Optometrists

29-1041.00
Median wage $136,570/yr42,790 employed (US)Rank #891 of 923 scored · top 97% by substitution

Diagnose, manage, and treat conditions and diseases of the human eye and visual system. Examine eyes and visual system, diagnose problems or impairments, prescribe corrective lenses, and provide treatment. May prescribe therapeutic drugs to treat specific eye conditions.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution9
Exposure10
Augmentation49

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

10 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%9

panel mean rating 1.4/5 → substitution pressure 9/100

Technical feasibility todayw 20%11

panel mean rating 1.5/5 → substitution pressure 11/100

Cost vs. human wagew 15%8

panel mean rating 1.3/5 → substitution pressure 8/100

Adoption barriersw 20%inverted — strong barriers lower the score5

panel mean rating 4.8/5 (barrier strength) → substitution pressure 5/100

Sector adoption velocityw 10%16

panel mean rating 1.7/5 → substitution pressure 16/100

Task breakdown (10 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.

Educate and counsel patients on contact lens care, visual hygiene, lighting arrangements, and safety factors.

29

CI 25–34 · exposure 25 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Optometry remains a human-centric, in-person profession with slow digitization of clinical interactions. Most practices use standard handouts and verbal counseling; AI-driven patient education is not yet standard in practice, and adoption is limited to early-stage pilots rather than industry-wide deployment.
Sector adoption velocityclaude-sonnet-52/5Healthcare, especially private optometry practices, adopts AI more slowly due to regulatory caution, patient-contact norms, and reliance on in-person exams.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-generated educational materials, visual diagrams, or reminder systems can assist optometrists in patient counseling by providing polished resources or follow-up messaging, improving consistency and patient recall. However, the core counseling task still requires human presence and judgment, so augmentation is moderate rather than transformative.
Augmentation potentialclaude-sonnet-54/5AI can generate handouts, personalize care instructions based on prescription data, and answer routine patient questions, meaningfully augmenting the optometrist's counseling efficiency.
Task automatabilityclaude-haiku-4-5-202510012/5Patient education requires adaptive dialogue, personalized explanation, and real-time responsiveness to patient confusion or concerns. While AI could draft educational materials or scripts, the counseling component—reading patient cues, adjusting explanation complexity, building trust—remains difficult for current systems without human oversight.
Task automatabilityclaude-sonnet-52/5AI chatbots can deliver generic educational content on contact lens care and eye safety, but personalized counseling tied to a specific patient's exam findings and physical demonstration/fitting checks requires in-person clinical judgment.'
Adoption barriersclaude-haiku-4-5-202510014/5Patient counseling on medical matters (contact lens safety, visual hygiene) occurs in a licensed clinical context where the optometrist has legal and fiduciary responsibility. Liability for incorrect or inadequate counseling attaches to the clinician; regulatory frameworks expect human judgment and accountability, creating meaningful substitution barriers.
Adoption barriersclaude-sonnet-53/5No strict licensing requirement mandates the optometrist personally deliver this counseling, but liability concerns, patient trust, and the need to tailor advice to individual eye conditions create moderate friction.
Cost vs. human wageclaude-haiku-4-5-202510012/5Even assuming feasible AI, oversight and integration costs for patient-facing health counseling are high, and the task is time-efficient when bundled with exams. AI would need to be substantially cheaper than the few minutes an optometrist spends per patient to justify replacement.
Cost vs. human wageclaude-sonnet-53/5AI-generated educational materials or chatbot Q&A are cheap to produce, but the overall counseling task still requires clinician time for personalization, keeping costs comparable rather than dramatically lower.
Technical feasibility todayclaude-haiku-4-5-202510012/5Chatbots and educational videos exist but lack the interactive depth, clinical judgment, and one-to-one rapport necessary for reliable patient counseling. Deployed products are narrow (static videos) or unreliable (generic chatbots); no mature production system performs this task as an optometrist would.
Technical feasibility todayclaude-sonnet-52/5Patient education apps and AI chatbots exist for general eye health information, but no deployed product reliably replaces personalized in-office counseling integrated with the exam and lens fitting.

Analyze test results and develop a treatment plan.

23

CI 20–25 · exposure 25 · augmentation 63 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Optometry practices have been slow to adopt AI-driven clinical decision support at scale; most adoption remains in larger chains or research settings. The sector is relatively low-digitization and resistant to ceding clinical authority, with pilots common but production deployment of autonomous treatment planning rare.
Sector adoption velocityclaude-sonnet-52/5Healthcare, and optometry in particular, adopts AI diagnostic aids slowly due to regulatory approval processes, liability concerns, and clinical validation requirements, with most tools still in pilot or narrow-use stages.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist by automatically summarizing test results, flagging abnormalities, and suggesting diagnostic considerations, thereby speeding the optometrist's analysis and reducing cognitive load. However, augmentation is limited to input preparation and decision support rather than transforming the core planning step.
Augmentation potentialclaude-sonnet-54/5AI-based imaging analysis, risk stratification tools, and clinical decision support can meaningfully speed up interpretation of specific tests and inform the optometrist's planning process, even though the human remains responsible for the final plan.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in analyzing certain test results (e.g., visual field data, refraction outcomes), developing a comprehensive treatment plan requires clinical judgment, patient history synthesis, and individualized decision-making that current AI systems cannot reliably perform end-to-end. The task demands integration of multiple data streams and patient-specific contraindications that exceed narrow diagnostic automation.
Task automatabilityclaude-sonnet-52/5Interpreting clinical test results and formulating a treatment plan requires integrating patient history, exam findings, and clinical judgment in ways current AI cannot reliably replicate end-to-end for diverse presentations, though AI can assist with pattern recognition in specific test types like OCT or visual fields.
Adoption barriersclaude-haiku-4-5-202510014/5Optometry is a licensed profession with regulatory requirements that the optometrist must personally perform or sign off on clinical assessment and treatment planning. Liability for incorrect treatment plans, malpractice exposure, and statutory requirements that a licensed optometrist make clinical decisions create substantial legal and organizational barriers to full automation.
Adoption barriersclaude-sonnet-55/5Developing a treatment plan is a licensed clinical act that legally must be performed or signed off on by a qualified optometrist, creating a hard regulatory and liability barrier.
Cost vs. human wageclaude-haiku-4-5-202510012/5Current AI diagnostic aids require significant infrastructure, human oversight, and integration costs. When amortized across the optometrist's loaded wage for the combined analysis-and-planning task, AI remains cost-comparable or more expensive than direct human performance.
Cost vs. human wageclaude-sonnet-52/5AI tools can cheaply flag abnormalities in specific tests, but the overall analysis-and-planning task still requires a licensed optometrist's review, so total cost savings versus the human clinician are limited today.
Technical feasibility todayclaude-haiku-4-5-202510012/5AI products exist for isolated diagnostic support (retinal imaging analysis, automated refraction), but no deployed system reliably performs the full end-to-end task of analyzing results and synthesizing a treatment plan at production scale. Clinical validation and regulatory requirements mean such systems remain largely research or pilot stage.
Technical feasibility todayclaude-sonnet-52/5Some diagnostic decision-support tools exist for narrow tasks (e.g., retinal image analysis, glaucoma risk scoring) but no deployed product autonomously synthesizes full optometric test results into a comprehensive treatment plan in production.

Prescribe, supply, fit and adjust eyeglasses, contact lenses, and other vision aids.

10

CI 0–20 · exposure 13 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Optometry is a licensed clinical profession with minimal evidence of AI-driven automation or agent displacement in production settings; adoption remains concentrated in administrative support rather than clinical decision-making or physical adjustment.
Sector adoption velocityclaude-sonnet-52/5Healthcare delivery, especially licensed clinical services with physical components, adopts AI more slowly than digital-only sectors due to regulation, liability, and hands-on requirements.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with aspects such as analyzing refraction measurements, suggesting lens configurations, or automating preliminary refractive tests, helping optometrists work more efficiently while the professional remains responsible for fitting and adjustment.
Augmentation potentialclaude-sonnet-53/5AI-assisted autorefraction, imaging analysis, and lens design software can meaningfully speed up parts of the exam and fitting workflow while the optometrist retains final judgment and physical execution.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires precise fitting and physical adjustment of vision aids, complex three-dimensional spatial reasoning about facial anatomy, and real-time iterative feedback from the patient—none of which current AI systems can execute end-to-end without direct human intervention.
Task automatabilityclaude-sonnet-52/5Diagnostic refraction and prescribing require licensed clinical judgment, and physical fitting/adjustment of eyewear requires hands-on manipulation, so AI cannot yet do this end-to-end at equal quality with major time savings.
Adoption barriersclaude-haiku-4-5-202510015/5This task is legally restricted to licensed optometrists and ophthalmologists in most jurisdictions; prescribing and fitting vision aids requires professional licensure and carries liability for visual outcomes, creating hard regulatory barriers to automation.
Adoption barriersclaude-sonnet-55/5Prescribing corrective lenses is a legally regulated act requiring a licensed optometrist or ophthalmologist, and fitting/adjustment involves physical contact and liability for vision-affecting devices.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task involves regulated clinical decision-making, physical fabrication, fitting, and adjustment—all requiring trained optometrist labor; AI cost savings are negligible compared to the high-skill human labor required.
Cost vs. human wageclaude-sonnet-52/5Automated refraction hardware can lower per-exam cost somewhat, but licensed optometrist oversight, physical fitting, and liability requirements keep overall costs comparable to human-delivered care.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI product performs the full supply, fit, and adjustment workflow autonomously; the task demands clinical judgment, physical manipulation, and legal authority that remain gatekept by licensed humans.
Technical feasibility todayclaude-sonnet-52/5Some autorefractors and online vision-test apps exist for prescription estimation, but they are supplementary tools, not products that independently prescribe, fit, or adjust eyewear in production.

Provide patients undergoing eye surgeries, such as cataract and laser vision correction, with pre- and post-operative care.

8

CI 0–16 · exposure 13 · augmentation 50 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Eye care practices are traditional; adoption of AI in this clinical context is slow despite digitization, with most adoption limited to imaging aids rather than workflow replacement.
Sector adoption velocityclaude-sonnet-51/5Healthcare delivery involving surgical care follows slow, heavily regulated adoption patterns with minimal AI displacement of hands-on clinical care roles.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can usefully assist optometrists by automating image analysis, flagging potential complications, and organizing monitoring data, raising efficiency in documentation and review without removing the optometrist from care decisions.
Augmentation potentialclaude-sonnet-53/5AI can assist with documentation, image analysis (e.g., OCT scans), scheduling, and patient education materials, improving efficiency around the core task without replacing the clinician's judgment or direct care.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist with some pre- and post-operative monitoring (e.g., analyzing imaging, flagging complications), the task critically requires in-person clinical judgment, patient education, and reassurance that only a licensed optometrist can provide. No AI system today can perform the full spectrum of pre/post-operative care end-to-end with 50% time savings at equal quality.
Task automatabilityclaude-sonnet-51/5This requires hands-on clinical examination, direct patient interaction, and licensed judgment before and after surgery; no AI system can perform the physical exam, counseling, and decision-making end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Licensing laws and professional liability require that a licensed optometrist evaluate patients and sign off on surgical care decisions; this is a hard regulatory barrier preventing substitution.
Adoption barriersclaude-sonnet-55/5Optometrists must be licensed, and pre/post-operative surgical care involves direct liability, physical examination, and legal scope-of-practice requirements that mandate human, licensed involvement.
Cost vs. human wageclaude-haiku-4-5-202510011/5The human cost is modest (routine follow-up care by optometrists), and AI integration would require oversight infrastructure that does not offset the wage cost of the professional clinician who must remain accountable.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot substitute for the clinical care itself, there is no viable AI cost comparison—human licensed provider costs remain the only option for this task.
Technical feasibility todayclaude-haiku-4-5-202510012/5Deployed AI products can support image analysis and complication detection, but no system reliably performs the full clinical care workflow independently; this remains research-stage for autonomous execution and requires human supervision in all production settings.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs pre/post-operative eye surgery care; AI is at best used for adjunct documentation or imaging analysis, not the care delivery itself.

Prescribe therapeutic procedures to correct or conserve vision.

8

CI 0–16 · exposure 13 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI for autonomous prescription in optometry is minimal; the sector remains heavily dependent on licensed professionals performing the prescriptive function. Regulatory constraints and patient safety liability create structural resistance to replacing human clinical judgment.
Sector adoption velocityclaude-sonnet-52/5Healthcare/optometry adoption of AI is growing but remains slow and cautious for clinical decision-making and prescribing due to regulation and liability concerns.
Augmentation potentialclaude-haiku-4-5-202510013/5AI-powered diagnostic tools and refractive-error analysis can meaningfully assist optometrists by automating preliminary measurements and flagging anomalies, improving the efficiency of the examination process while the optometrist retains prescriptive authority.
Augmentation potentialclaude-sonnet-53/5AI can assist with diagnostic imaging analysis, decision support, and documentation, helping optometrists inform treatment choices, though the prescribing judgment remains human-driven.
Task automatabilityclaude-haiku-4-5-202510012/5While AI can assist in analyzing visual data and suggesting corrective parameters, the task of prescribing therapeutic procedures requires nuanced clinical judgment about individual patient biology, comorbidities, and contraindications that current systems cannot reliably perform end-to-end. No AI system can yet replace the optometrist's diagnostic reasoning and prescription authority at scale.
Task automatabilityclaude-sonnet-51/5Prescribing therapeutic procedures requires clinical examination, judgment, and legal authority to diagnose and treat a specific patient; current AI cannot perform this end-to-end.wt
Adoption barriersclaude-haiku-4-5-202510015/5Prescribing therapeutic procedures is a licensed clinical function requiring a licensed optometrist or ophthalmologist to legally issue prescriptions in virtually all jurisdictions. Regulatory and legal barriers prevent autonomous AI systems from performing this task without human professional sign-off.
Adoption barriersclaude-sonnet-55/5Prescribing therapeutic procedures is a licensed medical act requiring optometrist authorization, with significant liability and regulatory oversight preventing automation.
Cost vs. human wageclaude-haiku-4-5-202510011/5The AI infrastructure and oversight required to support prescription of therapeutic procedures remains expensive relative to the optometrist's hourly rate, especially given regulatory and liability requirements that would necessitate human review and approval of any AI recommendation.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot legally or reliably perform this task alone, there is no functional cost substitute; any AI use still requires full optometrist involvement and cost.
Technical feasibility todayclaude-haiku-4-5-202510012/5Decision-support tools exist to aid optometrists in analyzing vision tests and refractive errors, but no deployed AI product independently prescribes therapeutic vision procedures with reliability sufficient for clinical deployment. Research prototypes exist but lack the validation and integration needed for autonomous clinical use.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently prescribes therapeutic vision procedures; AI diagnostic aids exist but do not replace the prescribing act itself.

Examine eyes, using observation, instruments, and pharmaceutical agents, to determine visual acuity and perception, focus, and coordination and to diagnose diseases and other abnormalities, such as glaucoma or color blindness.

5

CI 3–7 · exposure 5 · augmentation 63 · importance 5.0/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Adoption of AI assistive tools (image analysis, screening) is emerging in some practices, but replacement of the optometrist's role is not occurring; sector adoption remains concentrated in larger clinics and research settings, not mainstream practice.
Sector adoption velocityclaude-sonnet-52/5Healthcare/optometry is a moderately digitizing sector with AI diagnostic aids in early deployment (e.g., retinal imaging screening), but physical examination workflows remain largely unchanged and adoption of full-task automation is minimal.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist with specific narrower tasks—analysis of retinal images, visual field interpretation, and disease screening—which improves efficiency and diagnostic support. However, the core clinical examination and judgment remain human-dependent, limiting augmentation to partial acceleration of selected sub-tasks rather than transforming overall productivity.
Augmentation potentialclaude-sonnet-54/5AI-based image analysis tools (e.g., for diabetic retinopathy or glaucoma screening) can meaningfully assist optometrists in flagging abnormalities and supporting diagnostic decisions during the exam.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires direct in-person examination using specialized instruments (phoropters, tonometers, slit lamps) and patient interaction that AI cannot perform. Clinical judgment in interpreting instrument readings combined with patient observations and disease diagnosis requires the integrative human expertise that AI systems today cannot replicate end-to-end with equal quality.
Task automatabilityclaude-sonnet-51/5This task requires hands-on physical examination, administration of pharmaceutical agents, and direct patient interaction with instruments that AI cannot physically operate or administer today.
Adoption barriersclaude-haiku-4-5-202510015/5Significant legal and regulatory barriers exist: optometrists are licensed professionals required by law in most jurisdictions to perform patient examinations, prescribe lenses, and diagnose ocular diseases. Liability and malpractice considerations strongly protect the profession, as diagnostic errors directly harm patient outcomes.
Adoption barriersclaude-sonnet-55/5Licensed optometrists are legally required to perform eye exams and prescribe pharmaceutical diagnostic agents, with strict regulatory and liability requirements around diagnosis of disease.
Cost vs. human wageclaude-haiku-4-5-202510011/5An optometrist's loaded wage is substantial (≥$60–80/hour), and the full examination requires physical presence and specialized equipment. Current AI assistance tools (image analysis software) cost pennies per use but do not replace the full examination, making direct cost comparison inappropriate; the task itself cannot be fully automated to create an economically viable substitution.
Cost vs. human wageclaude-sonnet-51/5There is no AI system capable of performing the physical examination end-to-end, so no meaningful cost comparison to a human optometrist exists for this task as a whole.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist with image analysis of retinal imaging or visual field data interpretation, no deployed product performs the full examination—instrument use, patient interaction, real-time adjustment based on patient responses, and comprehensive clinical diagnosis—reliably in production. Some AI tools exist for specific narrow tasks (e.g., diabetic retinopathy detection) but not the full scope.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs the physical eye examination or administers pharmaceutical agents; AI-assisted diagnostic imaging analysis exists but is a narrow sub-component, not the full task.

Consult with and refer patients to ophthalmologist or other health care practitioner if additional medical treatment is determined necessary.

5

CI 3–7 · exposure 5 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510012/5Healthcare sectors adopt AI cautiously due to regulatory and liability constraints; referral automation remains largely experimental and adoption is slow except as narrow decision-support tools rather than autonomous systems.
Sector adoption velocityclaude-sonnet-52/5Healthcare referral workflows are adopting AI slowly for decision support, but the referral judgment itself remains a manual, regulated clinical process with limited automation penetration.
Augmentation potentialclaude-haiku-4-5-202510013/5AI can assist by analyzing clinical measurements, flagging red-flag symptoms, or suggesting relevant specialist types, which would usefully inform the optometrist's referral decision while the human retains clinical and legal responsibility.
Augmentation potentialclaude-sonnet-53/5AI diagnostic tools (e.g., retinal imaging analysis) can flag abnormalities that support the optometrist's referral decision, offering moderate assistance without replacing judgment.
Task automatabilityclaude-haiku-4-5-202510011/5This task requires clinical judgment to determine medical necessity and appropriate specialist referral—decisions that depend on nuanced patient assessment, medical history synthesis, and professional accountability that current AI systems cannot reliably perform end-to-end.
Task automatabilityclaude-sonnet-51/5Referral requires clinical judgment about a patient's specific condition, risk factors, and coordination with another licensed provider, which is not something AI can execute end-to-end today.
Adoption barriersclaude-haiku-4-5-202510015/5Optometrists are licensed professionals whose medical judgment and referral authority is legally mandated; liability for inappropriate or missed referrals rests with the practitioner, creating hard barriers to full automation.
Adoption barriersclaude-sonnet-55/5Referral decisions are part of licensed clinical practice with legal and liability implications, requiring a licensed optometrist to make and document the determination.
Cost vs. human wageclaude-haiku-4-5-202510011/5The task centers on human clinical judgment and patient interaction; automation offers minimal cost advantage when human optometrists must ultimately make and communicate the referral decision anyway.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this task independently, so no meaningful cost comparison favors AI over the optometrist's judgment and liability coverage.
Technical feasibility todayclaude-haiku-4-5-202510012/5While AI can assist in flagging potential referral candidates or summarizing clinical data, no deployed system reliably makes independent referral decisions or handles the consultative interaction required between optometrist and specialist without human oversight.
Technical feasibility todayclaude-sonnet-51/5No deployed product autonomously decides on and executes referrals to ophthalmologists; this remains a clinician-driven decision-making process.

Prescribe medications to treat eye diseases if state laws permit.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Adoption of AI for autonomous medication prescription in optometry is near-zero because of regulatory barriers and liability concerns. While AI-assisted diagnostic tools see some pilot adoption, no sector-wide automation of the prescriptive act itself is occurring.
Sector adoption velocityclaude-sonnet-52/5Healthcare prescribing workflows are adopting AI slowly for diagnostics and documentation, but the prescribing decision itself remains untouched by automation in practice.
Augmentation potentialclaude-haiku-4-5-202510012/5AI can assist with image analysis, condition detection, and treatment recommendations to support optometrist decision-making, but current systems provide limited augmentation for the full prescriptive task itself. Most assistance is indirect (supporting diagnosis) rather than directly amplifying prescription authority.
Augmentation potentialclaude-sonnet-53/5AI can assist with diagnostic imaging analysis, drug interaction checks, and documentation, helping optometrists make and record prescribing decisions faster.
Task automatabilityclaude-haiku-4-5-202510011/5Prescribing medications requires clinical judgment, diagnosis of disease pathology, knowledge of patient medical history and contraindications, and legal authority. Current AI systems cannot perform the full diagnostic and prescriptive process end-to-end with the necessary reliability and accountability.
Task automatabilityclaude-sonnet-51/5Diagnosing and prescribing medication for eye disease requires clinical judgment, direct patient examination, and legal accountability that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5State optometry licensing laws explicitly regulate who can prescribe medications. In most jurisdictions, only licensed optometrists (and in some states, only specific certified optometrists) are legally authorized to prescribe. This creates a hard legal barrier that prevents autonomous AI prescription.
Adoption barriersclaude-sonnet-55/5Prescribing is tightly regulated, requiring state licensure and legal authorization, making this one of the strongest barrier-protected tasks.
Cost vs. human wageclaude-haiku-4-5-202510011/5AI systems capable of supporting diagnostic components still require significant human oversight, integration costs, and liability management. The all-in cost of automation infrastructure and the regulatory requirements make this more expensive than traditional optometrist-led prescription.
Cost vs. human wageclaude-sonnet-51/5Since AI cannot legally or practically replace the prescribing act, there is no valid AI-only cost comparison; the human optometrist remains essential.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product reliably performs autonomous medication prescription in optometry. While AI can assist with image analysis or condition classification, the final prescriptive decision and legal responsibility remain with the licensed optometrist in all current real-world deployments.
Technical feasibility todayclaude-sonnet-51/5No deployed product independently prescribes ocular medications to patients; AI is at most a decision-support tool used by a licensed prescriber.

Provide vision therapy and low-vision rehabilitation.

1

CI 0–3 · exposure 0 · augmentation 38 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare sectors, especially clinical vision care, adopt automation slowly due to regulatory, liability, and quality-of-care constraints. Vision therapy remains a hands-on, in-person clinical service with minimal AI integration in practice today.
Sector adoption velocityclaude-sonnet-52/5Healthcare delivery involving hands-on rehabilitation adopts AI slowly, with most current use limited to diagnostic imaging or administrative support rather than therapy delivery.
Augmentation potentialclaude-haiku-4-5-202510012/5AI might assist with some peripheral tasks like generating patient education materials or tracking progress metrics, but offers limited augmentation for the core clinical work of conducting therapy sessions and making real-time adaptive adjustments based on patient response.
Augmentation potentialclaude-sonnet-53/5AI can assist with tracking patient progress, generating exercise plans, or analyzing visual field data, but the core therapeutic delivery still depends on the practitioner.
Task automatabilityclaude-haiku-4-5-202510011/5Vision therapy and low-vision rehabilitation require real-time clinical assessment, patient interaction, physical manipulation of equipment, and adaptive instruction tailored to individual patient needs and responses. Current AI cannot perform these hands-on, interactive clinical interventions.
Task automatabilityclaude-sonnet-51/5This task requires hands-on assessment, physical exercises, adaptive device fitting, and personalized therapeutic interaction that current AI cannot perform end-to-end.
Adoption barriersclaude-haiku-4-5-202510015/5Vision therapy and low-vision rehabilitation are regulated clinical services that must be performed or directly supervised by a licensed optometrist or physician. Legal and licensing requirements create hard barriers to automation; a licensed professional must personally conduct and sign off on patient care.
Adoption barriersclaude-sonnet-55/5This is a licensed clinical activity requiring an optometrist's direct assessment, judgment, and often physical intervention, with strong liability and regulatory requirements for human delivery.
Cost vs. human wageclaude-haiku-4-5-202510011/5Vision therapy and rehabilitation are inherently labor-intensive, requiring skilled professional time and physical presence. Any AI assistance would supplement rather than replace the optometrist, making the combined cost higher than the human performing the task alone.
Cost vs. human wageclaude-sonnet-51/5There is no viable AI substitute performing this clinical service, so cost comparison favors the human provider entirely.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed AI products reliably perform vision therapy or low-vision rehabilitation as a standalone clinical service. These tasks require licensed optometric judgment, patient-specific adaptation, and direct patient contact that existing AI systems cannot replicate in a clinical setting.
Technical feasibility todayclaude-sonnet-51/5No deployed AI product independently delivers vision therapy or low-vision rehabilitation; this remains a clinician-delivered, hands-on service.

Remove foreign bodies from the eye.

0

CI 0–0 · exposure 0 · augmentation 25 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
DimensionModelRatingRationale
Sector adoption velocityclaude-haiku-4-5-202510011/5Healthcare adoption of autonomous procedures on sensitive anatomy is extremely slow and regulatory-constrained. No production deployment of AI systems for eye foreign body removal exists in clinical practice.
Sector adoption velocityclaude-sonnet-51/5Hands-on clinical eye procedures show essentially no AI adoption or displacement trend; this is a physical, low-digitization task within healthcare delivery.
Augmentation potentialclaude-haiku-4-5-202510012/5AI imaging may assist in locating and visualizing foreign bodies before manual extraction, but the core task of safe removal remains human-dependent. Augmentation value is limited to pre-procedure imaging rather than transforming the procedure itself.
Augmentation potentialclaude-sonnet-52/5AI can assist with diagnostic imaging or documentation around the visit, but offers minimal direct assistance during the physical act of foreign body removal itself.
Task automatabilityclaude-haiku-4-5-202510011/5Removing foreign bodies from the eye requires precise physical manipulation in a living patient's eye, which current AI systems cannot perform. This task demands real-time sensorimotor control, sterile technique, and immediate response to patient movement—entirely outside the scope of today's AI capabilities.
Task automatabilityclaude-sonnet-51/5Removing foreign bodies from the eye requires fine motor manipulation of physical instruments under direct visualization of a delicate organ, which current AI systems cannot perform end-to-end at all.
Adoption barriersclaude-haiku-4-5-202510015/5Removing foreign bodies from the eye is a regulated clinical procedure that requires a licensed optometrist or ophthalmologist to perform or directly supervise. Legal, liability, and patient-safety requirements create hard barriers to automation.
Adoption barriersclaude-sonnet-55/5This is an invasive clinical procedure requiring licensed practitioner hands-on care, with direct liability, physical contact, and regulatory scope of practice requirements.
Cost vs. human wageclaude-haiku-4-5-202510011/5The specialized equipment, validation, and liability overhead for any automated eye intervention far exceed the cost of a trained optometrist performing the procedure. No economic comparison favors automation today.
Cost vs. human wageclaude-sonnet-51/5There is no AI system that can substitute for this physical procedure, so cost comparison favors the human provider entirely.
Technical feasibility todayclaude-haiku-4-5-202510011/5No deployed product performs end-to-end foreign body removal from the eye. While imaging AI exists, the actual extraction requires surgical or clinical robotics with haptic feedback and autonomous decision-making in a sensitive anatomical context that remains research-stage.
Technical feasibility todayclaude-sonnet-51/5No deployed product performs physical foreign body removal from the eye; this remains entirely a manual clinical procedure performed by trained practitioners.

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.