Clinical Neuropsychologists
19-3039.03Median wage $110,840/yr18,820 employed (US)Rank #796 of 923 scored · top 86% by substitution
Assess and diagnose patients with neurobehavioral problems related to acquired or developmental disorders of the nervous system, such as neurodegenerative disorders, traumatic brain injury, seizure disorders, and learning disabilities. Recommend treatment after diagnosis, such as therapy, medication, or surgery. Assist with evaluation before and after neurosurgical procedures, such as deep brain stimulation.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution15
Exposure14
Augmentation54
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
18 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%16
panel mean rating 1.7/5 → substitution pressure 16/100
Cost vs. human wagew 15%15
panel mean rating 1.6/5 → substitution pressure 15/100
Adoption barriersw 20%inverted — strong barriers lower the score13
panel mean rating 4.5/5 (barrier strength) → substitution pressure 13/100
Sector adoption velocityw 10%20
panel mean rating 1.8/5 → substitution pressure 20/100
Task breakdown (18 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.
Read current literature, talk with colleagues, and participate in professional organizations or conferences to keep abreast of developments in neuropsychology.
36CI 16–55 · exposure 30 · augmentation 63 · importance 4.4/5 · click for rater detail
Read current literature, talk with colleagues, and participate in professional organizations or conferences to keep abreast of developments in neuropsychology.
36| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Clinical neuropsychologists work in healthcare and academic settings with moderate digitization; adoption of AI for professional development is still in the pilot and exploratory phase, not yet mainstream in production workflows. |
| Sector adoption velocity | claude-sonnet-5 | 3/5 | Healthcare and academic medicine show moderate AI adoption for research assistance and literature review, though full integration into CME/professional development workflows is still emerging. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by curating and summarizing literature, identifying relevant papers, and organizing conference schedules, but the clinician must still actively evaluate and engage; these tools enhance but do not transform the full learning process. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI substantially augments the literature-reading component by quickly summarizing and flagging relevant new research, meaningfully boosting efficiency while the human still engages with colleagues and conferences directly. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task is fundamentally about staying current through active engagement—reading literature selectively, interpersonal dialogue with colleagues, and participation in professional networks. While AI can summarize papers or fetch conference agendas, the core work of active learning, discernment of relevance, and relationship-building cannot be automated end-to-end with time savings. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can summarize literature, surface relevant papers, and synthesize research trends, saving substantial reading time, but networking with colleagues and conference participation involve social/professional engagement AI cannot replace. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Professional development and staying current are ethical and professional obligations embedded in licensure and continuing education requirements; there is also an inherent expectation that clinicians maintain personal engagement with the field through conferences and peer exchange. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement dictates how a professional must stay current, though professional organizations often require personal participation for CE credits and networking, creating mild friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for literature search and summarization cost a fraction of a clinician's time, but the task itself is discretionary professional development rather than a high-volume operational service, so cost comparison is not the binding constraint. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI literature review tools are cheap relative to a neuropsychologist's time for the reading portion, but the task also includes in-person professional activities where AI offers no cost substitution, keeping overall ratio moderate. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Deployed products can help with literature retrieval and summarization (e.g., semantic search, abstract generation), but no system reliably performs the full task of professional development—which requires judgment about which sources matter, meaningful colleague interaction, and active conference participation—in production. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Deployed tools like literature summarization assistants and research search engines (e.g., Elicit, PubMed AI tools) reliably help with the reading component, but no product handles the collegial/conference networking aspect. |
Write or prepare detailed clinical neuropsychological reports, using data from psychological or neuropsychological tests, self-report measures, rating scales, direct observations, or interviews.
31CI 25–37 · exposure 33 · augmentation 63 · importance 5.0/5 · click for rater detail
Write or prepare detailed clinical neuropsychological reports, using data from psychological or neuropsychological tests, self-report measures, rating scales, direct observations, or interviews.
31| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Clinical neuropsychology remains a high-touch, regulated discipline with slow digital transformation. Adoption of AI tools has been limited to narrow assistive functions (formatting, data entry support), not replacement of report writing, reflecting professional skepticism and regulatory constraints. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized psychological assessment, has been slower to adopt generative AI tools compared to sectors like finance or general professional services, with pilots more common than production use. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by organizing test data, flagging outliers, and generating preliminary text frameworks that a neuropsychologist then refines, improving efficiency modestly. However, the core cognitive work—synthesizing findings into clinical interpretation—remains heavily human-dependent, limiting transformative augmentation potential. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully speed up drafting, summarizing test data, and organizing narrative sections, allowing the neuropsychologist to focus on interpretation and clinical judgment while remaining the final author of record. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can draft text from test data and assist with organizing information, clinical neuropsychological reports require integrating complex behavioral observations, clinical judgment about cognitive deficits, and narrative synthesis that current systems struggle to perform reliably. The high stakes of diagnostic conclusions and need for clinician interpretation across multiple data sources means AI cannot meet the ≥50% time-saving bar end-to-end without substantial human rework. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can draft substantial portions of a neuropsychological report by synthesizing test scores, observations, and templated language, but integrating nuanced clinical judgment across multiple data sources and producing a defensible diagnostic narrative still requires significant human review and correction. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical neuropsychological reports must be authored or signed off by a licensed neuropsychologist; licensure and liability law create a hard requirement for human professional accountability. Additionally, insurance and medical-legal contexts demand human professional judgment and signature, creating strong regulatory and legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Neuropsychological reports require interpretation and sign-off by a licensed psychologist for clinical, legal, and insurance purposes, creating strong regulatory and liability barriers against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for report generation still require substantial clinician review, correction, and refinement, limiting cost savings. The loaded wage of a clinical neuropsychologist is high, and AI overhead for oversight and safety checks remains significant relative to the labor savings achieved. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | AI can cut drafting time substantially and inference costs are low, but the need for licensed clinician review, data integration, and liability oversight keeps the effective all-in cost only moderately below the clinician's own time cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed clinical product reliably generates a complete, defensible neuropsychological report from raw test data. Existing AI tools can assist with formatting or preliminary summaries, but clinical neuropsychologists do not trust systems to independently produce the nuanced diagnostic integration this task requires, particularly given liability and licensure implications. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some EHR-integrated AI drafting tools and LLM-based scribes exist for clinical documentation, but there are no widely deployed, validated products specifically for full neuropsychological report generation from raw test batteries used reliably in production clinics today. |
Establish neurobehavioral baseline measures for monitoring progressive cerebral disease or recovery.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.1/5 · click for rater detail
Establish neurobehavioral baseline measures for monitoring progressive cerebral disease or recovery.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Neuropsychology remains a human-centered specialty with slow AI adoption; most clinical practices still rely on traditional paper-and-pencil or minimally digitized assessment, and there is no evidence of rapid production-level AI displacement in this domain. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized clinical assessment, adopts AI tools slowly due to regulatory, liability, and validation requirements, with pilots more common than widespread production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by automating score calculations, flagging data quality issues, and summarizing prior baseline data, moderately improving clinician efficiency, but the core task of clinical judgment in test selection and interpretation requires human expertise throughout. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI-assisted scoring, normative comparisons, and automated report drafting can meaningfully speed up test administration and interpretation while the neuropsychologist retains clinical judgment and sign-off. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help score and analyze standardized neuropsychological test results, establishing baseline measures requires clinical judgment about test selection, patient capacity to participate, and interpretation within individual context—tasks requiring human expertise and real-time adaptive decision-making that current AI systems cannot reliably perform end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Establishing baseline measures requires selecting/administering standardized tests, clinical interviewing, and interpreting results in context of patient history, which involves nuanced professional judgment AI cannot fully replicate; only scoring/data aggregation portions are automatable today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Neuropsychological baseline establishment must be performed or directly supervised by a licensed clinical neuropsychologist to be legally and professionally defensible; results inform critical clinical decisions about disease progression and treatment, creating strong liability and regulatory barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical neuropsychological assessment is typically restricted to licensed psychologists/neuropsychologists for diagnostic and legal/insurance purposes, creating strong regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI tools for neuropsychological assessment support require significant setup, clinician validation, and oversight; the all-in cost remains comparable to or exceeds the time a neuropsychologist would spend, particularly when factoring integration and quality assurance. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While automated scoring software is cheap, the overall task still requires a highly trained clinician's time for administration, interpretation, and documentation, so all-in cost savings versus human labor are modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI-assisted scoring and data analysis tools exist in research and some clinical settings, but no deployed product reliably performs the full task of baseline establishment (including test selection, patient assessment, and clinical synthesis) without substantial clinician oversight and decision-making. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some digital neuropsych testing platforms exist that score and normalize results, but no deployed AI product independently establishes valid clinical baselines without a licensed neuropsychologist's administration and interpretation. |
Compare patients' progress before and after pharmacologic, surgical, or behavioral interventions.
25CI 25–25 · exposure 25 · augmentation 63 · importance 4.0/5 · click for rater detail
Compare patients' progress before and after pharmacologic, surgical, or behavioral interventions.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare and clinical psychology sectors adopt AI slowly relative to information and finance; neuropsychology clinics remain largely traditional, with tools like EHR integration and reporting aids in early adoption, not widespread production deployment for autonomous interpretation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and clinical psychology have historically been slower to adopt AI diagnostic tools broadly due to regulatory, liability, and workflow integration challenges, with pilots more common than production deployment.' |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by automating data compilation, flagging outliers, computing effect sizes, and generating preliminary statistical summaries, allowing neuropsychologists to focus on clinical interpretation and contextual analysis. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by aggregating test results, flagging significant score changes, and drafting comparative summaries, improving efficiency while the neuropsychologist retains interpretive responsibility.' |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Comparing test scores, cognitive measures, and behavioral metrics before and after intervention is partially automatable (data aggregation, statistical analysis), but the clinical interpretation requires expert judgment about confounds, measurement reliability, and clinical significance that AI cannot reliably perform independently. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help organize and summarize test scores and progress notes, but the clinical interpretation of neuropsychological change (distinguishing practice effects, disease progression, treatment response) requires expert judgment that current systems cannot reliably replicate end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical neuropsychologists are licensed professionals whose scope of practice includes test interpretation and clinical judgment; professional liability, diagnostic responsibility, and regulatory standards require a human expert's sign-off on conclusions about intervention efficacy. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical judgments about treatment efficacy and patient status typically require a licensed neuropsychologist's sign-off, given liability, diagnostic significance, and regulatory expectations in healthcare documentation.' |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools for data analysis and statistical reporting are inexpensive, but they serve only a small portion of the task (computation), while the core clinical judgment remains expensive human work, yielding modest overall cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI-assisted scoring or summarization tools are cheap to run, but the overall task still requires substantial expert oversight and interpretation, keeping total cost comparable to or only modestly below human-only performance.' |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI systems can compute statistical comparisons and flag changes in numerical data, but no deployed product reliably performs the full clinical interpretation and integration of neuropsychological test batteries in the way a licensed neuropsychologist would, especially with complex patient histories. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | No deployed product autonomously compares pre/post-intervention neuropsychological profiles and renders clinical conclusions; existing tools are limited to score aggregation and basic statistical change indices used as clinician aids.' |
Participate in educational programs, in-service training, or workshops to remain current in methods and techniques.
21CI 16–25 · exposure 17 · augmentation 63 · importance 4.0/5 · click for rater detail
Participate in educational programs, in-service training, or workshops to remain current in methods and techniques.
21| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare and clinical professions adopt AI slowly for routine tasks; mandatory in-person continuing education is a regulatory and professional norm with limited digitization, and AI has no role in actual participation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare/psychology sectors are slower adopters of AI for compliance-driven professional development activities compared to tech-forward fields. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing key concepts from educational materials, generating study guides, or helping organize learned techniques, thereby moderately enhancing a clinician's ability to retain and synthesize information from training. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can help identify relevant courses, summarize research, generate study aids, and reinforce learning between sessions, meaningfully supporting the human's ongoing education. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Participation in educational programs requires human engagement, discussion, reflection, and professional judgment to evaluate relevance to clinical practice. AI cannot meaningfully substitute for the interactive learning and skill development inherent in this task. |
| Task automatability | claude-sonnet-5 | 2/5 | Attending and absorbing training content requires genuine human learning and professional engagement; AI can curate materials but cannot 'participate' on behalf of the person.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Professional licensure and continuing education requirements typically mandate human participation in accredited programs; many states legally require documented personal attendance at workshops for license renewal and competency maintenance. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Licensing boards typically require documented personal attendance and completion of continuing education hours, making delegation to AI largely disallowed. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI tools might reduce prep time for educational materials, but the core task—actually participating and learning—requires human time investment. The cost savings are marginal compared to the clinician's loaded wage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Since AI cannot substitute for the human's actual participation, there's no comparable cost basis for full task replacement, though search/summarization tools are cheap. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can help curate or summarize educational content, no deployed product can autonomously participate in live training sessions, workshops, or hands-on skill development activities that require professional presence and active engagement. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI tools exist to summarize CE content or recommend courses, but no product actually completes required participation or in-service training for a licensed professional. |
Provide education or counseling to individuals and families.
20CI 11–29 · exposure 13 · augmentation 50 · importance 4.7/5 · click for rater detail
Provide education or counseling to individuals and families.
20| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Clinical neuropsychology is a specialized field with slow digital transformation. While some practices use educational videos and telehealth, autonomous or primary AI counseling is rare; uptake remains limited to pilot adjuncts in academic or large medical centers. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and behavioral health sectors are adopting AI tools for documentation and triage, but direct patient/family counseling by AI remains rare and cautiously piloted. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by drafting patient-friendly educational summaries, generating custom counseling scripts, or flagging key discussion points, thereby saving preparation time and standardizing content. However, the clinician must still deliver, adapt, and respond to individual and family needs in real time. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help neuropsychologists prepare educational materials, summarize test results in plain language, and draft counseling talking points, meaningfully aiding preparation even though delivery remains human. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate educational materials or draft counseling talking points, the task requires responsive, empathetic interaction with vulnerable individuals and families dealing with neurological conditions. Current AI systems lack the real-time adaptive capacity, emotional attunement, and crisis-response judgment needed for meaningful counseling, and cannot reliably replace a human clinician's presence. |
| Task automatability | claude-sonnet-5 | 1/5 | Providing clinical education and counseling to patients and families requires empathetic, adaptive, contextual human interaction and clinical judgment that current AI cannot deliver end-to-end at equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical neuropsychologists are licensed professionals; counseling and education touch on mental health care, which carries regulatory requirements, duty-of-care liability, and informed-consent expectations. Families and patients typically expect a qualified human clinician, and organizational risk management limits autonomous AI deployment. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical counseling typically requires licensed professional judgment, informed consent processes, and liability considerations that create strong barriers to full AI substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | AI-driven psychoeducational content (chatbots, interactive modules) costs less than clinician time per interaction, but integration, clinical oversight, and liability management narrow the savings. The cost advantage is modest and offset by need for human backup. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI-generated educational materials are cheap, they cannot substitute for the counseling interaction itself, so cost comparison for the full task favors the human significantly. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Chatbots and AI tutors exist for generic health education, but deployed products do not reliably perform clinical counseling or family education in neuropsychological contexts at production scale. Clinicians remain the primary delivery mechanism; AI tools are at most supplementary in real practice. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently conducts clinical neuropsychology patient/family counseling in production; at most chatbots offer generic psychoeducation, not this specific clinical task. |
Interview patients to obtain comprehensive medical histories.
19CI 14–25 · exposure 17 · augmentation 63 · importance 5.0/5 · click for rater detail
Interview patients to obtain comprehensive medical histories.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Adoption of autonomous AI for medical history-taking in clinical neuropsychology is minimal; the field remains human-centered due to regulatory constraints, patient expectations, and the high cost of error, with AI mainly in pilot or assistive roles rather than production replacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized clinical assessment fields like neuropsychology, has been slower and more cautious in adopting AI for direct patient interaction compared to information/finance sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by drafting interview summaries, flagging potential medical contradictions, or suggesting follow-up questions based on patient responses, raising clinician efficiency in documentation and review, though the core interview itself remains human-led. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribes, pre-visit questionnaires, and summarization tools can meaningfully streamline documentation and prompt clinicians with relevant follow-up questions, enhancing efficiency while the clinician retains control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Obtaining comprehensive medical histories requires nuanced, adaptive questioning that responds to patient context, emotional state, and emerging clinical patterns. Current AI cannot reliably conduct the full semi-structured interview with the requisite flexibility and judgment needed to probe inconsistencies or emotional cues at equal quality to human clinicians. |
| Task automatability | claude-sonnet-5 | 2/5 | While AI can transcribe and summarize interviews, eliciting a comprehensive medical history requires adaptive follow-up questioning, rapport-building, and clinical judgment about symptom significance that current systems cannot reliably replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong regulatory and liability barriers exist: neuropsychologists are licensed professionals whose clinical judgment on history-taking is legally and ethically their responsibility; patients and institutions expect human contact for sensitive mental health and neurological histories, and malpractice liability for missed information falls on the clinician, not the AI vendor. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical history-taking for neuropsychological evaluation typically requires a licensed professional for liability, diagnostic accuracy, and patient trust reasons, creating strong regulatory and professional barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI systems capable of medical interviewing remain limited and expensive to integrate into clinical workflows; the cost of inference, validation, and necessary clinician oversight makes them comparable to or more expensive than direct human interview for this high-stakes task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI transcription/summarization tools are cheap, but since a licensed clinician must still conduct or verify the interview, overall cost savings versus the human-led process are modest. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While chatbots can conduct templated Q&A, no deployed clinical product reliably performs comprehensive medical history-taking as a standalone clinical tool; systems that exist are narrow, require heavy human supervision, and lack validation for clinical-grade history completeness and accuracy. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some intake chatbots and ambient scribe tools exist and are used for structured symptom checklists, but no deployed product independently conducts nuanced neuropsychological history-taking in production at scale. |
Conduct neuropsychological evaluations such as assessments of intelligence, academic ability, attention, concentration, sensorimotor function, language, learning, and memory.
14CI 3–25 · exposure 13 · augmentation 50 · importance 5.0/5 · click for rater detail
Conduct neuropsychological evaluations such as assessments of intelligence, academic ability, attention, concentration, sensorimotor function, language, learning, and memory.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Clinical neuropsychology is a specialized, regulated field with slow digital transformation relative to other health sectors. Adoption of AI assistive tools is limited; most practices still rely on in-person administration and paper-based or legacy scoring systems rather than AI-driven workflows. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and clinical psychology adopt AI slowly due to regulatory, liability, and validation requirements, with pilots for scoring assistance but no broad deployment for full evaluations. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist clinicians by automating test scoring, flagging patterns, generating interpretive summaries, and providing normative comparisons, allowing the neuropsychologist to focus on behavioral observation and clinical synthesis. This modest augmentation aids efficiency but does not transform the core diagnostic process, which remains human-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with test scoring, report drafting, and data pattern analysis, offering moderate productivity gains while the neuropsychologist still performs and interprets the assessment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Neuropsychological evaluations require synthesis of complex test administration, behavioral observation, and clinical judgment integrated with patient history. Current AI can score some standardized tests and suggest interpretations, but cannot reliably conduct the full evaluation including test selection, adaptive administration based on patient presentation, and nuanced clinical synthesis needed to meet 50% time-savings-at-equal-quality threshold. |
| Task automatability | claude-sonnet-5 | 1/5 | Neuropsychological evaluations require live administration of standardized tests, direct behavioral observation, and clinical interpretation of subtle patient responses that current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical neuropsychological evaluations are heavily regulated: only licensed psychologists or neuropsychologists can legally conduct these assessments and be held liable for diagnostic and treatment recommendations. Medical/legal liability, insurance requirements, and patient care standards create substantial legal barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a licensed clinical activity requiring a doctoral-level psychologist's judgment, legal liability, and often in-person testing protocols mandated by professional and regulatory standards. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Integration costs for AI systems within neuropsych workflows, combined with required clinician oversight and potential liability exposure, make the all-in AI cost comparable to or higher than conducting evaluations directly with licensed neuropsychologists, especially given the need for validation and quality assurance. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot perform the core hands-on assessment, there is no viable AI substitute cost to compare; the human specialist remains the only option for delivering the service. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI tools can assist with scoring and interpretation of specific neuropsych tests, no deployed product performs end-to-end neuropsychological evaluations reliably. Clinical neuropsychologists remain gatekeepers due to the need for real-time behavioral observation, test adaptation, and accountability for diagnostic accuracy in high-stakes clinical decisions. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product administers or conducts full neuropsychological batteries in production; existing tools are limited to scoring support or research-stage pattern analysis. |
Identify and communicate risks associated with specific neurological surgical procedures, such as epilepsy surgery.
14CI 3–25 · exposure 13 · augmentation 50 · importance 3.8/5 · click for rater detail
Identify and communicate risks associated with specific neurological surgical procedures, such as epilepsy surgery.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare AI adoption remains cautious and slow in high-stakes surgical contexts; neuropsychology and surgical risk communication are specialty practices with low digital maturity and strong professional gatekeeping, showing pilot interest but minimal production displacement. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized neuropsychology and surgical risk counseling, has slow, cautious AI adoption due to regulatory, liability, and high-stakes accuracy concerns. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by aggregating surgical outcome literature, calculating population-level risk statistics, and organizing complication taxonomies, enabling neuropsychologists to spend more time on individual patient framing and shared decision-making; augmentation is useful but partial. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help synthesize research literature, summarize patient records, or draft risk communication materials, aiding the neuropsychologist's preparation without replacing the judgment or delivery. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI can assist in synthesizing published surgical risk literature and generating risk summaries, but cannot independently evaluate patient-specific anatomical complexities, comorbidities, or intraoperative contingencies that define actual procedural risk. Clinical judgment linking neuroimaging, functional status, and individual surgical anatomy remains irreducibly human. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires integrating patient-specific neuropsychological testing, medical history, and surgical risk factors into a nuanced clinical judgment and communicating it empathetically; no AI system can perform this end-to-end today with equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: clinical neuropsychologists hold licensure and are held legally responsible for risk disclosure quality; informed consent documentation ties directly to professional liability; medical practice regulations require a licensed provider to own risk communication and obtain informed consent. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This involves licensed medical/psychological judgment, informed consent processes, and legal liability for surgical risk communication, requiring a credentialed professional by law and standard of care. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI inference cost is low, but integration into surgical workflows and the necessary human expert review for accuracy and liability mean total cost per reliable risk communication remains comparable to or higher than direct neuropsychologist time. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute product, so cost comparison favors the human specialist entirely; any AI use would only supplement, not replace, at added cost for oversight. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed product reliably performs independent surgical risk identification and communication for neurological cases; existing AI tools offer literature retrieval or risk calculation aids but require expert oversight and cannot replace the nuanced clinical assessment required for this task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs pre-surgical risk assessment and communication for neurosurgery patients; this remains firmly within specialist clinical practice. |
Distinguish between psychogenic and neurogenic syndromes, two or more suspected etiologies of cerebral dysfunction, or between disorders involving complex seizures.
12CI 3–21 · exposure 13 · augmentation 50 · importance 4.7/5 · click for rater detail
Distinguish between psychogenic and neurogenic syndromes, two or more suspected etiologies of cerebral dysfunction, or between disorders involving complex seizures.
12| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Clinical neuropsychology remains a high-touch, specialty field with slow digitization. While academic medical centers experiment with AI decision support, meaningful production adoption of autonomous diagnostic AI in this domain is minimal, and cultural and regulatory inertia slows uptake. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare diagnostics adopt AI cautiously due to regulation and liability; neuropsychological differential diagnosis specifically sees minimal AI deployment currently. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can usefully assist by organizing and pattern-matching neuropsychological test results, flagging relevant literature, and organizing imaging findings, thereby raising clinician efficiency in the information-gathering phase. However, the final differential reasoning and clinical judgment remain human-led. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by analyzing test data patterns, flagging inconsistencies, summarizing literature on etiologies, or supporting documentation, aiding but not replacing clinical reasoning. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task requires integrating complex clinical history, neuropsychological testing patterns, imaging data, and differential diagnosis reasoning. While AI can assist with pattern recognition in test results and literature synthesis, current systems cannot reliably perform the full differential diagnostic reasoning end-to-end without substantial human oversight, and achieving 50% time savings with equal quality is not demonstrated at scale. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires integrating clinical interviews, test batteries, imaging, and nuanced judgment about etiology, which current AI cannot perform end-to-end at equal quality with major time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical diagnosis of seizure disorders and neuropsychological differential diagnosis are legally and professionally bounded to licensed neuropsychologists and physicians. Liability for misdiagnosis, regulatory requirements (medical practice acts), and professional oversight requirements create strong barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Diagnosis of complex neurological/psychiatric conditions legally requires a licensed neuropsychologist or physician, with high liability for misdiagnosis. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The infrastructure, integration, regulatory oversight, and human review required to deploy AI for neuropsychological differential diagnosis makes the all-in cost comparable to or higher than a trained clinical neuropsychologist's time, especially given the high stakes of diagnostic error. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the diagnostic process, so any AI cost would be additive to, not replacing, the specialist's fee, making it more expensive not cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed clinical product reliably performs independent differential diagnosis of psychogenic versus neurogenic syndromes or complex seizure etiology without neuropsychologist supervision. AI tools exist for image analysis and literature retrieval, but the synthesis and clinical judgment required for this distinction remain research-stage or narrow-scope. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs differential diagnosis of psychogenic vs neurogenic syndromes reliably; this remains a specialist clinical judgment task, not automated in practice. |
Design or implement rehabilitation plans for patients with cognitive dysfunction.
10CI 0–20 · exposure 13 · augmentation 50 · importance 3.8/5 · click for rater detail
Design or implement rehabilitation plans for patients with cognitive dysfunction.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Clinical neuropsychology is a low-digitization, human-intensive specialty where rehabilitation planning remains fundamentally relational and individualized. Adoption of AI for core clinical decisions is minimal; practices remain bound by licensure and professional standards that prioritize human oversight. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized clinical neuropsychology, adopts AI slowly due to regulatory, liability, and workflow integration barriers, with pilots more common than production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by summarizing test results, suggesting evidence-based interventions from literature, or organizing patient data, but current tools offer only marginal productivity gains. The task's demand for real-time clinical judgment and patient-specific adaptation limits the scope of meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can meaningfully assist by suggesting evidence-based exercises, tracking patient progress data, and drafting plan documentation, boosting clinician efficiency while the clinician retains design authority. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Designing and implementing rehabilitation plans requires individualized clinical judgment, synthesis of complex neuropsychological test data, patient history, and adaptive tailoring based on therapeutic response—capabilities far beyond current AI systems. No general-purpose AI can reliably perform the full cycle of plan design and adaptive implementation at quality and speed comparable to a qualified neuropsychologist. |
| Task automatability | claude-sonnet-5 | 2/5 | Designing rehabilitation plans requires integrating clinical assessment data, patient-specific goals, and judgment about neuroplasticity and functional prognosis that current AI cannot reliably synthesize end-to-end.It could draft template plans but not autonomously design or implement individualized clinical interventions. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Clinical rehabilitation plan design and implementation is typically a licensed neuropsychologist function; legal, regulatory, and malpractice frameworks require qualified human professionals to take responsibility for patient outcomes. Medical liability and the cognitive vulnerability of the target population create hard barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Designing and implementing treatment plans for cognitive dysfunction is a licensed clinical activity with direct liability and legal scope-of-practice requirements mandating a qualified neuropsychologist's sign-off. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The marginal cost of AI assistance (if available at all) remains low relative to specialist neuropsychologists, but the high liability, specialized expertise requirement, and need for human oversight mean AI cannot achieve cost-per-outcome parity with human clinicians on this safety-critical task. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | AI could cheaply generate draft exercise protocols, but the clinical judgment, assessment interpretation, and patient interaction required still necessitate a licensed neuropsychologist, keeping overall costs comparable to human-led care. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with literature retrieval and generate template suggestions, no deployed product reliably designs or implements personalized cognitive rehabilitation plans end-to-end. Existing systems lack the clinical reasoning, real-time patient assessment integration, and accountability required for independent execution in clinical settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently designs or implements neuropsychological rehabilitation plans in clinical practice; existing cognitive rehab software is used as a tool under clinician direction, not as an autonomous planner. |
Diagnose and treat conditions involving injury to the central nervous system, such as cerebrovascular accidents, neoplasms, infectious or inflammatory diseases, degenerative diseases, head traumas, demyelinating diseases, and various forms of dementing illnesses.
9CI 3–16 · exposure 13 · augmentation 63 · importance 5.0/5 · click for rater detail
Diagnose and treat conditions involving injury to the central nervous system, such as cerebrovascular accidents, neoplasms, infectious or inflammatory diseases, degenerative diseases, head traumas, demyelinating diseases, and various forms of dementing illnesses.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare sectors show cautious, slower AI adoption than information/finance; neuropsychology in particular remains resistant due to clinical judgment requirements, liability concerns, and low digitization of some assessment modalities. Pilots exist but production displacement is minimal. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized neurological diagnosis and treatment, adopts AI cautiously due to regulatory, liability, and safety concerns, with most use confined to pilot decision-support tools. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI substantially assists neuropsychologists through automated neuroimaging analysis, test scoring/interpretation dashboards, literature search, and pattern detection across large datasets. These tools meaningfully raise clinician productivity and diagnostic confidence while the neuropsychologist retains decision authority. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with literature review, imaging pattern recognition, cognitive test scoring, and differential diagnosis suggestions, improving efficiency while the clinician retains full responsibility for diagnosis and treatment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with pattern recognition in imaging and analysis of test batteries, the task requires integration of complex patient history, neurological examination, functional assessment, and clinical judgment to formulate a diagnosis and treatment plan. Current AI cannot reliably perform the full diagnostic and treatment workflow end-to-end at quality parity with neuropsychologists. |
| Task automatability | claude-sonnet-5 | 1/5 | Diagnosing and treating complex CNS conditions requires integrating clinical exam, neuroimaging, patient history, and hands-on assessment plus treatment planning that AI cannot perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Diagnosis and treatment of neurological conditions are legally restricted to licensed physicians and credentialed clinical neuropsychologists in most jurisdictions. Malpractice liability, patient safety considerations, and regulatory requirements (medical licensing) create hard barriers to full automation without human clinical authority. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Diagnosis and treatment of CNS disorders legally requires a licensed physician/neuropsychologist, with high liability, malpractice exposure, and regulatory oversight preventing AI from independently performing this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of comprehensive neuropsychological evaluation (neuroimaging, testing, clinician time, oversight) combined with AI inference and integration remains higher than or comparable to human neuropsychologist fees, especially given the need for clinician review and liability mitigation. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Given the task cannot be performed autonomously, AI cannot substitute for the clinician's cost; any AI use adds cost as a supplementary tool rather than replacing the licensed professional's labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools exist for specific subtasks (e.g., detecting lesions on MRI, scoring neuropsychological tests), but no deployed product reliably performs the full diagnostic and treatment task independently. Clinical neuropsychology requires real-time patient interaction, adaptive assessment, and differential diagnosis reasoning that exceeds current product capabilities in production. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently diagnoses and treats these neurological/neuropsychological conditions; AI is used at most for narrow decision support like imaging analysis, not the full diagnostic-treatment task. |
Diagnose and treat neural and psychological conditions in medical and surgical populations, such as patients with early dementing illness or chronic pain with a neurological basis.
9CI 3–16 · exposure 13 · augmentation 50 · importance 4.6/5 · click for rater detail
Diagnose and treat neural and psychological conditions in medical and surgical populations, such as patients with early dementing illness or chronic pain with a neurological basis.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare AI adoption for neuropsychological diagnosis remains limited; most neuropsychologists continue to rely on traditional assessment batteries and clinical interviews, with AI use confined to narrow administrative or data-processing roles rather than core diagnostic or treatment decisions. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially specialized neuropsychology, shows slow, cautious AI adoption due to regulatory, liability, and clinical validation requirements, with pilots more common than production deployment. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist neuropsychologists by automating test scoring, flagging patterns in neuroimaging or cognitive data, and generating preliminary differential-diagnosis summaries, helping clinicians work more efficiently while maintaining human oversight and final clinical authority. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with cognitive test scoring, literature review, documentation, and pattern recognition in imaging or test data, but the diagnostic and treatment judgment remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with pattern recognition in neuropsychological test scoring and data analysis, the task requires integrated clinical judgment across patient history, behavioral observation, neuroimaging interpretation, and treatment planning that depends on human expertise and real-time adaptation to individual patient presentations. |
| Task automatability | claude-sonnet-5 | 1/5 | Diagnosing and treating complex neurological/psychological conditions requires integrated clinical judgment, physical/neuropsychological examination, and longitudinal patient relationships that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Diagnosis and treatment of neural-psychological conditions are legally and ethically reserved to licensed clinical neuropsychologists; malpractice liability for diagnostic error is severe, regulatory frameworks (medical licensing boards) mandate human credentials, and patients typically expect and require human clinical judgment for treatment decisions. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This task requires licensed medical/psychological practitioners with legal authority to diagnose and treat patients, with high liability and mandatory human accountability for clinical decisions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | A trained clinical neuropsychologist commands a high loaded wage ($150k+); AI systems for neuropsychological assessment have significant setup, training, and oversight costs, and cannot yet fully replace the specialized expertise needed, making the cost ratio unfavorable for substitution. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the licensed clinical work itself, so any cost comparison favors the human who must perform the core diagnostic and treatment functions regardless of AI tool costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI tools exist for neuroimaging analysis and test administration support, but diagnosis and treatment of complex neural-psychological conditions demands nuanced clinical reasoning, differential diagnosis weighing, and therapeutic relationship management that current systems cannot perform reliably end-to-end in production clinical settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently diagnoses or treats such conditions in clinical populations; AI tools remain adjuncts to human clinicians in research or narrow decision-support roles. |
Consult with other professionals about patients' neurological conditions.
8CI 0–16 · exposure 13 · augmentation 50 · importance 4.6/5 · click for rater detail
Consult with other professionals about patients' neurological conditions.
8| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare adoption of AI for clinical consultation remains slow and heavily regulated; consultation tasks specifically require human professionals by regulation and clinical practice standards, limiting any adoption velocity toward automation. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially clinical specialty consultation, adopts AI slowly due to regulatory, liability, and workflow integration constraints despite growing AI use in documentation and diagnostics support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by summarizing patient records, flagging relevant literature, or organizing prior test results to support the neuropsychologist's preparation for consultation, moderately enhancing the professional's efficiency in gathering and organizing information before speaking with colleagues. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help by summarizing patient records, pulling relevant literature, or drafting consult notes, improving efficiency without replacing the substantive consultation. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Consultation requires nuanced clinical judgment, interpersonal negotiation, and integration of complex patient histories that current AI cannot perform end-to-end. While AI can summarize data, the task's core—synthesizing expert opinion with other professionals—demands human professional reasoning. |
| Task automatability | claude-sonnet-5 | 2/5 | Interprofessional consultation about a specific patient's neurological condition requires synthesizing clinical judgment, nuanced communication, and shared accountability that current AI cannot replicate end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and regulatory barriers exist: clinical neuropsychologists must be licensed, credentialed professionals; consultation on neurological conditions is typically governed by licensing boards and healthcare regulations that require human professional accountability and direct clinical responsibility. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Licensure, liability for clinical decisions, and legal/ethical requirements for a qualified professional to render clinical opinions make this a hard barrier task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task involves high-value professional interaction; automating it would require AI systems that perform at consultant-level quality, which is not economically viable compared to the actual cost of human neuropsychologist consultation time. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Since AI cannot substitute for the consultation itself, there is no valid cost comparison—the human expert remains necessary and any AI tool is supplementary, not substitutive. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed products reliably perform clinical consultation between professionals. AI systems can support documentation or suggest information, but they cannot authentically consult, establish clinical consensus, or take responsibility for clinical decisions in production healthcare settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently conducts clinical consultations between neuropsychologists and other providers; AI is not trusted to represent clinical judgment in these exchanges. |
Provide psychotherapy, behavior therapy, or other counseling interventions to patients with neurological disorders.
5CI 4–6 · exposure 0 · augmentation 38 · importance 3.5/5 · click for rater detail
Provide psychotherapy, behavior therapy, or other counseling interventions to patients with neurological disorders.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare and mental health sectors remain conservative on autonomous AI in clinical roles; adoption of AI-delivered psychotherapy is minimal and cautious. Organizations continue to deploy AI only in support roles (scheduling, documentation) rather than front-line treatment. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and mental health sectors show slow, cautious AI adoption for direct clinical care, with pilots mostly limited to administrative or screening support rather than therapy delivery. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with limited tasks (session note generation, evidence library lookup, routine psychoeducation scripts), but current systems offer minimal genuine augmentation of the core therapeutic skill—clinical judgment, rapport building, and real-time intervention adaptation remain entirely human-dependent. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can support session note-taking, symptom tracking, and psychoeducational material generation, offering moderate productivity gains while the neuropsychologist retains full clinical control. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Psychotherapy and counseling require genuine therapeutic alliance, real-time emotional attunement, and adaptive responsiveness to patient state—capabilities that current AI systems cannot provide. The task fundamentally depends on human presence and relational depth that AI cannot replicate. |
| Task automatability | claude-sonnet-5 | 1/5 | Delivering psychotherapy or behavior therapy to patients with neurological disorders requires clinical judgment, adaptive rapport, and real-time risk assessment that current AI cannot perform end-to-end at equal quality. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Strong legal and ethical barriers exist: clinical psychotherapy must be delivered by licensed practitioners (neuropsychologists, psychologists, counselors); malpractice liability, informed consent, and duty-of-care requirements are non-delegable to unregulated AI systems. Regulatory frameworks explicitly require human licensure. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Psychotherapy for patients with neurological disorders requires state licensure, clinical supervision, and legal accountability, making unsupervised AI delivery essentially prohibited. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-generated dialogue is cheap per unit, but oversight, liability management, and the need for human clinician involvement to ensure safety and compliance substantially raise effective cost. The human clinician remains necessary, limiting cost advantage. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While inference costs are low, the near-total lack of viable AI substitutes means the effective cost of AI-only delivery is not comparable to a licensed clinician, since liability and safety oversight costs remain high. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs independent psychotherapy or behavior therapy for neurological patients. While chatbots can simulate conversation, they lack clinical judgment, cannot assess harm risk, and are not licensed to treat patients independently. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently conducts psychotherapy for neurologically impaired patients in production; existing chatbot-based mental health tools are adjunct, not substitutive, and unproven for this specialized population. |
Diagnose and treat psychiatric populations for conditions such as somatoform disorder, dementias, and psychoses.
4CI 0–7 · exposure 5 · augmentation 50 · importance 4.4/5 · click for rater detail
Diagnose and treat psychiatric populations for conditions such as somatoform disorder, dementias, and psychoses.
4| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Psychiatric and neuropsychological practice remains predominantly human-driven with minimal AI displacement in diagnosis or treatment. High liability, regulatory scrutiny, and the centrality of therapeutic relationship mean adoption is moving slowly in pilots only; primary care screening tools show modest uptake but clinical neuropsychology practices themselves lag in automation adoption. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and clinical psychiatry adopt AI slowly for diagnostic/treatment decisions due to regulatory, liability, and safety concerns, despite faster uptake in administrative or documentation support. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist neuropsychologists by organizing symptom data, suggesting differential diagnoses from structured assessments, or flagging red-flag medical mimics—moderately enhancing diagnostic efficiency. However, augmentation is limited by the requirement that clinicians independently validate all AI suggestions through clinical judgment and direct patient assessment, which remains the foundation of the work. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with cognitive test scoring, literature review, documentation, and pattern recognition in neuroimaging or test data, meaningfully aiding but not replacing the clinician's diagnostic and treatment role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Diagnosis and treatment of psychiatric conditions requires nuanced clinical judgment, differential diagnosis across overlapping symptom presentations, and real-time therapeutic rapport. Current AI systems cannot conduct unstructured psychiatric interviews, assess for malingering or deception, or deliver evidence-based psychotherapy—core components of the task that cannot be decomposed into automatable subtasks. |
| Task automatability | claude-sonnet-5 | 1/5 | Diagnosing and treating complex psychiatric conditions requires integrated clinical judgment, physical/behavioral examination, and legal authority to make diagnoses and treatment decisions that AI cannot perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Clinical diagnosis and psychotherapy are licensed, regulated activities in most jurisdictions; only credentialed neuropsychologists or psychiatrists can legally render psychiatric diagnoses. Liability exposure for misdiagnosis in psychiatric conditions is severe, regulatory bodies (state licensing boards, Medicare/insurance) require human accountability, and patient safety in this population creates strong legal and organizational barriers. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Diagnosis and treatment of psychiatric conditions require licensure, clinical liability, and legal authorization that only credentialed neuropsychologists/psychiatrists can provide. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Clinical neuropsychologists command high salaries ($100k–$150k+) justified by years of specialized training. AI diagnostic tools require significant clinical oversight, validation, and integration costs to meet regulatory and liability standards, making the cost per reliable diagnosis comparable to or exceeding human performance. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this full task, so cost comparison favors the human clinician who is legally required regardless of AI cost efficiencies in subcomponents. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with symptom screening questionnaires and generate differential diagnoses from structured data, no deployed clinical product independently diagnoses or treats complex psychiatric populations. Psychiatric diagnosis requires integration of clinical history, behavioral observation, and exclusion of medical mimics that exceed current AI reliability; treatment requires therapeutic alliance and real-time adjustment. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product independently diagnoses or treats dementias, psychoses, or somatoform disorders in clinical practice; AI tools are at best decision-support aids used by licensed clinicians. |
Diagnose and treat pediatric populations for conditions such as learning disabilities with developmental or organic bases.
3CI 0–5 · exposure 5 · augmentation 50 · importance 4.7/5 · click for rater detail
Diagnose and treat pediatric populations for conditions such as learning disabilities with developmental or organic bases.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Clinical neuropsychology remains a highly regulated, human-centered profession with slow digitization; adoption of AI-driven automation in diagnostic or treatment contexts is negligible because of liability, licensure constraints, and the standard of care requirement for direct clinical contact. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare, especially pediatric clinical assessment and treatment, remains a slow-adopting sector for AI-driven task replacement due to regulatory, liability, and trust constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by automating scoring of standardized tests, flagging patterns in neuropsychological batteries, or organizing case documentation, which raises clinician efficiency on administrative tasks; however, core diagnostic and treatment decisions remain the neuropsychologist's responsibility. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with test scoring, report drafting, literature review, and flagging patterns in cognitive test data, but the core diagnostic and treatment relationship remains human-led. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Diagnosing and treating pediatric learning disabilities requires nuanced clinical judgment, behavioral observation, rapport-building with children, differential diagnosis across developmental and organic etiologies, and treatment planning that integrates family and educational context. Current AI systems cannot perform this complex integration end-to-end at diagnostic or therapeutic quality. |
| Task automatability | claude-sonnet-5 | 1/5 | Diagnosing and treating pediatric neuropsychological conditions requires hands-on assessment, clinical judgment, and therapeutic relationship-building that current AI cannot perform end-to-end, especially with children where nonverbal cues and adaptive interaction are critical. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Diagnosis and treatment of pediatric conditions is legally and ethically restricted to licensed clinical psychologists or neuropsychologists; liability for misdiagnosis is severe; and informed consent and direct human contact with the child and family are mandatory regulatory and professional requirements. |
| Adoption barriers | claude-sonnet-5 | 5/5 | This is a licensed clinical activity requiring a doctoral-level psychologist with specialized credentials, direct patient contact, and legal/ethical accountability for diagnosis and treatment of minors. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The clinical expertise, specialized training, and liability risk borne by neuropsychologists far exceed current AI operational costs, and no AI system can yet substitute for the clinician's billable time in assessment and treatment planning. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute delivering this service, so cost comparison favors the human clinician entirely; any AI tools only add cost as adjuncts, not replacements. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can assist with standardized test scoring and some pattern recognition in neuropsychological data, no deployed product reliably performs independent diagnostic assessment or treatment planning for pediatric learning disabilities. Clinical neuropsychology requires interactive evaluation that AI cannot currently replicate in production settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product independently diagnoses or treats pediatric learning disabilities with developmental/organic bases; this remains firmly within specialized clinical practice. |
Educate and supervise practicum students, psychology interns, or hospital staff.
0CI 0–0 · exposure 0 · augmentation 38 · importance 3.8/5 · click for rater detail
Educate and supervise practicum students, psychology interns, or hospital staff.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task is embedded in clinical education governance and professional standards that move slowly and resist de-professionalization. No data suggest sectors are adopting AI to replace clinical supervision; instead, AI might assist documentation, but supervision itself remains human-driven. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare training and supervision structures are highly regulated and slow-moving, with essentially no AI adoption for the supervisory function itself. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with scheduling, generating feedback templates, or flagging gaps in case documentation, but these supports are peripheral to the core supervisory relationship. Meaningful augmentation would be limited and modest compared to the centrality of human mentorship. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can help by providing training materials, case summaries, or feedback drafts, but the core supervisory and educational relationship remains human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Educating and supervising trainees requires dynamic relationship-building, real-time assessment of competence gaps, adaptive feedback calibrated to individual learners, and modeling of clinical judgment in ambiguous situations. Current AI cannot replicate the interactive, mentorship-driven core of this task. |
| Task automatability | claude-sonnet-5 | 1/5 | Supervising trainees and educating staff requires live judgment, mentorship, relationship-building, and accountability for trainee competence that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Supervision of clinical trainees is a legally and ethically mandated function tied to licensure, board certification, and institutional liability. Regulatory bodies (APA, AACN, accrediting bodies) require that a licensed neuropsychologist supervise practicum and internship hours. This is a hard barrier to AI substitution. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Clinical supervision of interns/practicum students is typically legally and professionally mandated to be performed by a licensed psychologist, with liability for trainee actions resting on the supervisor. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of AI systems capable of even partial supervisory function (including guardrails and human review) would exceed the cost of clinical faculty time, especially given the low automation achievable and high oversight burden. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this task, so no meaningful cost comparison favors AI; any attempt would require extensive human oversight negating savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products perform this task reliably. While AI can deliver static educational content or flag errors in work samples, the supervisory relationship—ongoing assessment, motivation, mentoring, and accountability for trainee competence—remains a human-centric function with no production equivalents. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs clinical supervision or trainee evaluation in hospital settings; this remains firmly a human responsibility. |
Related occupations — Life, Physical & Social Science
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.