Home Health Aides
31-1121.00Rank #869 of 923 scored · top 94% by substitution
Monitor the health status of an individual with disabilities or illness, and address their health-related needs, such as changing bandages, dressing wounds, or administering medication. Work is performed under the direction of offsite or intermittent onsite licensed nursing staff. Provide assistance with routine healthcare tasks or activities of daily living, such as feeding, bathing, toileting, or ambulation. May also help with tasks such as preparing meals, doing light housekeeping, and doing laundry depending on the patient's abilities.
Sub-scores
0–100 · band = confidence interval from rater disagreement
Substitution11
Exposure7
Augmentation32
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
15 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%8
panel mean rating 1.3/5 → substitution pressure 8/100
Technical feasibility todayw 20%6
panel mean rating 1.2/5 → substitution pressure 6/100
Cost vs. human wagew 15%10
panel mean rating 1.4/5 → substitution pressure 10/100
Adoption barriersw 20%inverted — strong barriers lower the score24
panel mean rating 4.0/5 (barrier strength) → substitution pressure 24/100
Sector adoption velocityw 10%4
panel mean rating 1.2/5 → substitution pressure 4/100
Task breakdown (15 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.
Entertain, converse with, or read aloud to patients to keep them mentally healthy and alert.
35CI 28–43 · exposure 17 · augmentation 50 · importance 4.1/5 · click for rater detail
Entertain, converse with, or read aloud to patients to keep them mentally healthy and alert.
35| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Home health is a fragmented, low-digitization sector dominated by small independent providers and public/non-profit agencies with limited tech budgets; adoption of AI companions remains pilot-stage despite years of development, reflecting both sector conservatism and resistance to replacing human care presence. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, high-touch physical sector with slow AI adoption for direct patient interaction tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI-assisted reading (text-to-speech with adaptive pacing) and conversational prompts could help a human aide manage multiple patients or identify topics of interest, modestly raising efficiency, though the core therapeutic value depends on human presence and empathy. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can supplement with music, audiobooks, or conversational apps to give aides tools for engaging patients, but it only partially assists rather than transforming the task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires genuine emotional engagement, adaptive conversation based on patient preferences and cognitive state, and the ability to read subtle social cues—capabilities that current AI systems cannot reliably deliver in a way that provides authentic mental health benefit to vulnerable patients. |
| Task automatability | claude-sonnet-5 | 2/5 | AI voice companions can converse or read aloud, but genuine social engagement, emotional attunement, and physical presence for a vulnerable patient are not fully replaceable end-to-end today.5.0 |
| Adoption barriers | claude-haiku-4-5-20251001 | 2/5 | While there is no legal licensing requirement for this specific task, significant organizational and ethical friction exists: patient and family preference for human contact, liability concerns if isolation increases, and regulatory/accreditation standards in many care settings that expect human interaction as part of quality care. |
| Adoption barriers | claude-sonnet-5 | 2/5 | No licensing requirement specifically covers casual conversation/entertainment, but patient preference for human contact and the interpersonal nature of caregiving create moderate friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 4/5 | An AI system (chatbot + text-to-speech) deployed on a tablet or smart device costs far less per hour than the loaded wage of a home health aide ($25–35/hour fully loaded), with marginal deployment costs near zero once built. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Voice assistants and reading apps are cheap per interaction, but they don't replace the human aide who is already present for other duties, so marginal cost savings are limited. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | While AI chatbots and text-to-speech systems exist, no deployed product reliably replaces a human caregiver's ability to provide meaningful entertainment, personalized conversation, and therapeutic presence at the quality level needed for isolated or cognitively vulnerable patients. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Some companion chatbots and smart speakers exist for eldercare entertainment, but they are narrow, novelty-stage deployments rather than reliable substitutes for human conversation in home health settings. |
Maintain records of patient care, condition, progress, or problems to report and discuss observations with supervisor or case manager.
29CI 25–34 · exposure 25 · augmentation 50 · importance 4.5/5 · click for rater detail
Maintain records of patient care, condition, progress, or problems to report and discuss observations with supervisor or case manager.
29| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Healthcare, especially home health, lags in AI adoption due to regulatory conservatism, fragmented IT systems, and liability concerns; while pilot documentation aids exist, production deployment of autonomous record-keeping systems remains rare in this sector. |
| Sector adoption velocity | claude-sonnet-5 | 2/5 | Home health care is a low-digitization, high physical-contact sector with fragmented small employers and limited IT infrastructure, resulting in slow AI tool adoption compared to office-based professional services. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist by suggesting what observations to record, summarizing patient notes, or flagging potential issues for human review, but the aide must ultimately synthesize clinical judgment and validate all entries. This co-creation raises productivity moderately without replacing the human's decision-making role. |
| Augmentation potential | claude-sonnet-5 | 3/5 | Dictation apps and structured EHR templates can meaningfully speed up documentation and standardize reporting, giving moderate productivity benefits while the aide remains responsible for accuracy and judgment. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can help draft notes and organize observations, the task requires nuanced judgment about which patient details are clinically significant and contextual understanding of individual patient trajectories that current AI struggles with reliably. The end-to-end task of deciding what to record, synthesizing observations, and determining what warrants supervisor escalation involves domain expertise and legal liability that prevents 50% time savings at equal quality today. |
| Task automatability | claude-sonnet-5 | 2/5 | AI can help draft or structure notes from dictated observations, but the actual observation, judgment about patient condition, and interpersonal reporting to a supervisor require in-person human presence and cannot be fully automated end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Healthcare records are legally protected (HIPAA) and clinical documentation carries liability; most jurisdictions require a licensed healthcare professional to verify, sign, or take responsibility for patient care records. Regulatory and legal requirements create hard barriers to full AI substitution. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing barrier for basic recordkeeping, but liability concerns around accurate health documentation and requirement for a responsible caregiver to verify and report information create meaningful friction. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Current AI documentation tools require significant infrastructure integration, oversight, and error-correction by clinical staff, making the all-in cost per task-equivalent comparable to or higher than a home health aide's wage for this portion of their work. Liability exposure and quality assurance overhead further inflate costs. |
| Cost vs. human wage | claude-sonnet-5 | 3/5 | Speech-to-text and simple documentation assistants are cheap relative to aide wages, but human aides still must perform the underlying observation and judgment, so cost savings are partial rather than full task replacement. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Electronic health record (EHR) systems exist but primarily assist with data entry rather than autonomous record-keeping; clinical documentation tools struggle with capturing context-dependent patient observations and lack reliable output for high-stakes healthcare settings. No deployed product reliably performs independent clinical record synthesis and supervisor flagging without human oversight and correction. |
| Technical feasibility today | claude-sonnet-5 | 2/5 | Voice-to-text and note-summarization tools exist and are used in some home care agencies, but reliable, widely deployed systems that autonomously handle patient condition documentation and communication with case managers are not yet standard in this low-tech workforce. |
Check patients' pulse, temperature, and respiration.
24CI 18–30 · exposure 20 · augmentation 63 · importance 3.9/5 · click for rater detail
Check patients' pulse, temperature, and respiration.
24| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Home health is a fragmented, low-digitization sector with many small agencies and in-home settings; while remote monitoring pilots exist, production adoption of AI-driven vital-sign automation remains limited and slow compared to information-sector automation. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically-embedded, small-provider-dominated sector with minimal AI/robotic adoption for direct patient care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 4/5 | Wearable and ambient monitoring devices substantially augment aides by providing continuous or on-demand vital-sign data, alerts, and documentation support, allowing them to focus on patient care and intervention rather than manual measurement and record-keeping. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled wearable sensors and monitoring apps can assist aides by continuously tracking vitals and flagging anomalies, improving efficiency and safety even though the human still performs manual checks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While remote monitoring devices can automate pulse, temperature, and respiration measurement, a home health aide still must physically position equipment, interpret readings in context, and ensure patient cooperation—core human elements that current AI cannot fully replace end-to-end without substantial human oversight. |
| Task automatability | claude-sonnet-5 | 2/5 | Requires hands-on physical measurement of vital signs on a person, which current AI systems cannot perform without embodied robotic capability; only the interpretation of readings could be automated, not the measurement itself. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Home health aide services are often mandated by insurance and regulatory frameworks that require a licensed or certified human to be physically present; medical assessment of vital signs and patient safety responsibility typically falls on the human aide, creating strong legal and licensing barriers to full automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for taking vital signs, but home health care involves physical human contact, trust, and safety monitoring that create practical and liability-related friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | Automated vital-sign monitors have high upfront hardware and setup costs; the ongoing cost per check (device maintenance, integration, oversight) remains comparable to or exceeds the time cost of a home health aide taking manual measurements, especially at scale across many patients. |
| Cost vs. human wage | claude-sonnet-5 | 2/5 | Wearable/IoT vital sign monitors have upfront and maintenance costs and still typically require human oversight or intervention, so total cost is not clearly cheaper than a home health aide performing this task manually. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Standalone vital-sign monitoring devices exist and work reliably, but they require human placement and interpretation; no deployed AI system independently performs the full task (patient contact, device application, reading documentation) without human intervention in real home-health workflows. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product physically checks a patient's pulse, temperature, and respiration autonomously in home care settings today; wearable sensors exist but require separate hardware and setup, not general AI action. |
Plan, purchase, prepare, or serve meals to patients or other family members, according to prescribed diets.
18CI 5–30 · exposure 13 · augmentation 50 · importance 4.0/5 · click for rater detail
Plan, purchase, prepare, or serve meals to patients or other family members, according to prescribed diets.
18| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 2/5 | Home health is a fragmented, labor-intensive sector with low digitization and high reliance on in-person care. Adoption of AI for meal planning is modest; the physical and service aspects remain stubbornly human-dependent. Pilot programs exist but production-scale automation is minimal. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically intensive sector with minimal AI agent deployment for hands-on caregiving tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can meaningfully assist by suggesting compliant menus based on patient restrictions, automating grocery lists, and flagging nutritional imbalances—raising a human aide's productivity in the planning and purchasing phases. However, augmentation in preparation and serving is limited given the manual and interpersonal nature of those tasks. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI can meaningfully assist with meal planning by generating diet-compliant menus, grocery lists, and recipe suggestions based on prescribed dietary restrictions, even though it cannot perform the physical tasks. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could assist with meal planning and purchasing logistics, the physical preparation and serving of meals requires human presence and dexterity. Current systems can suggest menus based on dietary restrictions but cannot end-to-end replace the hands-on cooking, plating, and personalized service aspects of this task. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical presence to shop for groceries, cook, and serve food in a client's home, none of which current AI systems can perform end-to-end without robotic embodiment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Significant barriers exist: patients often require personalized attention during meals for safety and comfort, dietary compliance is medically supervised and may require professional judgment, and liability concerns arise if a non-human system errs on dietary restrictions or food safety. Many jurisdictions and care standards expect human caregivers for meal service in healthcare contexts. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No strict licensing requirement for meal prep itself, but dietary compliance for medical conditions and in-home trust/safety expectations create moderate friction against full automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-assisted meal planning and grocery ordering are cheaper than human labor for those components alone, but the cost of integrating such systems with the physical execution (food prep, serving) and the oversight required for dietary compliance makes the overall solution only marginally cheaper than employing a home health aide. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot perform the physical shopping, cooking, and serving components, so there is no viable AI-only cost comparison; a human aide remains necessary for all physical labor. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Deployed products exist for meal planning and diet management software, but no end-to-end system reliably performs the full workflow (planning, purchasing, preparing, serving) autonomously. The preparation and serving components remain beyond current AI and robotics capabilities in real home settings. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product plans, purchases, prepares, and serves meals in a home setting; meal-planning apps exist but the physical execution remains entirely human. |
Provide patients and families with emotional support and instruction in areas such as caring for infants, preparing healthy meals, living independently, or adapting to disability or illness.
14CI 5–23 · exposure 13 · augmentation 38 · importance 3.9/5 · click for rater detail
Provide patients and families with emotional support and instruction in areas such as caring for infants, preparing healthy meals, living independently, or adapting to disability or illness.
14| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health services remain labor-intensive, fragmented, and concentrated among small providers and individuals with limited digitization. Adoption of AI for core instruction and emotional support remains negligible; most innovation targets administrative tasks rather than direct patient interaction. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, high-touch physical sector with minimal AI agent deployment for direct care and emotional support tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI can assist home health aides by generating customized meal plans, care instructions, or resource guides based on patient conditions, freeing time for direct emotional support and hands-on care. However, the augmentation is limited to informational components; the core relational work remains human-dependent. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could provide reference material or generate educational content (e.g., meal plans, care tips) that aides might use, but it offers limited direct enhancement to the emotional support component of this task. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI can generate instructional content on infant care, meal preparation, and disability adaptation, the emotional support component and real-time responsiveness to family dynamics requires human presence and judgment. Current AI lacks the embodied, relational capabilities and contextual understanding needed to deliver meaningful emotional support at scale. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical presence, genuine human empathy, and relationship-based emotional support that cannot be delivered end-to-end by current AI systems. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Home health aide roles are regulated in many jurisdictions and often require certification; care tasks involving vulnerable populations (infants, elderly, disabled) face liability and duty-of-care constraints. The requirement for human contact and trust—especially for emotional support—creates strong organizational and regulatory barriers to substitution. |
| Adoption barriers | claude-sonnet-5 | 4/5 | While not always formally licensed, the emotional/interpersonal nature of care, liability concerns, and strong human-contact expectations create high resistance to substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 2/5 | AI-generated instructional materials are inexpensive, but they require human oversight, customization, and emotional labor integration to match the value delivered by home health aides. The loaded cost of a home health aide is modest relative to the full-spectrum service (instruction plus emotional presence), making AI cost-parity difficult to achieve. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | AI cannot substitute for the human presence and trust-building required, so any AI-only alternative would fail to deliver the service at all, making cost comparison favor the human. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 2/5 | Chatbots and educational apps exist to provide some instructional content, but they operate in narrow domains and lack the ability to assess family circumstances, adapt advice to individual needs, or provide genuine emotional comfort. No deployed product reliably performs the full task of emotional support plus tailored instruction in real home environments. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product provides in-home emotional support and hands-on instruction to patients and families; this remains firmly in the human domain. |
Perform a variety of duties as requested by client, such as obtaining household supplies or running errands.
12CI 5–19 · exposure 8 · augmentation 25 · importance 3.7/5 · click for rater detail
Perform a variety of duties as requested by client, such as obtaining household supplies or running errands.
12| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | This sector is characterized by small independent providers, low technology adoption, high reliance on personal relationships, and regulatory constraints. Home health remains one of the least digitized care sectors with minimal AI deployment. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically grounded sector with minimal AI/robotic adoption for hands-on errand tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could marginally assist by suggesting errands to consolidate, optimizing shopping lists, or providing navigation support, but these are peripheral to the core task. The human aide's judgment about client needs and execution of errands remains central and largely unaugmented by current AI. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI apps (e.g., grocery delivery, list-making, scheduling assistants) can help organize or expedite errands, but they don't materially transform the aide's core physical task execution. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | This task involves physical presence and real-world errand execution (shopping, delivery) that current AI agents cannot perform autonomously. While AI can help plan errands or optimize routes, the core work—obtaining supplies, interacting with merchants, handling cash/cards—remains firmly human. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical presence, mobility, and manual interaction with the physical world (shopping, running errands, fetching items) that current AI systems cannot perform without embodiment. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Strong barriers exist: home health aides require state licensure and certification in most jurisdictions, involve direct client contact and trust relationships, and handle vulnerable populations (elderly, disabled). Legal liability and care standards create hard adoption friction. |
| Adoption barriers | claude-sonnet-5 | 3/5 | No licensing requirement specifically bars automation of errands, but personal trust, safety, and physical-world reliability create real friction against non-human execution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI cannot yet substitute for this task at any cost advantage because it requires physical presence and execution in unstructured environments. The loaded labor cost of a home health aide remains far cheaper than the infrastructure needed for autonomous task completion. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing physical errands, so any comparison favors the human aide entirely; AI cannot deliver the output at all, let alone cheaply. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can reliably execute household errands end-to-end in the real world. Current systems lack embodiment, real-time navigation in unfamiliar environments, and ability to handle physical transactions and unexpected variations. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product autonomously runs errands or physically obtains household supplies for a client; this remains outside current robotics/AI product capability at scale. |
Direct patients in simple prescribed exercises or in the use of braces or artificial limbs.
9CI 5–14 · exposure 8 · augmentation 38 · importance 3.5/5 · click for rater detail
Direct patients in simple prescribed exercises or in the use of braces or artificial limbs.
9| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home healthcare is a low-tech, highly fragmented sector with many small providers, minimal capital investment in automation, and regulatory constraints on remote-only delivery. Adoption of autonomous AI in this domain is negligible despite sector growth. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically-intensive sector with minimal AI agent deployment for hands-on patient care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 3/5 | AI video demonstrations, exercise libraries, or reminders could assist aides in planning and educating patients, but the core task—hands-on direction and physical correction—benefits only modestly from current AI tools. Augmentation exists at the planning level, not execution. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could provide reference videos, reminder schedules, or exercise instructions to supplement the aide's training, but offers little real-time assistance during the physical task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 2/5 | While AI could theoretically demonstrate exercises via video or provide bracing instructions, this task fundamentally requires physical presence to adjust patient positioning, correct form in real-time, and respond to patient feedback—capabilities that are beyond current AI systems. Physical contact and dynamic adjustment are essential and cannot be achieved remotely at scale. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires hands-on physical guidance, spotting for safety, and real-time tactile/visual correction of a patient's movement and equipment fit, which current AI cannot perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 4/5 | Home health aides are regulated in most jurisdictions; Medicare/Medicaid reimbursement typically requires documented human-delivered care, and patient safety liability strongly favors human oversight. Patients also expect and often require human contact for reassurance and injury prevention during exercise. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Physical safety concerns, liability for injury from improper exercise or brace use, and the need for hands-on correction create strong barriers to full automation, though not a strict licensing requirement in all jurisdictions. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The human health aide wage is modest ($30–35k annually), and the task requires certified human presence by regulation and safety standards. AI video or text guidance offers minimal replacement value without physical execution, making the cost comparison heavily favor retaining the human. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical task, so cost comparison favors the human aide entirely; any AI approach would require robotics far exceeding current cost-effectiveness. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs hands-on exercise direction or prosthetic fitting without human supervision. Current AI lacks embodied presence and physical interaction capability; existing telehealth systems are assistive only, not autonomous performers of this task. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product provides in-home physical direction of exercises or fitting/adjustment of braces and prosthetics; this remains firmly in the physical, human-delivered domain. |
Care for patients by changing bed linens, washing and ironing laundry, cleaning, or assisting with their personal care.
5CI 0–10 · exposure 0 · augmentation 25 · importance 4.2/5 · click for rater detail
Care for patients by changing bed linens, washing and ironing laundry, cleaning, or assisting with their personal care.
5| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health care is a low-digitization, highly fragmented sector with aging, cost-conscious operators. No measurable AI or robotic adoption for core care tasks has occurred; the sector remains labor-intensive and human-dependent. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically embodied sector with essentially no automation penetration into hands-on caregiving tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Simple tools (task-scheduling apps, fall-alert systems) provide minor assistance, but AI offers no meaningful augmentation for the core physical and interpersonal care work described in this task statement. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can help with scheduling, reminders, or documentation around care visits, but offers little direct assistance during the physical tasks of bathing, dressing, or cleaning. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical manipulation in uncontrolled home environments (changing linens, washing, ironing, cleaning) and intimate personal care assistance that current robotics cannot reliably perform. Even specialized care robots have severe dexterity and adaptability limitations for these activities. |
| Task automatability | claude-sonnet-5 | 1/5 | This is physical, hands-on labor (lifting patients, bathing, changing linens) requiring dexterity and mobility that no current AI system or robot can perform reliably in real home settings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Home care work is heavily regulated; many jurisdictions legally require a licensed or certified human to provide direct personal care and assess patient condition changes. Client dignity, safety oversight, and liability create high regulatory and human-contact barriers to automation. |
| Adoption barriers | claude-sonnet-5 | 3/5 | While not licensed like a nurse, personal care involves intimate physical contact, safety liability, and strong patient/family preference for human caregivers, creating real but not statutory barriers. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The capital cost of robotics capable of this work (if it existed) would far exceed the loaded wage of home health aides ($15–25/hour in most U.S. markets), with substantial ongoing maintenance and operation overhead. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI/robotic substitute at any cost for this physical caregiving task, so AI cost comparison is moot and effectively far more expensive than a human aide. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed products reliably perform bed-changing, laundry care, or personal hygiene assistance at scale in real home settings. Research prototypes exist but have not achieved production-grade reliability for autonomous execution. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs in-home personal care, bathing, or housekeeping for patients; robotics for such unstructured physical tasks remains research-stage. |
Provide patients with help moving in and out of beds, baths, wheelchairs, or automobiles and with dressing and grooming.
3CI 0–5 · exposure 0 · augmentation 13 · importance 4.4/5 · click for rater detail
Provide patients with help moving in and out of beds, baths, wheelchairs, or automobiles and with dressing and grooming.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health care is a traditionally low-digitization, labor-intensive, fragmented sector with minimal AI/robotics deployment. Economic and regulatory constraints keep adoption in the laggard category. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically intensive sector with minimal AI/robotics adoption for direct care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | Limited assistive applications exist: wearable AI could remind caregivers of best practices or alert to fall risk, but AI cannot meaningfully augment the core physical assistance aspects of transfers, bathing, or dressing that dominate this task. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no meaningful assistance to the physical act of moving, dressing, or grooming patients. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task involves direct physical contact and manipulation of human bodies in varied, unpredictable positions and states. Current AI systems lack embodied robotics dexterous enough to safely assist patients with mobility and personal care without substantial human oversight and intervention. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a physical manual task requiring direct hands-on assistance with transfers and personal care; no current AI system (software or robotics) can perform this end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | This task is heavily protected by liability law, duty of care requirements, and regulatory frameworks requiring human licensure and presence. State regulations mandate that home health aides be present and accountable for patient safety during personal care. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical contact, safety liability for falls/injury, and often licensing/certification requirements for home health aides create strong barriers against automation of this task. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The hardware, software, and safety infrastructure required to deploy autonomous physical assistance robots would vastly exceed the loaded wage of a home health aide ($30k–$40k annually), making the economic case untenable today. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing this physical labor, so any hypothetical robotic solution would be far more expensive than a human aide today. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed commercial product reliably performs full-spectrum patient mobility assistance and personal grooming at scale in home settings. While research prototypes exist, production systems capable of safely and independently handling transfers and grooming remain unavailable. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical patient transfers, bathing, or dressing assistance; robotic assistance for this remains research-stage and not in real-world production. |
Bathe patients.
3CI 0–5 · exposure 0 · augmentation 13 · importance 4.3/5 · click for rater detail
Bathe patients.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health care remains a low-digitization, labor-intensive sector with minimal AI adoption. No measurable displacement or production deployment of automated bathing systems exists in the home health industry today. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically intensive sector with minimal AI/robotic adoption for direct physical care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | Current AI offers no meaningful assistance in the physical act of bathing a patient. While AI might help with scheduling or documentation, it does not augment the core bathing task itself, which remains entirely human-dependent. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI can support scheduling, care documentation, or monitoring reminders around bathing, but offers negligible assistance during the actual physical act of bathing a patient. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Bathing patients requires physical manipulation of a person's body in a private, intimate setting—something no current robot or AI system can perform safely or at equal quality. The task is fundamentally embodied and involves real-time adaptation to patient mobility, comfort, and dignity that autonomous systems cannot yet achieve. |
| Task automatability | claude-sonnet-5 | 1/5 | Bathing patients requires physical manipulation, lifting, sensitive touch, and real-time judgment of skin condition/safety in a highly variable home environment—no current AI system can perform this physical caregiving task. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Bathing patients is a licensed or regulated function in most jurisdictions; many states legally require a qualified caregiver to perform or directly supervise patient hygiene. Additionally, intimate physical contact creates liability and organizational/regulatory barriers that prevent substitution with automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical care involves safety, dignity, and liability concerns, often requiring trained/certified personnel and close human supervision, though not always a formally licensed professional. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The capital cost of a capable humanoid robot, plus ongoing maintenance and supervision, far exceeds the loaded hourly wage of a home health aide. No cost-competitive commercial solution exists for this task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI or robotic solution for this physical task, so any hypothetical system would require expensive specialized hardware far exceeding the cost of a human aide. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs full patient bathing in production settings today. While robotic prototypes exist in research contexts, they lack the dexterity, safety assurance, and human-touch judgment required for dependable clinical use at scale. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs patient bathing; robotic bathing assistance exists only in narrow research/prototype contexts, not in production home health settings. |
Accompany clients to doctors' offices or on other trips outside the home, providing transportation, assistance, and companionship.
3CI 0–5 · exposure 0 · augmentation 25 · importance 3.8/5 · click for rater detail
Accompany clients to doctors' offices or on other trips outside the home, providing transportation, assistance, and companionship.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health care is a low-digitization, human-intensive sector dominated by small providers and individual workers. Adoption of even basic digital tools remains slow, and the sector has not begun deploying autonomous transportation or robot companions at meaningful scale. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically-embodied sector with minimal AI agent deployment for hands-on client accompaniment tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could assist with appointment scheduling, transportation routing, or medical record management, but these are peripheral to the core task of physical accompaniment and assistance. The central requirement—being present with the client—offers minimal surface for AI augmentation. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could help schedule appointments, arrange transportation logistics, or provide route optimization, but offers little assistance for the actual accompaniment and companionship task itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical presence, real-time adaptability to client needs, hands-on assistance, and genuine human companionship—capabilities that current AI systems cannot provide. Transportation, mobility support, and emotional presence cannot be meaningfully automated by deployed AI today. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires physical presence, driving or transit assistance, and real-world mobility support that current AI systems cannot perform at all, let alone with time savings. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Licensing and liability are substantial barriers: home health aides operate under regulatory oversight, duty-of-care standards, and legal accountability for client safety. Many jurisdictions require a licensed human to provide direct in-home and out-of-home personal assistance, and liability for harm to vulnerable populations creates hard legal and organizational friction. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Physical presence, driving responsibility, liability for client safety, and the inherently personal/companionship nature of the task create strong barriers to any automated substitution. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The loaded cost of a human home health aide is relatively low (~$15–25/hour in many markets), and any autonomous system capable of safe transportation and physical assistance would require significant capital and operating costs that exceed this wage rate. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute for physical accompaniment and transportation, so any AI cost comparison is moot—human labor is the only option and thus effectively cheaper by default. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can autonomously provide in-person transportation, physical assistance, and companionship to vulnerable clients. While ride-sharing and companion robots exist in limited capacity, they do not reliably handle the full scope of medical appointment accompaniment with the safety and care standards required. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product accompanies clients physically to appointments; this remains entirely outside the scope of software or robotic products available today. |
Change dressings.
3CI 0–5 · exposure 0 · augmentation 25 · importance 3.6/5 · click for rater detail
Change dressings.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home healthcare is a low-digitization, labor-intensive sector with limited capital investment in automation; adoption of robotic wound care remains negligible in practice. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a physically-intensive, low-digitization sector where AI adoption for hands-on caregiving tasks remains minimal to nonexistent. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI could provide decision support (e.g., wound classification from images, dressing-type suggestions) but offers minimal assistance to the core physical and tactile elements of the dressing change itself. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI could offer minor assistance such as documentation prompts or wound photo analysis for infection signs, but it does not meaningfully enhance the physical act of dressing changes. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Changing dressings requires physical manipulation of bandages and wound sites, sterile technique, and real-time assessment of wound condition—tasks that demand embodied presence and dexterity far beyond current AI robotic capabilities in uncontrolled home environments. |
| Task automatability | claude-sonnet-5 | 1/5 | Changing dressings requires physical manipulation, sterile technique, and tactile assessment of wounds, which current AI systems cannot perform end-to-end as they lack embodied physical capability. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Wound dressing changes are often tied to skilled nursing or clinical protocols; many jurisdictions require documented training or supervision, and liability concerns around infection risk and improper technique create hard barriers to unsupervised automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical contact, hygiene/infection control standards, and liability for improper wound care create strong practical barriers, though not always a strict licensing requirement for home health aides specifically. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The capital, integration, and safety-oversight costs of a robotic system capable of sterile wound dressing far exceed the loaded wage of a home health aide for this task. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute for the physical task, so any comparison would require a robotic system that does not exist commercially, making AI effectively far more costly or simply unavailable. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs wound dressing changes in home settings today; this remains a task requiring licensed or trained human healthcare workers with tactile feedback and situational judgment. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical wound dressing changes; this remains firmly in the domain of human caregivers with no robotic or AI product doing this in home settings. |
Massage patients or apply preparations or treatments, such as liniment, alcohol rubs, or heat-lamp stimulation.
3CI 0–5 · exposure 0 · augmentation 0 · importance 3.3/5 · click for rater detail
Massage patients or apply preparations or treatments, such as liniment, alcohol rubs, or heat-lamp stimulation.
3| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health care remains a low-digitization, labor-intensive sector with limited adoption of advanced robotics. Physical care tasks in this sector show minimal real-world AI or robotic displacement. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, physically intensive sector with minimal AI/robotic adoption for direct physical care tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 1/5 | AI offers no meaningful assistance for the core task of applying massage or physical treatments; digital tools cannot augment the hands-on tactile work or enhance patient outcomes in this context. |
| Augmentation potential | claude-sonnet-5 | 1/5 | AI offers essentially no meaningful assistance to the physical act of massaging or applying treatments, though it might tangentially help with scheduling or care documentation unrelated to this specific task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | Massage and physical treatments require direct tactile contact, force calibration, and real-time patient feedback that current robotic systems cannot reliably deliver. The task involves sensing tissue response and adjusting pressure/technique dynamically—capabilities far beyond today's general-purpose AI or robotic systems. |
| Task automatability | claude-sonnet-5 | 1/5 | This is a hands-on physical care task requiring direct touch, mobility, and adaptive physical manipulation of a patient's body—current AI systems have no ability to perform physical massage or apply topical treatments. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Licensed healthcare providers and home health aides are often legally required to perform or directly supervise hands-on patient treatments. Liability, state regulations, and patient safety requirements create hard barriers to substitution or full automation. |
| Adoption barriers | claude-sonnet-5 | 4/5 | Direct physical contact with vulnerable patients, safety concerns (burns from heat lamps, skin reactions), and the human-contact nature of caregiving create strong practical and liability barriers to automation, though not a strict licensure requirement. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | Current robotic massage or treatment-application systems are extremely expensive to purchase and maintain, require skilled technicians, and provide limited functional scope compared to a home health aide's hourly wage. The total cost of ownership far exceeds human labor. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI-driven substitute for physical application of treatments, so any hypothetical robotic solution would require expensive specialized hardware far exceeding the cost of a human aide. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product reliably performs therapeutic massage or physical treatment application in home or clinical settings. While specialized robotic arms exist in research, they lack the dexterity, safety certification, and real-world validation needed for unsupervised patient care. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product performs physical massage or applies liniment/heat treatments to patients; robotics for this specific caregiving task remain research-stage at best. |
Administer prescribed oral medications, under the written direction of physician or as directed by home care nurse or aide, and ensure patients take their medicine.
0CI 0–0 · exposure 0 · augmentation 38 · importance 4.3/5 · click for rater detail
Administer prescribed oral medications, under the written direction of physician or as directed by home care nurse or aide, and ensure patients take their medicine.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health is a low-digitization, human-contact-intensive sector with strong regulatory constraints and an aging population expecting human caregivers. Adoption of AI or robotics for core care tasks remains minimal in real deployment. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health care is a low-digitization, high-touch, in-person sector with minimal AI agent deployment for hands-on physical caregiving tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI might assist with reminders or record-keeping for medication schedules, but it cannot augment the core task of physical medication administration and direct patient observation, which remain entirely human-dependent. |
| Augmentation potential | claude-sonnet-5 | 3/5 | AI-enabled reminder systems, medication trackers, and alert tools can help aides and families track schedules and flag missed doses, offering moderate assistance without replacing the physical task. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task requires physical presence in the home, direct observation of the patient taking medication, and real-time judgment about patient compliance and adverse reactions—capabilities no current AI system possesses. The task is fundamentally embodied and relational, involving human contact and decision-making that cannot be automated end-to-end today. |
| Task automatability | claude-sonnet-5 | 1/5 | This requires physical presence to hand medication to a patient, observe ingestion, and respond to reactions—no AI system can physically administer or verify medication intake today. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Medication administration by unlicensed individuals is tightly regulated; state boards of nursing and federal laws (e.g., CMS regulations) typically require a licensed nurse to oversee or directly administer medications. A human aide can do this only under written physician direction and nurse supervision, creating a hard legal and licensing barrier. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Medication administration is a regulated caregiving act requiring physician direction and often licensed/certified personnel, with high liability for errors—strong legal and safety barriers block automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | The task requires a physical human presence in the patient's home; even theoretical AI/robotic equivalents would require expensive hardware, maintenance, and integration vastly more costly than a home health aide's loaded wage. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no viable AI substitute performing the physical act, so any comparison defaults to the human being the only option, making AI effectively unusable/not cheaper. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed product can physically deliver medications, observe patient intake, or make clinical judgments about medication administration in a home setting. This remains entirely outside the scope of deployed AI or robotic systems in practical use. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed product performs physical medication administration or in-person compliance verification; smart pill dispensers and reminder apps exist but do not replace the hands-on task. |
Care for children with disabilities or who have sick parents or parents with disabilities.
0CI 0–0 · exposure 0 · augmentation 25 · importance 4.0/5 · click for rater detail
Care for children with disabilities or who have sick parents or parents with disabilities.
0| Dimension | Model | Rating | Rationale |
|---|---|---|---|
| Sector adoption velocity | claude-haiku-4-5-20251001 | 1/5 | Home health aide work is performed in highly fragmented, low-digitization settings (private homes) with predominantly small employers and limited capital for automation. Current adoption of AI in home care is negligible. |
| Sector adoption velocity | claude-sonnet-5 | 1/5 | Home health and personal care sectors are among the least digitized, with minimal AI agent deployment for direct physical caregiving tasks. |
| Augmentation potential | claude-haiku-4-5-20251001 | 2/5 | AI offers minimal meaningful assistance in direct caregiving tasks. Simple scheduling or reminder systems might marginally reduce administrative burden, but AI cannot assist with the core physical and emotional work of caring for children with disabilities. |
| Augmentation potential | claude-sonnet-5 | 2/5 | AI tools (scheduling, care-plan reminders, communication aids) can support the aide administratively, but offer little to no assistance for the core caregiving activity itself. |
| Task automatability | claude-haiku-4-5-20251001 | 1/5 | This task involves hands-on physical care, emotional support, and safety monitoring of vulnerable children that cannot be performed by current AI systems. The task requires direct human contact, judgment in real-time crisis situations, and the ability to respond to complex, unpredictable child behaviors. |
| Task automatability | claude-sonnet-5 | 1/5 | This task requires direct physical caregiving, emotional support, and hands-on supervision of children in a home setting, none of which current AI systems can perform end-to-end. |
| Adoption barriers | claude-haiku-4-5-20251001 | 5/5 | Legal and regulatory frameworks require licensed or certified human caregivers for children with disabilities and health needs. Liability, duty of care, and mandatory reporting obligations create hard barriers that prevent AI substitution and require a qualified human to remain legally responsible. |
| Adoption barriers | claude-sonnet-5 | 5/5 | Caregiving for vulnerable children involves safety, liability, child welfare regulations, and requires a physically present, often background-checked and trained human, creating hard barriers to automation. |
| Cost vs. human wage | claude-haiku-4-5-20251001 | 1/5 | AI systems cannot currently perform this task at all, making direct cost comparison meaningless. The loaded wage for home health aides remains far cheaper than any hypothetical solution combining robotics, AI, and necessary oversight. |
| Cost vs. human wage | claude-sonnet-5 | 1/5 | There is no AI substitute delivering this physical, relational service, so cost comparison favors the human by default—AI cannot produce the output at all. |
| Technical feasibility today | claude-haiku-4-5-20251001 | 1/5 | No deployed AI product can reliably perform childcare or disability support tasks. Current AI cannot physically assist children, make nuanced decisions about their welfare, or substitute for the human presence and trust required in these sensitive caregiving relationships. |
| Technical feasibility today | claude-sonnet-5 | 1/5 | No deployed AI product provides physical child care or in-home caregiving; this remains firmly outside current product capabilities. |
Related occupations — Healthcare Support
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.