Ophthalmic Medical Technicians
29-2057.00Median wage $45,570/yr71,010 employed (US)Rank #620 of 923 scored · top 67% by substitution
Assist ophthalmologists by performing ophthalmic clinical functions. May administer eye exams, administer eye medications, and instruct the patient in care and use of corrective lenses.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution23
Exposure21
Augmentation44
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
20 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%22
panel mean rating 1.9/5 → substitution pressure 22/100
Technical feasibility todayw 20%20
panel mean rating 1.8/5 → substitution pressure 20/100
Cost vs. human wagew 15%18
panel mean rating 1.7/5 → substitution pressure 18/100
Adoption barriersw 20%inverted — strong barriers lower the score29
panel mean rating 3.8/5 (barrier strength) → substitution pressure 29/100
Sector adoption velocityw 10%22
panel mean rating 1.9/5 → substitution pressure 22/100
Task breakdown (20 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.
Take and document patients' medical histories.
49CI 39–59 · exposure 50 · augmentation 75 · importance 5.0/5 · click for rater detail
Take and document patients' medical histories.
49| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 3/5 | Healthcare has moderate AI adoption for documentation assist (voice recognition, template auto-population), but systematic replacement of medical history-taking remains slow due to regulation, liability aversion, and the need for clinician oversight—adoption is broader in large health systems than in smaller ophthalmology practices. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially smaller specialty clinics like ophthalmology, adopts clinical AI tools more slowly than digital-native sectors due to compliance, EHR integration complexity, and clinician trust issues. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI can significantly assist technicians by auto-filling common fields, transcribing spoken history, and flagging missing elements, which can accelerate the documentation workflow while the technician remains responsible for accuracy and completeness. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI scribes and structured intake tools meaningfully speed up documentation and reduce technician burden while the technician remains responsible for verifying and finalizing the record. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | AI can partially automate medical history capture through conversational interfaces or form-filling from patient input, but requires human oversight to identify critical details, clarify ambiguities, and handle complex medical narratives—falling short of the 50% time-saving bar for the full task end-to-end. |
| Task automatability | claude-sonnet-5 | 3/5 | AI can transcribe and structure verbal history-taking or extract data from intake forms with substantial time savings, but nuanced follow-up questioning and clinical judgment about relevance still typically require human involvement.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Medical history documentation is regulated under HIPAA, patient consent, and medical practice standards; liability for missing or incorrect history falls on the practice and clinician, creating strong incentives to retain human verification and sign-off rather than fully automate the intake process. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No strict licensure requires a human to take history, though documentation must be accurate for medical-legal and clinical care purposes, creating moderate oversight requirements. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted documentation tools reduce technician labor somewhat, but the cost of the software, oversight review time, and integration with existing EHRs means the total cost-per-task is comparable to or only modestly cheaper than straightforward human data entry. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | AI scribe/documentation tools cost a small fraction of a technician's loaded wage per encounter, though some human review remains necessary, slightly reducing the savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Deployed products (EHR systems with AI documentation assist, voice-to-text medical capture) exist and are used in some practices, but they still produce material documentation gaps and require human review to ensure completeness and accuracy in a medical context. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Ambient scribe and clinical documentation products (e.g., AI scribes in EHRs) are deployed in some ophthalmology practices, but adoption is uneven and accuracy for specialized ophthalmic history still needs verification. |
Measure and record lens power, using lensometers.
48CI 21–75 · exposure 50 · augmentation 63 · importance 4.9/5 · click for rater detail
Measure and record lens power, using lensometers.
48| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic practices are moderate-to-low in digitization and automation adoption compared to information or finance sectors. While some clinics use digital lensometers, routine integration of AI or robotic measurement is rare; adoption remains pilot-phase at best. |
| Sector adoption velocity | claude-sonnet-5 | 4/5 | Automatic lensometers are already standard in optical retail and clinical settings, representing widespread, mature adoption rather than pilot-stage use. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Digital lensometers with electronic data capture and automatic record logging can assist technicians by reducing manual transcription and speeding up workflow. However, the core measurement task itself (positioning, reading) remains largely technician-driven, limiting transformative augmentation. |
| Augmentation potential | claude-sonnet-5 | 4/5 | Even where full automation is used, technicians benefit from faster, more accurate readings and digital recordkeeping that streamlines their workflow. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | A lensometer measurement is fundamentally a physical device interaction requiring precise positioning and optical alignment of a lens. While image recognition could theoretically identify readings on older analog displays, the physical manipulation and positioning of lenses into the device cannot be automated without specialized robotics, making end-to-end automation without significant setup infeasible. |
| Task automatability | claude-sonnet-5 | 4/5 | Automated lensmeters/lensometers already measure and digitally record lens power with minimal human input, meeting the ≥50% time-saving bar, though a technician still positions lenses and verifies output. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Quality assurance and liability considerations in optical/medical settings create friction: measurements affect prescription accuracy and patient safety, so oversight and validation by trained personnel are expected. Regulatory standards for ophthalmic testing may also implicitly or explicitly require human skill verification. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensure specifically requires a human to perform this measurement; automated devices are already standard equipment, though final clinical sign-off in some contexts remains with staff. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The measurement itself is fast (under a minute) and routine; human technician labor cost is low per measurement. Automating would require capital investment in robotic systems and AI integration that far exceeds the hourly cost of a technician performing the task. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated lensometers are a one-time capital cost that quickly becomes far cheaper per measurement than paying a technician to do manual readings repeatedly. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI system reliably performs the physical act of lens positioning and measurement today. Digital lensometers may output data electronically, but capturing and interpreting those outputs requires integration into a broader system, and no standalone product does this lens-measurement task reliably in clinical production settings. |
| Technical feasibility today | claude-sonnet-5 | 4/5 | Automatic lensometers are mature, widely deployed products in optical practices and labs, reliably measuring sphere, cylinder, axis, and prism today. |
Call patients to inquire about their post-operative status or recovery.
40CI 29–51 · exposure 38 · augmentation 50 · importance 4.0/5 · click for rater detail
Call patients to inquire about their post-operative status or recovery.
40| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare adoption of autonomous patient outreach remains slow outside routine reminder calls; ophthalmology practices are typically small to mid-sized and risk-averse. Pilots exist, but production deployment of AI-driven post-op status assessment is rare in this sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare administrative/clinical workflows adopt AI more slowly than other sectors due to compliance, EHR integration complexity, and patient trust concerns, even though automated appointment reminders are common. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist technicians by auto-dialing, prompting with question scripts, and flagging high-risk keywords (e.g., 'pain', 'vision loss') for escalation, meaningfully speeding up call workflows while the technician retains clinical judgment and documentation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can pre-screen patients via automated questionnaires or calls, flagging concerns for technician follow-up, which usefully augments but doesn't replace the judgment-based portions of the task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While call initiation and basic scripted inquiries are automatable, assessing post-operative recovery status typically requires understanding patient nuance, detecting complications, and deciding whether escalation is needed—tasks that current AI struggles with reliably. An AI could handle only the initial contact and basic screening, but most of the clinical judgment remains human. |
| Task automatability | claude-sonnet-5 | 3/5 | Simple status-check calls with structured questions could be handled by voice AI/IVR systems, but detecting concerning symptoms and deciding on escalation still benefits from human judgment, limiting full end-to-end automation. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Ophthalmic care is regulated; automated systems must comply with medical record-keeping (HIPAA), patient safety protocols, and clinician liability for missed post-op complications. Most healthcare systems require human clinical staff to document and act on findings, creating a hard barrier to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement to make a phone call, but clinical liability concerns, patient preference for human contact after surgery, and need for appropriate escalation of symptoms create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 3/5 | Automated call systems are inexpensive per call, but integration into clinical workflows, compliance with HIPAA, and the cost of errors (missed complications) offset savings; total cost is roughly comparable to employing part-time call staff. |
| Cost vs. human wage | claude-sonnet-5 | 4/5 | Automated calling/IVR systems are inexpensive per call compared to technician time, especially for routine status checks, though integration and occasional escalation handling add some cost. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Chatbots and automated calling systems exist, but deploying them to ophthalmic post-op patients without human oversight carries clinical risk; no mature product is known to run this task end-to-end in ophthalmology clinics at scale without fallback to human nurses. |
| Technical feasibility today | claude-sonnet-5 | 3/5 | Automated patient outreach and post-op check-in call products exist (voice bots, automated surveys) and are used in some healthcare systems, but adoption for clinical follow-up with nuanced triage is still limited and often supplements rather than replaces human calls. |
Conduct visual field tests to measure field of vision.
37CI 25–49 · exposure 42 · augmentation 63 · importance 4.7/5 · click for rater detail
Conduct visual field tests to measure field of vision.
37| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 4/5 | Ophthalmic practices, especially larger clinics and hospital systems, have adopted automated perimetry widely over the past two decades. AI-assisted quality control and interpretation tools are seeing increasing adoption in digitized ophthalmology settings. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare diagnostic support roles adopt automation slowly due to regulatory, safety, and physical-presence constraints; visual field testing equipment has been automated for years but AI-driven full task replacement is not occurring at pace. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI significantly augments technician productivity by automating equipment calibration, detecting retest artifacts, flagging suspicious responses in real time, and providing automated initial quality assessment, allowing technicians to focus on patient interaction and rare difficult cases. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with test result analysis, artifact detection, and flagging unreliable results, offering moderate productivity gains to technicians and interpreting clinicians while the technician still conducts the physical test. |
| Task automatability | claude-haiku-4-5-20251001 | 3/5 | Automated visual field testing equipment (e.g., Humphrey Field Analyzer) can perform the measurement itself, but technician interpretation, patient positioning, artifact detection, and response validation still require human oversight. AI could reduce technician time by ~40–50% through automated protocols and initial quality checks, meeting the threshold partially. |
| Task automatability | claude-sonnet-5 | 2/5 | Automated perimetry machines already run the test itself, but the technician's role of positioning the patient, ensuring proper fixation, coaching reliability, and interpreting patient responses during testing still requires hands-on human presence not fully replaceable by current AI. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Visual field testing results are used for clinical diagnosis and treatment decisions, creating liability exposure if automation errors occur. Medical oversight requirements, state licensure of technicians, and physician responsibility for test interpretation create legal and regulatory barriers to full substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is a clinical diagnostic procedure typically performed by a certified/licensed ophthalmic technician under supervision, with patient-contact and quality-control requirements that create strong regulatory and liability barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated perimetry equipment has high upfront capital cost and ongoing maintenance, while AI-assisted interpretation adds software licensing. The savings relative to technician labor are modest because a technician is still required for patient management, artifact handling, and quality control. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | The equipment is already a capital investment requiring a trained operator; there is no cheaper AI substitute that removes the need for a technician present to run and monitor the test on real patients. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 4/5 | Mature automated perimetry devices are deployed in clinical practice worldwide and reliably capture visual field data. AI-assisted analysis and quality assurance systems exist in production, though full end-to-end automation without technician oversight remains limited. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Perimeters are automated diagnostic devices, but they are not AI systems performing the technician's job end-to-end; no deployed AI product independently conducts patient setup, calibration, and quality monitoring for visual field tests. |
Operate ophthalmic equipment, such as autorefractors, phoropters, tomographs, or retinoscopes.
32CI 16–47 · exposure 41 · augmentation 63 · importance 4.9/5 · click for rater detail
Operate ophthalmic equipment, such as autorefractors, phoropters, tomographs, or retinoscopes.
32| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmology clinics and labs are adopting automated measurement devices incrementally, but technician displacement through full automation remains rare; most adoption consists of equipment upgrades that complement rather than replace technicians. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare technical support roles show slow AI/robotics adoption for hands-on patient equipment operation, with automation mostly limited to data interpretation rather than physical instrument handling. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | AI-assisted measurement guidance, real-time quality feedback, and automated data capture significantly enhance technician productivity and measurement consistency, allowing faster throughput and reducing manual data entry and repositioning time. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist by interpreting scan results, flagging abnormalities, or improving instrument calibration/software, but the physical operation and patient handling remain human-driven. |
| Task automatability | claude-haiku-4-5-20251001 | 4/5 | Modern AI vision systems and robotic automation can operate standardized ophthalmic equipment like autorefractors and phoropters with high precision, reducing operator interaction significantly. However, some equipment still requires subtle human judgment in patient positioning or adjusting for individual anatomy, limiting full end-to-end automation without setup. |
| Task automatability | claude-sonnet-5 | 2/5 | Physical operation of ophthalmic instruments on patients requires hands-on positioning, patient interaction, and real-time adjustment that current AI systems cannot perform without robotic embodiment, which is not yet standard equipment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Medical device regulations (FDA, international standards) require validated human-supervised operation, and patient safety liability creates strong incentives for human oversight. Insurance and clinical protocols typically mandate trained personnel certification and accountability. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct patient contact and equipment operation typically requires trained/certified personnel under clinical protocols, and liability for misuse of diagnostic equipment on patients creates strong barriers to substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current robotic and AI systems for ophthalmology equipment operation are capital-intensive and require ongoing maintenance, integration, and validation; the total cost per test likely exceeds the loaded wage of a trained ophthalmic technician. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-only substitute performing this physical task, so no meaningful cost comparison favors AI; a human technician remains necessary for equipment operation. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 3/5 | Robotic arms and automated measurement systems exist in laboratory and research settings, but few mature commercial products reliably perform complete ophthalmic measurements in clinical workflows without human oversight. Integration into existing clinical practice remains limited. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously operates autorefractors, phoropters, tomographs, or retinoscopes on patients in clinical practice; automated diagnostic instruments still require a technician to position and operate them. |
Measure visual acuity, including near, distance, pinhole, or dynamic visual acuity, using appropriate tests.
28CI 25–30 · exposure 25 · augmentation 50 · importance 4.9/5 · click for rater detail
Measure visual acuity, including near, distance, pinhole, or dynamic visual acuity, using appropriate tests.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic practices are moderately digitized but remain conservative in automating patient-facing clinical tasks. Adoption of automated screening has been slow and limited to specific settings (health fairs, telehealth screening); traditional in-office acuity testing by technicians remains standard across most practices. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially outpatient ophthalmology clinics, has been slower to adopt full automation for direct patient-facing diagnostic tasks compared to information/professional service sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted instruments (e.g., automated refraction with human interpretation, digital charts with AI-flagged anomalies) can meaningfully accelerate technician workflow and reduce eye strain from manual testing. However, augmentation is incremental rather than transformative, as the core task remains technician-driven. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Digital and computerized visual acuity systems can streamline recording, standardize test administration, and reduce technician time on repetitive measurement tasks, improving efficiency while the technician remains involved. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Visual acuity measurement requires precise instrument operation and patient interaction to ensure accurate results. While automated refraction devices exist, they still require human setup, patient cooperation, and judgment to validate measurements—no current AI system can fully replace the technician's role in administering these tests end-to-end with equal quality and ≥50% time savings. |
| Task automatability | claude-sonnet-5 | 2/5 | Some visual acuity testing can be partially automated with computerized charts and autorefractors, but the full hands-on process including patient positioning, communication, and judgment calls for pinhole/dynamic testing still requires human execution. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Ophthalmic technicians operate under standing orders from licensed eye care providers, and visual acuity measurement directly feeds clinical diagnosis. Regulatory requirements, liability concerns (misdiagnosis risk), and the clinical necessity for human validation create significant adoption barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not always requiring a licensed physician, this task is typically performed within a clinical setting with quality and liability expectations, and many practices prefer trained personnel for accuracy and patient trust. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated visual acuity systems are expensive capital equipment with ongoing calibration and maintenance costs. When amortized and integrated into a clinic workflow, the per-task cost remains comparable to or higher than a trained technician's loaded wage, especially given the need for human oversight and backup. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Automated acuity testing devices have upfront capital and maintenance costs that may not yet undercut technician labor costs significantly, especially factoring integration and oversight needs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Automated vision-screening devices exist in research and some clinical settings, but they have material limitations in handling diverse patient presentations, special cases (children, low vision patients), and pinhole/dynamic acuity testing. No deployed product reliably performs this task autonomously at production scale in typical ophthalmic clinics. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated visual acuity devices exist (e.g., digital acuity charts, self-administered kiosks) but are not yet widely deployed as full replacements for technician-administered testing in most clinical settings. |
Measure corneal curvature with keratometers or ophthalmometers to aid in the diagnosis of conditions, such as astigmatism.
28CI 25–30 · exposure 25 · augmentation 50 · importance 4.7/5 · click for rater detail
Measure corneal curvature with keratometers or ophthalmometers to aid in the diagnosis of conditions, such as astigmatism.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmology remains a conservative, physician-led sector with slow digital adoption outside major academic centers. Keratometry is a specialized task in a highly regulated clinical environment, limiting rapid uptake of autonomous AI solutions in typical practice settings. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare/ophthalmology adoption of AI-driven diagnostic tools is growing but automation of hands-on measurement tasks remains slow due to hardware, clinical workflow, and regulatory constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist technicians by automating measurement extraction from images, flagging outliers, and speeding data interpretation, moderately improving workflow efficiency while the technician retains quality control and clinical judgment. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Modern automated keratometers and corneal topography systems with software analytics assist technicians by increasing measurement speed and consistency, though the core physical task and patient interaction remain human-performed. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze keratometry images and extract curvature data with reasonable accuracy, the task requires proper instrument positioning, patient alignment, and interpretation of measurements in clinical context. Current systems cannot reliably perform the full end-to-end measurement independently without substantial human oversight, falling short of the 50% time-saving threshold. |
| Task automatability | claude-sonnet-5 | 2/5 | This is a hands-on clinical measurement requiring physical positioning of a patient and instrument alignment; current AI cannot perform the physical measurement, though automated keratometers already exist as instruments (not AI per se) that reduce technician judgment burden slightly. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory bodies like the FDA classify keratometry devices and their calibration as requiring clinical validation and physician oversight. Liability for measurement errors in refractive surgery planning creates strong error-cost asymmetry, and many jurisdictions require a licensed or certified technician to perform or validate corneal measurements. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not always requiring a licensed physician, this measurement typically must be performed by a trained/certified ophthalmic technician per clinical protocols, and diagnostic interpretation is reserved for licensed clinicians, creating moderate procedural and regulatory friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The infrastructure cost (AI software integration, quality assurance, clinical validation) and required operator oversight make current AI solutions comparable to or more expensive than a trained ophthalmic technician performing the measurement directly. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Automated keratometry devices are cost-effective and already integrated into clinics, but they don't eliminate the need for a technician's time, so labor cost savings versus device cost is limited rather than order-of-magnitude cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | AI-assisted analysis of keratometry data exists in research and some specialty software, but no mature deployed product reliably performs unsupervised corneal curvature measurement at scale. Clinical deployment remains limited, with most systems functioning as analysis aids rather than autonomous measurement tools. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated keratometers/corneal topographers are deployed and widely used, but they are dedicated optical/digital instruments rather than AI systems, and still require a trained technician to position the patient and operate/interpret the device. |
Instruct patients in the care and use of contact lenses.
28CI 25–30 · exposure 25 · augmentation 63 · importance 4.3/5 · click for rater detail
Instruct patients in the care and use of contact lenses.
28| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption remains slow because the task is embedded in regulated healthcare delivery and requires human-patient trust and interactive verification; most practices still rely on technicians for this function, and risk aversion in healthcare slows substitution even where technology could assist. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare/vision care settings are generally slower adopters of AI for hands-on patient education compared to purely digital information sectors. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could usefully assist by generating pre-visit instructional videos, providing on-demand reference materials, or helping technicians draft personalized care plans, thereby augmenting the technician's instruction without removing the human element of live demonstration and patient rapport. |
| Augmentation potential | claude-sonnet-5 | 4/5 | AI can generate customized instructional materials, videos, and answer follow-up questions, meaningfully supporting technicians in delivering consistent and thorough patient education while they retain hands-on responsibility. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | AI could draft instructional content or create visual guides for contact lens care, but the task requires real-time responsiveness to individual patient questions, demonstration of proper handling techniques, and adaptive explanation based on observed comprehension—elements that current AI systems cannot reliably perform end-to-end with 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 2/5 | Patient instruction on contact lens care involves hands-on demonstration, checking patient technique, and answering individualized questions, which requires physical presence and adaptive interaction that current AI cannot fully replicate. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Patient safety and regulatory compliance create substantial barriers: contact lens fitting and instruction typically fall under optometric or ophthalmic supervision, and liability for incorrect use (infection, corneal damage) means human professional accountability is legally and ethically required; many jurisdictions require a licensed practitioner to oversee or co-sign patient education. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement mandates a technician perform this specific instruction, but liability concerns around improper lens care leading to eye infections and patient preference for in-person guidance create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of developing, integrating, and maintaining AI systems for patient instruction, plus human oversight for safety and liability, would likely exceed or roughly match the cost of a technician delivering the instruction, especially given low volume per patient. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | While AI-generated educational content is cheap, the hands-on component still requires paid clinical staff time for demonstration and troubleshooting, limiting overall cost savings. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While chatbots and video tutorials exist, they cannot substitute for live instruction that monitors patient dexterity, answers emerging concerns, or adjust explanations based on a patient's specific eye anatomy or cognitive needs; deployed products lack the embodied interaction and safety verification required. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | AI chatbots and instructional videos exist for general contact lens care education, but no deployed product reliably performs hands-on patient training and verification of proper insertion/removal technique in a clinical setting. |
Take anatomical or functional ocular measurements of the eye or surrounding tissue, such as axial length measurements.
25CI 25–25 · exposure 25 · augmentation 50 · importance 4.9/5 · click for rater detail
Take anatomical or functional ocular measurements of the eye or surrounding tissue, such as axial length measurements.
25| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic practices remain moderately digitized and conservative in automation adoption; measurement devices are specialized and operated by certified technicians, with slow uptake of fully autonomous measurement solutions due to clinical validation requirements. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare, especially direct patient-contact clinical measurement tasks, adopts automation more slowly than office/knowledge sectors due to regulatory and safety requirements. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI assists by auto-analyzing acquired images, flagging measurement quality issues, and suggesting values, which can improve technician efficiency and consistency in confirming measurements. However, the human technician remains in the loop for device operation and final validation. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled devices can assist by automating measurement calculations, flagging anomalies, and improving consistency, but the technician remains essential for the physical task and quality control. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can analyze ocular images and measurements, the core task of *taking* anatomical measurements (e.g., axial length via A-scan ultrasound or optical biometry) requires specialized equipment operation and patient positioning that demands real-time judgment and manual adjustment. Current AI cannot operate the physical measurement devices or handle patient variability end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | Taking axial length and other ocular measurements requires precise physical positioning of patients and handling of devices (biometers, tonometers, keratometers) which AI cannot physically perform; only the interpretation/analysis of the resulting data could be augmented. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical and regulatory requirements mandate that qualified personnel operate ophthalmic measurement devices and validate measurements for surgical planning and diagnosis. The task typically requires certification and is tightly integrated with clinical protocols requiring human oversight and liability. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Ocular measurements are part of medical care requiring trained/certified technicians, with liability and accuracy concerns (e.g., pre-surgical biometry for IOL calculation) creating strong barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized ophthalmic measurement equipment (A-scan, optical biometry, topography) is expensive and must be operated by trained staff; AI augmentation does not yet reduce the per-measurement cost below the technician's loaded wage, as human operation remains essential. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Biometry devices are expensive capital equipment requiring a technician to operate them; there is no cheaper AI-only substitute since physical measurement equipment and human oversight are still required. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some AI systems assist in image analysis and measurement interpretation post-acquisition, but no deployed product reliably performs the full measurement-taking task without a technician operating the device and positioning the patient. The equipment operation and quality control remain human-dependent. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Automated ophthalmic devices already capture some measurements semi-automatically, but a trained technician is still required to operate equipment, position patients, and validate readings in production clinical settings. |
Assist patients to select eyewear.
23CI 11–35 · exposure 13 · augmentation 50 · importance 2.7/5 · click for rater detail
Assist patients to select eyewear.
23| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic practices remain moderately digitized; adoption of AI for eyewear selection is nascent, with most optometry and eyewear retail still relying on human technicians for the selection and fitting process. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare and optical retail sectors are slower AI adopters for hands-on patient-facing tasks compared to information/finance sectors, though AR try-on tools are emerging in retail. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by recommending frame styles based on face shape or suggesting inventory matches, helping technicians narrow options and speed up the initial consultation phase, though the human remains central to fit and final selection. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-powered virtual try-on and recommendation tools can help narrow choices or visualize frames, giving some productivity boost to the technician-patient interaction without replacing the human role. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Selecting eyewear requires human judgment about aesthetics, face geometry, comfort, lifestyle needs, and personal preference—all contextual factors that depend on visual fit and subjective comfort that current AI systems cannot reliably assess end-to-end without human involvement. |
| Task automatability | claude-sonnet-5 | 2/5 | This task requires physical fitting, personal interaction, and judgment about facial fit and patient preference that current AI cannot execute end-to-end.ationale simplified |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: patients expect human expertise and personal fitting in ophthalmic care, insurance/liability frameworks assume human technician involvement, and physical fit assessment (bridge width, temple length) requires hands-on adjustment that AI cannot perform. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically for eyewear selection, but patient preference for in-person guidance and fitting accuracy creates practical friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | The cost of AI systems to handle frame recommendation and basic styling (plus oversight) is comparable to or exceeds the labor cost of a technician, especially given the need for integration with inventory and fitting systems. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Any AI assistance (e.g., virtual try-on software) requires additional hardware/software investment and human oversight, offering limited cost savings over a technician who already handles multiple tasks during the visit. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI can provide basic frame recommendations via chatbots or image analysis, no deployed product reliably handles the full selection process independently; most real-world adoption still requires human technician judgment on fit and adjustment. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product physically assists patients in selecting and fitting eyewear in clinical settings; virtual try-on tools exist but are consumer-facing retail add-ons, not integrated into ophthalmic technician workflows. |
Conduct binocular disparity tests to assess depth perception.
19CI 16–21 · exposure 20 · augmentation 38 · importance 3.6/5 · click for rater detail
Conduct binocular disparity tests to assess depth perception.
19| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic practices remain traditional and conservative in automation adoption, with most clinics still relying on technician-administered equipment. Digitization of eye care diagnostics is slower than in information-intensive sectors, with limited evidence of AI agent deployment in production. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare diagnostic testing involving direct patient contact has seen slow AI adoption for physical exam tasks, though software-assisted diagnostics are growing in adjacent areas like imaging analysis. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist by automating data logging, flagging anomalous test results, or suggesting borderline findings for technician review, improving workflow efficiency. However, the core task of administering the test and interpreting immediate patient responses still centers on human skill. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with recording, tracking, or flagging abnormal results but offers limited direct enhancement to the physical administration of the test itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Binocular disparity tests involve specialized optical equipment and require interpretation of patient responses that depend on precise calibration and individual variability. While AI could assist in data collection and analysis, the task requires real-time patient interaction, equipment adjustment, and clinical judgment that current systems cannot reliably perform end-to-end. |
| Task automatability | claude-sonnet-5 | 2/5 | This requires physical patient interaction, positioning, and precise use of specialized equipment (e.g., stereopsis tests) that current AI cannot perform hands-on; only minor scoring/interpretation subcomponents could be automated.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Clinical eye care is heavily regulated and typically requires licensed personnel (optometrists or ophthalmologists) to oversee diagnostic procedures. Liability and patient safety concerns create strong friction against fully automated depth perception assessment without professional sign-off. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Clinical vision testing typically requires a trained/certified technician performing hands-on assessment as part of a regulated diagnostic workflow, creating strong practical and quasi-regulatory barriers to full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The specialized optical equipment, calibration requirements, and need for human oversight make automation economically unviable. Current ophthalmic AI tools are expensive relative to the modest labor cost of a technician performing this routine procedural task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | Without a viable automated substitute, AI cannot replace the human-performed physical test, so cost comparison favors the human technician entirely. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | No deployed AI system currently conducts binocular disparity tests independently in clinical settings. While computer vision could theoretically analyze test results, the actual administration—positioning patients, adjusting equipment, interpreting subjective responses—remains dependent on human technicians in production environments. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product autonomously administers binocular disparity/stereopsis testing in clinical settings today; this remains a manual clinical procedure performed by technicians. |
Conduct ocular motility tests to measure function of eye muscles.
16CI 7–25 · exposure 13 · augmentation 38 · importance 4.6/5 · click for rater detail
Conduct ocular motility tests to measure function of eye muscles.
16| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Ophthalmology has moderate AI adoption (e.g., retinal image analysis), but automation of motility testing specifically remains rare in production; most practices still rely on technician-administered tests, reflecting slow uptake of autonomous systems for this particular procedure. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare diagnostic testing involving physical patient examination has seen slow AI adoption compared to information-processing tasks, with most AI use confined to image analysis rather than the physical test administration. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted eye-tracking and automated saccade/pursuit measurement could help a technician flag abnormalities faster and reduce manual chart entry, modestly raising productivity without removing the need for clinical judgment and patient interaction. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with recording, analyzing eye movement data, or flagging abnormalities from video capture, but the core physical test administration and patient interaction remain unaided by AI today. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Ocular motility testing requires measuring eye movements and alignment via specialized equipment (cover tests, saccade tracking, pursuit movements). While AI can analyze video of eye movements post-hoc, current systems cannot reliably administer the full battery of tests, position patients correctly, interpret subtle clinical signs, or handle exceptions in real time—falling short of 50% time savings at equal quality. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires hands-on physical examination of a patient's eye movements using specialized instruments and direct observation, which current AI cannot perform end-to-end without robotic embodiment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Ocular motility testing is typically part of a comprehensive eye exam that may require a licensed professional to supervise or sign off on clinical findings; patient contact, physical positioning, and interpretation in context of clinical history create both regulatory and practical barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is a clinical diagnostic procedure typically requiring trained/certified personnel under physician oversight, with direct patient contact and liability considerations limiting substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Specialized ophthalmic equipment (eye trackers, automated refractors) and clinical oversight remain expensive; AI inference cost is modest but integration, calibration, and required human supervision make the total cost per test comparable to or higher than a technician performing it directly. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI system replacing this hands-on task, so AI is not currently cheaper than the human technician performing it. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Some computer vision systems can track eye gaze and detect gross movement abnormalities in controlled settings, but no deployed clinical product reliably performs comprehensive ocular motility assessment independent of a technician. Research prototypes exist, but production systems with medical-grade accuracy and liability clearance are not standard in practice. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts ocular motility testing on patients; this remains a manual clinical procedure performed by trained technicians. |
Clean or sterilize ophthalmic or surgical instruments.
15CI 5–25 · exposure 13 · augmentation 25 · importance 4.6/5 · click for rater detail
Clean or sterilize ophthalmic or surgical instruments.
15| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Most ophthalmic practices and surgical centers are small to mid-sized, digitization-limited organizations with slower adoption of advanced automation; while basic sterilizers are ubiquitous, intelligent robotic cleaning and verification remain niche. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare support and physical instrument handling tasks show minimal AI/robotic adoption; this is a low-digitization, hands-on task with no meaningful automation trend. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | Automated washers and sterilization systems do assist technicians by handling bulk cleaning and standardized cycles, reducing manual labor; however, the assistance is mainly on the mechanical parts, not on the quality control and sorting that remain labor-intensive. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no assistance for the physical acts of cleaning or sterilizing instruments; software tracking systems may log compliance but do not aid the task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While automated instrument washers and sterilization equipment exist, the task involves variable cleaning needs based on instrument type and contamination level, plus quality verification that requires human judgment. Current AI/robots cannot reliably handle the full end-to-end workflow with 50% time savings at equal safety outcomes. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manual task requiring handling, cleaning, and sterilization of delicate instruments; no current AI system can perform this end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Instrument sterilization is heavily regulated by FDA, OSHA, and clinical standards that mandate documented procedures and accountability; liability for contamination or instrument damage falls on the facility, creating strong legal and compliance barriers to full automation without human sign-off. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Infection control and sterilization protocols are regulated (e.g., by health authorities), requiring trained personnel to follow validated procedures, creating significant compliance and liability barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Medical-grade autoclaves and ultrasonic cleaners exist but require significant upfront capital investment, maintenance, and integration; the loaded wage of an ophthalmic technician is modest, making full automation cost-competitive only at large scale. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute, so any AI-based approach (e.g., robotic sterilization) would require costly specialized hardware exceeding current human labor costs. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Automated washers and sterilizers are deployed in medical settings, but they perform only part of the cleaning pipeline; human technicians must sort, inspect, pack, and verify. No end-to-end system reliably replaces the full task without human oversight. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI or robotic product performs ophthalmic instrument cleaning/sterilization in clinical settings today; this remains a manual staff duty. |
Maintain ophthalmic instruments or equipment.
12CI 5–19 · exposure 8 · augmentation 38 · importance 4.5/5 · click for rater detail
Maintain ophthalmic instruments or equipment.
12| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Ophthalmic practices remain traditionally managed with human technicians in physical clinical settings; digitization is limited and adoption of autonomous maintenance automation in these settings is minimal, with strong preference for certified human specialists. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Physical equipment upkeep in healthcare clinical settings is a low-digitization, hands-on domain with minimal AI/robotic adoption for this specific task. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI could assist through automated maintenance scheduling, parts inventory management, diagnostic suggestions for equipment faults, and documentation support, providing useful but not transformative productivity gains to human technicians performing the physical work. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-based diagnostic software or IoT sensors could flag maintenance needs or track calibration schedules, offering minor assistance, but the physical maintenance work itself sees little AI augmentation. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Routine maintenance checklists and record-keeping could be partially automated, but ophthalmic instruments require calibration, alignment, and troubleshooting that demand specialized technical knowledge and hands-on physical work that current AI cannot perform end-to-end. |
| Task automatability | claude-sonnet-5 | 1/5 | Maintaining physical ophthalmic instruments (cleaning, calibration, sterilization, repair) is a hands-on physical task that current AI cannot perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Regulatory oversight (FDA) applies to ophthalmic equipment maintenance and records; clinical environments require validated, auditable procedures; and equipment manufacturers often specify technician qualifications and certification requirements for warranty and liability purposes. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing law mandates a specific credentialed professional for equipment maintenance, but manufacturer protocols, calibration certification, and liability for patient-facing device accuracy create real organizational friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The labor cost of a trained ophthalmic technician is relatively modest compared to the cost of AI systems that would need to be integrated into clinical workflows, plus the cost of any robotic or physical automation apparatus required. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for physical maintenance, so AI cost cannot be compared favorably to a human technician's wage for this task. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI products reliably perform the full maintenance of delicate ophthalmic equipment in real clinical settings; while monitoring and alerting systems exist, actual maintenance requires human technician intervention with specialized tools and domain expertise. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical equipment maintenance in clinical ophthalmology settings; this remains a manual technician task. |
Assist physicians in performing ophthalmic procedures, including surgery.
10CI 0–20 · exposure 8 · augmentation 50 · importance 4.7/5 · click for rater detail
Assist physicians in performing ophthalmic procedures, including surgery.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Adoption of surgical robotic assistants in ophthalmology remains limited to large academic centers and specialized practices; the broader sector of community eye clinics and smaller surgical centers has seen minimal AI-assisted automation of technician roles. Current pace is slow relative to information-sector adoption. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare surgical support roles show minimal AI-driven displacement; adoption in this specific physical assistive function is essentially nonexistent. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted image analysis and surgical planning tools can enhance pre-operative preparation and post-operative review, improving technician efficiency in those phases. However, intraoperative augmentation remains limited, as real-time surgical AI assistance is not yet standard in ophthalmology. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can assist with pre-op imaging analysis, scheduling, and documentation surrounding the procedure, but does not meaningfully augment the physical in-procedure assistance itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Ophthalmic surgery assistance requires real-time coordination, precise instrument handling, and adaptive response to patient physiology—tasks that demand embodied presence and continuous human judgment. Current AI cannot reliably perform the core procedural-support functions that define this role, though it could assist with preparation and documentation. |
| Task automatability | claude-sonnet-5 | 1/5 | Physical, hands-on surgical assistance requiring real-time manual dexterity, sterile technique, and instrument handling cannot be performed by current AI systems. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Ophthalmic surgery carries high liability for error, and patient safety requires real-time human judgment and accountability. Regulatory frameworks (FDA, state medical boards) and malpractice liability strongly incentivize human presence and signature on procedures, creating substantial legal and organizational barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Surgical assistance involves direct patient contact, sterile field protocols, and often certification/licensure requirements, with high liability for errors—strong regulatory and professional barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Robotic surgical assistants exist but cost hundreds of thousands of dollars upfront and require specialized training, making them more expensive than hiring trained technicians when overhead and integration are factored in. The cost asymmetry strongly favors human technicians except in high-volume surgical centers. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for this physical task, so any comparison of inference cost to human wage is moot—AI cannot deliver the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably assists physicians during ophthalmic surgery itself. While AI tools exist for image analysis and pre-operative planning, the intraoperative assistance—positioning equipment, anticipating surgeon needs, managing instruments—remains outside the scope of any production system. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical surgical assistance duties; robotic surgery aids exist but operate under direct surgeon control, not as autonomous assistants replacing technicians. |
Assess refractive conditions of eyes, using retinoscopes.
10CI 7–13 · exposure 5 · augmentation 25 · importance 4.1/5 · click for rater detail
Assess refractive conditions of eyes, using retinoscopes.
10| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | While autorefractors are widely adopted in many clinics, they supplement rather than replace retinoscopy, and manual retinoscopy remains a core clinical skill; adoption of full automation of this task is very slow in actual practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Ophthalmic clinical settings adopt digital diagnostic tools like autorefractors but full replacement of manual retinoscopy skills via AI/robotics is not part of current sector adoption trends. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Digital displays or AI-assisted measurement feedback during retinoscopy might aid interpretation of results, but current systems offer minimal enhancement to the core manual task of performing the procedure itself. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI-based autorefractors and decision-support tools can complement refractive assessment, but they don't specifically enhance the manual retinoscopy task itself in a transformative way. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Retinoscopy requires manual manipulation of a handheld instrument, precise observation of pupillary reflex in a moving target (patient's eye), and real-time clinical judgment about lens prescription. Current AI cannot reliably perform this hands-on procedure end-to-end. |
| Task automatability | claude-sonnet-5 | 1/5 | Retinoscopy requires hands-on manipulation of a handheld instrument, real-time observation of the reflex, and physical positioning relative to the patient's eye, none of which current AI systems can perform end-to-end today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Retinoscopy is legally and clinically the domain of licensed ophthalmologists, optometrists, or certified ophthalmic technicians under direct supervision; regulatory and professional licensing requirements are high, and liability for incorrect refraction is substantial. |
| Adoption barriers | claude-sonnet-5 | 4/5 | This is a clinical diagnostic procedure typically requiring trained/certified personnel and involves direct patient contact and instrument use near the eye, creating strong safety, liability, and training barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Autorefractors (the closest automation proxy) still require technician oversight and are capital-intensive; the labor cost of a trained ophthalmic technician performing retinoscopy remains lower than system acquisition, maintenance, and integration when all-in costs are considered. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI system that performs this physical diagnostic task, so cost comparison favors the human entirely; any automation would require expensive robotics not yet available. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Automated refraction systems exist (e.g., autorefractors) but they operate differently and produce different data than manual retinoscopy; no deployed AI system can operate a retinoscope or replicate the clinical interpretation of retinoscopic findings as a technician would in real time. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously performs retinoscopy; this remains a manual clinical skill performed by trained technicians with no robotic or AI substitute in production. |
Conduct tonometry or tonography tests to measure intraocular pressure.
8CI 0–16 · exposure 8 · augmentation 25 · importance 4.9/5 · click for rater detail
Conduct tonometry or tonography tests to measure intraocular pressure.
8| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Ophthalmology remains a hands-on clinical specialty with slow adoption of automation in diagnostic procedures; no measurable market penetration of AI-based autonomous tonometry exists in practice. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Healthcare diagnostic procedures involving direct patient contact adopt automation slowly; while some tonometers have automated features, full replacement of the technician is not occurring. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist by automatically interpreting tonometry results or flagging anomalies in pressure data post-test, but offers minimal help during the actual test execution, which remains technician-driven. |
| Augmentation potential | claude-sonnet-5 | 2/5 | Modern non-contact tonometers and automated instruments assist the technician by simplifying readings, but this is instrument automation rather than AI-driven augmentation of decision-making or workflow. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | Tonometry and tonography require precise physical contact with the patient's eye and calibration of specialized equipment. While image analysis of pressure readings could be partially automated, the actual test execution—positioning, patient cooperation, device contact, and real-time adjustment—remains fundamentally manual and cannot meet the 50% time-saving threshold without human control. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires direct physical instrument contact with or proximity to the patient's eye and manual dexterity/patient management; no current AI system can perform the hands-on measurement procedure. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | These diagnostic tests involve direct patient contact and medical liability; in most jurisdictions, only licensed ophthalmic technicians or physicians may perform tonometry, creating a hard legal and professional licensing barrier to automation or delegation to non-human systems. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct patient contact with a diagnostic device touching or interacting with the eye typically requires trained/certified personnel and carries liability risk for eye injury, creating strong practical barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The equipment, calibration, patient interface, and skilled technician labor involved mean AI automation would require custom robotics and integration costs far exceeding the loaded wage of an ophthalmic technician for this specific task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute performing this physical clinical task, so cost comparison favors the human technician who must be present regardless. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI system can independently perform tonometry or tonography; these remain hands-on clinical procedures requiring direct patient interaction and specialized medical devices. Current AI has no production role in executing these tests end-to-end. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product autonomously conducts tonometry/tonography without a human technician operating the device; automated tonometers exist but still require a human to position and operate them on the patient. |
Adjust or make minor repairs to spectacles or eyeglasses.
7CI 5–10 · exposure 0 · augmentation 13 · importance 2.7/5 · click for rater detail
Adjust or make minor repairs to spectacles or eyeglasses.
7| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Ophthalmic retail and clinics are not early adopters of task automation; digitization is limited to scheduling and record-keeping. The distributed, low-volume nature of eyeglass repair in small optical shops and practices slows any automation adoption. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Optical dispensing and repair remains a low-digitization, hands-on retail/clinical service with minimal AI or robotics penetration. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could potentially assist by identifying common adjustment patterns or recommending frame modifications based on image analysis, but current tools offer minimal support. The technician's hands-on judgment and real-time customer feedback remain central to the task. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI provides negligible assistance for physical frame adjustment or repair, as this is a manual craft task with no significant digital or cognitive component to augment. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation of eyewear frames and lenses with precision fitting to individual face geometry and comfort preferences—work that demands dexterous robotic systems and real-time visual feedback in an unstructured environment. Current AI and automation cannot reliably perform end-to-end frame adjustment or minor lens repairs without human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manipulation task involving frame adjustment, screw tightening, and lens fitting that requires manual dexterity and tactile feedback AI cannot replicate today.atable |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Customer safety, product liability (defective adjustments causing discomfort or damage), and strong consumer preference for a trained human technician to verify fit create substantial friction against automation. Many optical retailers also position fitting as a personalized service. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically bars automation, but the task requires physical presence, tactile judgment, and customer interaction that create practical friction against remote or robotic substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The capital cost of precision robotic systems capable of handling delicate eyewear, plus integration and maintenance, far exceeds the modest labor cost of a technician performing these quick adjustments (typically 10–30 minutes per customer). |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI/robotic solution for this physical task, so any comparison to human labor cost favors the human by default. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial product performs eyeglass adjustment or minor repair autonomously. This task remains human-only in ophthalmic clinics and retail settings; no production-grade automation exists. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical spectacle repair; this remains a purely manual craft task performed by technicians with hand tools. |
Assist patients to insert or remove contact lenses.
5CI 5–5 · exposure 0 · augmentation 13 · importance 3.9/5 · click for rater detail
Assist patients to insert or remove contact lenses.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This task occurs in healthcare settings with established workflows, conservative adoption of physical automation, and strong preference for human technicians who can reassure and adapt to patient needs; no sector-wide adoption of automated lens insertion/removal exists. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare physical/hands-on tasks show minimal AI adoption; this specific manual procedure has no automation trend in clinical practice. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could potentially assist by providing real-time patient guidance, lens positioning feedback, or virtual try-on preview, but the hands-on manipulation itself must remain human-controlled; limited scope for meaningful augmentation. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers no meaningful assistance for the physical act of helping a patient insert or remove contact lenses. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Contact lens insertion/removal involves fine motor manipulation of a physical object near the eye, requiring real-time tactile feedback and adaptation to individual anatomy. Current AI lacks embodied robotic capabilities to perform this reliably and safely on patients. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical task requiring direct manipulation of a patient's eye, which no current AI system can perform end-to-end.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist due to liability exposure (eye injury risk), patient safety requirements, regulatory oversight of medical devices, and the implicit human-contact requirement for patient comfort and trust during an intimate eye procedure. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical patient contact with the eye involves safety, hygiene, and liability concerns that require a trained human present, though not necessarily a licensed physician. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Any hypothetical robotic system capable of this task would require significant hardware investment, maintenance, and real-time computer vision infrastructure—far exceeding the cost of a trained technician performing the task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI equivalent performing this physical task, so any AI cost comparison is moot; the human technician remains the only viable option. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial system can autonomously insert or remove contact lenses from a patient's eye. This task requires physical dexterity, real-time sensory feedback, and patient cooperation that current AI-powered robotics cannot reliably achieve in clinical settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI or robotic product performs contact lens insertion/removal on patients; this remains purely manual clinical work. |
Administer topical ophthalmic or oral medications.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.8/5 · click for rater detail
Administer topical ophthalmic or oral medications.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Healthcare remains a laggard sector for full task automation due to regulatory constraints, liability concerns, and the requirement for human-patient contact and judgment in clinical settings. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Healthcare direct patient care tasks involving physical medication administration show minimal AI adoption due to physical, regulatory, and safety constraints. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist by flagging drug interactions or reminding technicians of proper administration protocols, but the core task of physically administering medication offers limited augmentation opportunity since human execution remains mandatory. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could assist with medication reminders, dosage calculations, or documentation, but offers little direct assistance to the physical act of administering medication. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Administering medications requires direct patient contact, precise physical application (especially topical eye drops), and real-time assessment of patient response. Current AI cannot perform the physical manipulation or safely handle the medication delivery without human oversight. |
| Task automatability | claude-sonnet-5 | 1/5 | Physically administering medications to a patient's eye or orally requires direct physical manipulation and patient interaction that current AI systems cannot perform, as they lack embodiment.' |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medication administration is heavily regulated; healthcare licensing laws, liability frameworks, and controlled substance regulations require a licensed or credentialed human to perform or directly supervise drug administration. Legal and regulatory barriers are substantial. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Administering medications is a regulated clinical act typically requiring licensed or certified personnel, with direct liability for patient safety and legal scope-of-practice restrictions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The cost of developing a robotic system capable of safe medication administration, plus required validation and oversight, would far exceed the loaded wage of an ophthalmic technician performing this routine task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for physical medication administration, so cost comparison favors the human by default since AI cannot perform the task at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products perform medication administration independently. This task requires physical interaction with patients and controlled substances, which remain exclusively within the human technician domain in clinical practice. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product administers topical or oral medications to patients; this remains a hands-on clinical task performed by trained personnel. |
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.