutable cost comparison—human physicians remain mandatory. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product today performs independent prescription or treatment administration for pediatric patients at clinical scale. While AI assists in diagnostic support and drug interaction checking, the full decision and execution loop remains physician-dependent and legally requires licensed clinicians. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently prescribes, administers, or delivers vaccinations/treatment to children; AI exists only as decision-support, not as an autonomous clinical actor. |
Examine children regularly to assess their growth and development.
1CI 0–3 · exposure 0 · augmentation 38 · importance 4.9/5 · click for rater detail
Examine children regularly to assess their growth and development.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare automation in this domain remains limited to record-keeping and decision support, not clinical examination replacement. Physical pediatric examination is inherently resistant to automation due to regulatory and ethical constraints. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare overall has lagged in adopting autonomous AI for hands-on clinical tasks, though administrative and diagnostic support tools are growing; direct physical assessment remains untouched. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist marginally by analyzing growth charts, flagging developmental milestones to review, or organizing assessment data, but the core examination and judgment remain entirely human-dependent with limited augmentation potential. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by auto-plotting growth curves, flagging developmental delay patterns from structured data, and generating documentation, improving efficiency around the exam even though it can't replace it. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Examining children requires hands-on physical assessment, measurement, observation of behavior, and real-time clinical judgment that current AI cannot perform in-person. While AI could assist with data analysis post-examination, it cannot conduct the examination itself, which is the core of this task. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical examination requiring hands-on assessment of growth, developmental milestones, and physical findings cannot be performed end-to-end by current AI; it requires physical presence and tactile/visual clinical judgment in a live encounter. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legal and regulatory frameworks require a licensed physician to perform pediatric examinations and sign off on developmental assessments. Medical liability, child safety requirements, and professional licensing create hard barriers to any substitute for human clinical evaluation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Pediatric physical exams require a licensed physician or nurse practitioner by law and standard of care, with direct liability for missed diagnoses, making this a hard-barrier task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot yet perform this task end-to-end, making cost comparison premature. The human pediatrician's expertise is irreplaceable for the actual examination phase, making AI cost-prohibitive for task replacement. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the physical exam at all, so there is no viable cost comparison—human labor remains mandatory for this component of care. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can conduct physical examinations of children or reliably assess growth and development in real-time without a licensed clinician present. Remote assessment tools exist but require human interpretation and are not substitutes for direct clinical evaluation. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical pediatric exams autonomously; AI is used only for adjacent tasks like documentation or growth chart plotting, not the exam itself. |
Treat children who have minor illnesses, acute and chronic health problems, and growth and development concerns.
1CI 0–3 · exposure 0 · augmentation 50 · importance 4.9/5 · click for rater detail
Treat children who have minor illnesses, acute and chronic health problems, and growth and development concerns.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare automation, especially in primary care and pediatrics, faces organizational, regulatory, and liability friction. While AI tools for documentation and decision support are emerging, actual replacement of the pediatrician-patient clinical encounter is negligible in deployed practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially pediatric primary care, has been slow to adopt AI for actual clinical treatment due to regulatory, safety, and liability constraints, though administrative AI use is growing. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing patient history, flagging relevant differential diagnoses, retrieving evidence-based guidelines, or supporting documentation; these tools are beginning to enter clinical practice. However, augmentation remains limited because the physician must still perform the examination and make judgment calls on child-specific factors. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist pediatricians with differential diagnosis suggestions, growth chart analysis, documentation, and literature lookup, improving efficiency while the physician retains full clinical responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical examination, direct patient interaction, clinical judgment in diagnosis, and medical decision-making that cannot be fully automated. While AI can assist with symptom interpretation or literature review, the hands-on evaluation and treatment of children—especially involving physical examination, medication adjustment, and developmental assessment—cannot be performed end-to-end by current AI systems. |
| Task automatability | claude-sonnet-5 | 1/5 | Direct treatment of children requires physical examination, hands-on care, and clinical judgment under uncertainty that current AI cannot execute end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Pediatric care is heavily regulated; only licensed physicians (with pediatric training or general competence in pediatrics) can legally diagnose and prescribe treatment for children. Legal liability, malpractice risk, licensure requirements, and the requirement for a physician to examine and sign off on care create hard barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Practicing medicine on children requires a licensed physician, with strict liability, malpractice exposure, and regulatory requirements mandating human accountability for diagnosis and treatment. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI infrastructure, oversight, and integration to support pediatric assessment does not reduce costs compared to direct pediatrician time. A pediatrician's professional service is necessary; AI cannot replace the billable clinical encounter at a lower cost per patient visit. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot perform the actual treatment task, there is no substitutive cost comparison—human physician cost remains necessary regardless of AI tool costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product reliably performs comprehensive pediatric diagnosis and treatment independent of a licensed physician. Clinical decision support tools exist, but they require human physicians to conduct examinations, interpret findings, and make final treatment decisions; they do not substitute for the core task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently treats pediatric patients; AI tools exist only as clinician-facing decision support or documentation aids, not autonomous treatment providers. |
Direct and coordinate activities of nurses, students, assistants, specialists, therapists, and other medical staff.
1CI 0–3 · exposure 0 · augmentation 38 · importance 4.6/5 · click for rater detail
Direct and coordinate activities of nurses, students, assistants, specialists, therapists, and other medical staff.
1| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of AI for staff direction is negligible because legal authority and accountability cannot be delegated to AI; physicians retain and must exercise this role regardless of sector digitization. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative adoption of AI is growing but clinical leadership and staff coordination remain largely untouched by automation in practice. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI can assist with scheduling suggestions or flagging resource conflicts, but the core coordinative and directive judgment remains inherently human and physician-driven, limiting meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with scheduling, task tracking, and communication support among care teams, moderately easing coordination burden without replacing the coordinating role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Directing and coordinating medical staff requires real-time interpersonal judgment, conflict resolution, and accountability for clinical outcomes that current AI cannot perform end-to-end. AI lacks the authority, contextual understanding, and human accountability necessary to replace a physician's coordination role. |
| Task automatability | claude-sonnet-5 | 1/5 | Directing and coordinating a multidisciplinary clinical team requires real-time judgment, authority, and interpersonal leadership that current AI cannot substitute for end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers: only a licensed physician can direct and coordinate clinical staff and is legally liable for outcomes of coordinated care. Substitution by AI is prohibited by medical practice law. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Clinical supervision and delegation of care duties are legally and professionally restricted to licensed physicians, creating hard regulatory and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems, integration, and required human oversight would far exceed the value of automation, since a licensed physician must ultimately own the coordination decision and team accountability. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this managerial task, so cost comparison favors the human physician entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs this task in production. AI scheduling tools exist but do not substitute for the directive and coordinative authority a pediatrician exercises over multidisciplinary teams. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product manages or directs clinical staff activities; AI scheduling/communication tools exist but do not perform the supervisory coordination role itself. |
Operate on patients to remove, repair, or improve functioning of diseased or injured body parts and systems.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.3/5 · click for rater detail
Operate on patients to remove, repair, or improve functioning of diseased or injured body parts and systems.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of surgical robots remains limited to high-resource centers and specific procedure types; the vast majority of pediatric surgery is still performed by human surgeons with traditional instruments, reflecting slow and narrow adoption. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Surgical practice remains highly physical and human-performed; adoption of autonomous AI in actual surgical execution is negligible and not on a fast trajectory. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Surgical robots can assist with precision and visualization, but current systems offer limited augmentation beyond tool enhancement; they do not meaningfully amplify surgeon capability in real-time decision-making or adaptation during complex pediatric procedures. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI and robotic-assisted surgical systems, imaging analysis, and pre-op planning tools provide meaningful assistance to surgeons, though the physical operation itself remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Surgical operation requires hands-on physical intervention, real-time adaptation to anatomical variation, and immediate judgment in response to unexpected findings—capabilities far beyond current AI. No meaningful part of this task can be automated end-to-end today. |
| Task automatability | claude-sonnet-5 | 1/5 | Surgical operation on patients requires physical manipulation, real-time judgment, and dexterity that no current AI system can perform end-to-end; this is a hands-on physical task, not information processing.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Surgery is one of the most highly regulated and legally-protected medical acts; only licensed physicians can perform surgery, and malpractice liability, informed consent, and surgical credentials create strong legal and organizational barriers to any autonomous or delegated automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Surgery requires licensed physicians with malpractice liability, hospital credentialing, and legal accountability, making autonomous AI substitution essentially prohibited. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Surgical automation technology, where it exists, requires expensive hardware, extensive training, and integration into sterile operating-room environments. The all-in cost far exceeds the cost of a surgeon's time for equivalent surgical outcome. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task independently, so cost comparison favors the human surgeon entirely; robotic assistance adds cost rather than reducing it. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | While surgical robots exist (e.g., da Vinci), they function as teleoperated tools requiring a human surgeon in control; they do not perform surgery autonomously. No deployed product performs this task without direct human operative control. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product autonomously performs pediatric surgery; surgical robots (e.g., da Vinci) are human-controlled tools, not autonomous operators. |
Related occupations — Healthcare Practitioners & Technical
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.