Cashiers

41-2011.00

Median wage $32,880/yr3,089,410 employed (US)Rank #50 of 923 scored · top 5% by substitution

Receive and disburse money in establishments other than financial institutions. May use electronic scanners, cash registers, or related equipment. May process credit or debit card transactions and validate checks.

Sub-scores

0–100 · band = confidence interval from rater disagreement

Substitution57

Exposure54

Augmentation43

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

28 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

43%

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%54

panel mean rating 3.1/5 → substitution pressure 54/100

Technical feasibility todayw 20%55

panel mean rating 3.2/5 → substitution pressure 55/100

Cost vs. human wagew 15%57

panel mean rating 3.3/5 → substitution pressure 57/100

Adoption barriersw 20%inverted — strong barriers lower the score69

panel mean rating 2.2/5 (barrier strength) → substitution pressure 69/100

Sector adoption velocityw 10%53

panel mean rating 3.1/5 → substitution pressure 53/100

Task breakdown (28 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.

Compute and record totals of transactions.

CI 95–100 · exposure 100 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail and payment processing are highly digitized sectors with decades-long deployment of automated transaction systems; self-checkout, online retail, and POS automation are near-universal in developed economies, reflecting deep and rapid adoption.
Sector adoption velocity	claude-sonnet-5	5/5	Retail has broadly and rapidly adopted automated checkout and POS totaling systems for decades, with self-checkout and scan-based totals now standard in most stores.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI assists human cashiers by automating the computational burden, reducing errors, and allowing them to focus on customer service, but the human remains present in hybrid checkout scenarios rather than being fully replaced.
Augmentation potential	claude-sonnet-5	3/5	Where cashiers remain, POS systems assist by automatically calculating totals, taxes, and change, reducing manual arithmetic and error but still requiring human operation for many transactions.
Task automatability	claude-haiku-4-5-20251001	5/5	Computing and recording transaction totals is a straightforward mathematical operation that modern point-of-sale (POS) systems and AI already perform end-to-end with >50% time savings; this task requires minimal human judgment and is fully automatable by current technology.
Task automatability	claude-sonnet-5	5/5	POS systems, scanners, and payment terminals already compute and record transaction totals automatically; this task is largely already automated by non-AI and AI-assisted retail technology.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While retailers may retain human cashiers for customer service and theft prevention, there are minimal legal or regulatory barriers preventing automation of transaction computation itself; organizational inertia and customer preference for human interaction are modest friction points.
Adoption barriers	claude-sonnet-5	1/5	No licensing, liability, or regulatory requirement mandates a human compute transaction totals; self-checkout and automated systems are already widespread and legally unrestricted.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	The cost of running automated POS and payment computation (infrastructure, inference, maintenance) is orders of magnitude cheaper than the loaded wage of a human cashier performing the same task.
Cost vs. human wage	claude-sonnet-5	5/5	Automated scanning/POS hardware and software cost pennies per transaction compared to a cashier's wage for the same computation task.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	POS systems, self-checkout kiosks, and AI-integrated payment processors reliably compute transaction totals and record them in production systems across millions of retail locations worldwide today.
Technical feasibility today	claude-sonnet-5	5/5	Barcode scanners, POS software, and self-checkout kiosks reliably perform this function at massive scale in production across virtually all retail settings today.

Keep periodic balance sheets of amounts and numbers of transactions.

CI 95–100 · exposure 100 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail and food-service sectors have near-universal adoption of automated POS and accounting systems that generate periodic balance sheets; this is standard practice in production environments, not experimental.
Sector adoption velocity	claude-sonnet-5	5/5	Retail and point-of-sale systems have near-universal digital adoption, with automated end-of-shift/day balancing standard practice for decades.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI-assisted tools (real-time dashboards, anomaly detection, auto-reconciliation alerts) significantly boost productivity and accuracy for any human reviewing or managing balance sheet processes, keeping them informed and in oversight.
Augmentation potential	claude-sonnet-5	3/5	Where manual reconciliation still occurs, basic spreadsheet or calculator tools aid the cashier, though the task is largely already automated rather than merely augmented.
Task automatability	claude-haiku-4-5-20251001	5/5	Periodic balance sheet generation from transaction data is a fully automatable task; modern point-of-sale and accounting systems routinely perform this end-to-end with near-zero human intervention, achieving well over 50% time savings at equal or superior accuracy.
Task automatability	claude-sonnet-5	5/5	Reconciling transaction totals and generating balance summaries is a structured, rule-based data aggregation task that POS and accounting systems already perform automatically.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While no licensing requirement mandates human oversight of balance sheets, most retail organizations retain some manual review or audit steps for compliance and fraud detection, creating modest friction to full automation.
Adoption barriers	claude-sonnet-5	1/5	No licensing, legal, or human-contact requirement exists for internal bookkeeping/reconciliation tasks; software has replaced manual tallying in most retail settings already.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	The per-transaction cost of automated balance sheet generation (via integrated POS/accounting systems) is orders of magnitude cheaper than paying a cashier or accountant to manually tally and reconcile transactions.
Cost vs. human wage	claude-sonnet-5	5/5	Automated register/software reconciliation costs negligible marginal compute or licensing fees compared to paying a human to manually tally and record transaction totals.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature, deployed products (accounting software, POS systems, ERP platforms) reliably perform automated balance sheet reconciliation and reporting in production at scale across retail and food-service sectors daily.
Technical feasibility today	claude-sonnet-5	5/5	POS systems, cash registers, and retail accounting software already produce automated periodic balance/reconciliation reports as a standard, mature feature in production use.

Establish or identify prices of goods, services, or admission, and tabulate bills, using calculators, cash registers, or optical price scanners.

CI 92–100 · exposure 100 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail and hospitality have rapidly adopted self-checkout and automated POS systems over the past decade; major chains now operate largely automated checkout infrastructure, though some friction remains in small businesses and certain sectors.
Sector adoption velocity	claude-sonnet-5	5/5	Retail and grocery have rapidly and deeply adopted self-checkout and automated scanning systems over the past two decades, with continued expansion.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Once the bill is calculated and presented, there is little meaningful assistance AI can offer a human cashier; the task is already so automated that augmentation is nearly moot. Any remaining human role is oversight, not task execution.
Augmentation potential	claude-sonnet-5	4/5	Where human cashiers remain, scanners and registers already substantially speed up price identification and billing, though the task itself is largely automated rather than merely augmented.
Task automatability	claude-haiku-4-5-20251001	5/5	Identifying prices and calculating bills is almost entirely automatable: barcode scanners and point-of-sale systems already achieve this end-to-end with near-zero error rates and well over 50% time savings compared to manual entry. The core task is algorithmic and deterministic.
Task automatability	claude-sonnet-5	5/5	Price lookup and bill tabulation via barcode scanners and POS systems is already fully automated in most retail contexts, often requiring no human intervention beyond scanning or self-checkout supervision.
Adoption barriers	claude-haiku-4-5-20251001	2/5	While some businesses still prefer human checkout for customer service and loss-prevention reasons, there are no legal or regulatory barriers preventing full automation of price identification and bill tabulation. Organizational inertia and customer preference are weak barriers compared to truly restricted professions.
Adoption barriers	claude-sonnet-5	1/5	No licensing, legal, or liability barriers prevent automated price lookup and billing; this is already standard commercial practice with no regulatory restriction.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Self-checkout and automated billing systems cost a fraction of a human cashier's loaded wage per transaction; the infrastructure is amortized across thousands of transactions and operates continuously at near-zero marginal cost per task instance.
Cost vs. human wage	claude-sonnet-5	5/5	Automated scanning and POS hardware/software cost pennies per transaction versus a cashier's wage, and self-checkout lanes let one attendant oversee many stations.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Deployed products—self-checkout systems, modern POS terminals, and barcode/optical scanners—reliably perform price lookup and bill calculation in millions of retail and hospitality locations daily. This is mature, production-scale automation.
Technical feasibility today	claude-sonnet-5	5/5	Barcode/RFID scanning, self-checkout kiosks, and integrated POS systems are mature, widely deployed technologies used at massive scale across retail and grocery today.

Calculate total payments received during a time period, and reconcile this with total sales.

CI 84–95 · exposure 92 · augmentation 75 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail and food service sectors have rapidly adopted integrated POS systems that automate reconciliation as standard. This adoption is widespread and deep, with small- and mid-market retailers now routinely using systems that eliminate manual end-of-shift reconciliation.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and point-of-sale reconciliation is a mature, widely digitized process with near-universal adoption of automated POS/accounting reconciliation tools.
Augmentation potential	claude-haiku-4-5-20251001	4/5	Reconciliation software augments human decision-making by instantly flagging discrepancies, errors, and anomalies that a human would then review and investigate. This transforms oversight productivity even when a human retains final sign-off authority.
Augmentation potential	claude-sonnet-5	4/5	Even where full automation isn't trusted, AI-enabled reconciliation software significantly speeds up identifying discrepancies and summarizing totals for human review.
Task automatability	claude-haiku-4-5-20251001	5/5	Payment reconciliation is a core computational task that can be fully automated by point-of-sale (POS) systems, accounting software, and inventory management platforms. Current systems reliably calculate totals, cross-reference sales records, and flag discrepancies with minimal human intervention, easily meeting the 50% time-saving threshold.
Task automatability	claude-sonnet-5	4/5	Reconciling payments to sales totals is a structured, rules-based numeric task that POS and accounting systems already automate almost entirely, requiring only exception review by a human.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few legal or regulatory barriers prevent full automation; reconciliation is a mechanical process. Minor friction exists from legacy systems requiring integration and some retailers' preference for human audit as a control, but these are operational rather than legal constraints.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement exists for this task, though some internal controls/audit policies require human sign-off or dual verification for cash handling discrepancies.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated reconciliation via integrated POS systems costs a fraction of a cent per transaction after initial setup, vastly cheaper than paying a cashier or accounting clerk the full loaded wage (~$30k–$45k annually) to manually reconcile payments daily.
Cost vs. human wage	claude-sonnet-5	5/5	Once integrated into a POS system, the marginal cost of automated reconciliation is near zero compared to paying a cashier's time to manually tally and reconcile.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature POS and reconciliation software has been deployed at scale in retail for decades, reliably automating payment receipt calculation and sales reconciliation. Systems like Square, Toast, Shopify, and enterprise accounting platforms perform this task reliably in production daily across millions of transactions.
Technical feasibility today	claude-sonnet-5	5/5	POS systems, ERP software, and automated reconciliation tools (e.g., Square, Shopify, SAP) perform end-of-day cash/sales reconciliation reliably in production across retail today.

Weigh items sold by weight to determine prices.

CI 84–95 · exposure 92 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail has already adopted weight-based automated pricing systems nearly universally in developed economies; this is not a future possibility but an established, decades-old practice in grocery and food retail.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and grocery have widely adopted self-checkout and automated scale systems over the past decade, though full replacement of staffed lanes remains partial.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Automated scales assist cashiers by providing instant, accurate weight and price data, reducing manual calculation errors and speeding transaction processing, though the task is largely performed by the system itself rather than augmenting human judgment.
Augmentation potential	claude-sonnet-5	3/5	Where cashiers remain, integrated scanning-scale systems speed up the weighing and pricing step, though the task itself is largely automated rather than augmented.
Task automatability	claude-haiku-4-5-20251001	4/5	Most of the task is automatable: a scale integrated with POS systems can weigh items and look up prices from a database instantly. However, the remaining portion—handling ambiguous items, verifying weight legitimacy, and resolving scale errors—still benefits from human oversight, preventing a full 5 rating.
Task automatability	claude-sonnet-5	5/5	Self-checkout and POS-integrated scales already automate weighing and price calculation fully, requiring no human intervention beyond placing the item on the scale.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal requirement mandates human verification of weight-based pricing; POS systems are routine retail infrastructure. The only minor friction is staff training and customer comfort with automation, which is already normalized.
Adoption barriers	claude-sonnet-5	2/5	Some friction exists around loss prevention, produce misidentification, and customer preference for staffed lanes, but no licensing or legal requirement mandates human weighing.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	The cost of an automated scale and price lookup integration is a one-time capital expense amortized over years, making it orders of magnitude cheaper than paying a cashier for this specific task across thousands of transactions.
Cost vs. human wage	claude-sonnet-5	5/5	Automated scale/POS integration has negligible marginal cost per transaction compared to a human cashier's wage for the same weighing action.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Weighing scales with integrated POS systems have been standard in grocery retail for decades; modern systems automatically look up and apply prices from item databases in production at scale across thousands of stores.
Technical feasibility today	claude-sonnet-5	5/5	Self-service checkout systems with integrated scales are deployed at massive scale in grocery and retail stores worldwide today.

Sort, count, and wrap currency and coins.

CI 80–92 · exposure 92 · augmentation 38 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail chains, supermarkets, and banks have steadily adopted coin counters and note sorters over the past two decades. Modern e-commerce and self-checkout environments accelerate this trend, showing sustained, widespread deployment in the retail/financial sector.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and banking sectors have adopted coin-counting and cash-recycling machines broadly and for decades, representing mature, widespread adoption relative to many other automation categories.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Once automated counting is in place, there is minimal role for AI to assist a human performing this task; the human is largely removed from the loop. The task offers little opportunity for augmentation because the work is fully substitutable.
Augmentation potential	claude-sonnet-5	3/5	Where full automation isn't installed, counting machines still assist cashiers by speeding up and verifying counts, though the human typically remains responsible for final reconciliation.
Task automatability	claude-haiku-4-5-20251001	5/5	Sorting, counting, and wrapping currency and coins are highly repetitive, rule-based mechanical tasks. Coin counters and currency sorters with computer vision already perform this end-to-end faster than manual methods with equal accuracy, meeting the ≥50% time-saving bar.
Task automatability	claude-sonnet-5	4/5	Sorting, counting, and wrapping currency is a highly mechanical, rule-based task already well-suited to automated coin/currency counting machines and cash-recycling systems., meeting the 50% time-savings bar easily.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No licensing or regulatory barrier exists for automated counting equipment. The main friction is organizational (existing infrastructure, familiarity) rather than legal or liability-driven; adoption remains straightforward.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement for this specific sub-task; some organizational friction exists around cash-handling controls and dual-verification policies, but no legal mandate for human performance.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated coin and currency counting equipment costs pennies per transaction (amortized infrastructure + electricity), while a human cashier takes 15–30 minutes per cash drawer reconciliation at loaded wage cost, making AI solutions orders of magnitude cheaper.
Cost vs. human wage	claude-sonnet-5	4/5	Automated currency counting/sorting machines have low per-transaction operating cost versus paying a cashier's time for manual counting, though upfront hardware cost is a factor.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Coin counting and currency sorting machines are mature, deployed products used routinely in retail, banking, and casinos. These systems reliably perform the full task at scale in production environments with minimal error rates.
Technical feasibility today	claude-sonnet-5	5/5	Commercial coin counters, currency sorters, and cash-recycler machines are widely deployed in banks, retailers, and casinos today, reliably performing this exact task at scale.

Receive payment by cash, check, credit cards, vouchers, or automatic debits.

CI 80–86 · exposure 84 · augmentation 50 · importance 4.8/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Retail and hospitality sectors have rapidly deployed self-checkout, mobile payment, and automated point-of-sale systems over the past decade. Major chains (Walmart, Target, Amazon) operate significant cashier-less or automated payment environments at scale.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and grocery have adopted self-checkout and automated payment systems broadly and rapidly over the past decade, representing one of the more advanced physical-sector automation cases.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist cashiers by flagging suspicious transactions, predicting payment method likelihood, and streamlining receipts, but the core payment receipt task is already heavily automated rather than augmented in typical deployments.
Augmentation potential	claude-sonnet-5	3/5	Where human cashiers remain, POS systems and automated scanning meaningfully speed transaction processing and reduce errors, though the task itself is largely mechanical rather than judgment-based.
Task automatability	claude-haiku-4-5-20251001	4/5	Current payment processing systems (POS terminals, mobile payment processors) can handle cash, checks, cards, vouchers, and automatic debits with minimal human intervention, achieving substantial time savings. However, full end-to-end automation without human oversight for fraud detection, verification, and exception handling prevents a perfect 5 rating.
Task automatability	claude-sonnet-5	4/5	Self-checkout kiosks, payment terminals, and automated point-of-sale systems already handle the mechanics of processing cash, card, and digital payments with minimal human involvement, though cash handling and exception cases still need staffing.
Adoption barriers	claude-haiku-4-5-20251001	2/5	PCI-DSS and payment card industry regulations impose compliance requirements, but they do not mandate human involvement—only security standards that automation readily meets. Some jurisdictions and retailers retain human cashiers for customer service preference rather than legal requirement.
Adoption barriers	claude-sonnet-5	2/5	No licensing is required to process payments, but some friction exists from theft/fraud liability, cash handling regulations, and customer preference for staffed lanes in certain contexts.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	The per-transaction cost of automated payment processing (infrastructure amortized across high volumes) is orders of magnitude cheaper than human cashier labor, especially when factoring in wage, benefits, and time per transaction.
Cost vs. human wage	claude-sonnet-5	4/5	Self-checkout machines and card terminals amortize quickly and require far less labor per transaction than a dedicated cashier, though upfront hardware and loss-prevention monitoring add cost.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature, deployed payment processing technologies reliably handle all payment types mentioned at scale across millions of retail locations worldwide. Credit card readers, check scanners, and automated payment gateways are production-proven systems.
Technical feasibility today	claude-sonnet-5	5/5	Self-checkout and automated payment terminals are mature, widely deployed products in retail, grocery, and fast food at massive scale today.

Issue receipts, refunds, credits, or change due to customers.

CI 75–86 · exposure 80 · augmentation 63 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Self-checkout and automated POS systems are widespread in retail, with major chains rapidly expanding self-service infrastructure. Adoption has been fast and deep, particularly in grocery and quick-service restaurant sectors.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and grocery have adopted self-checkout and automated payment/refund systems widely and rapidly over the past decade, representing deep penetration relative to other physical-service sectors.
Augmentation potential	claude-haiku-4-5-20251001	4/5	POS systems augment cashiers significantly by handling calculations, record-keeping, and tender processing instantly, freeing staff for customer service and problem-solving. However, the task is largely procedural, so augmentation is bounded.
Augmentation potential	claude-sonnet-5	3/5	AI-enabled POS systems assist cashiers with faster transaction processing, fraud flags, and automated calculations, improving productivity even when a human remains involved.
Task automatability	claude-haiku-4-5-20251001	4/5	Self-checkout systems and point-of-sale (POS) automation already handle receipts, refunds, and change calculation end-to-end with minimal human intervention, achieving well over 50% time savings. However, complex refund scenarios (damaged goods, policy exceptions) still require human judgment, preventing a full 5 rating.
Task automatability	claude-sonnet-5	4/5	The transactional logic of issuing receipts, refunds, credits, or change is highly rule-based and already largely automated via POS systems and self-checkout kiosks, though physical cash handling still requires hardware or human involvement.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few legal barriers prevent automation; retailers can deploy self-checkout freely. Main friction points are customer preference for human interaction and fraud/shrinkage concerns with unmanned systems, but these are soft, not regulatory barriers.
Adoption barriers	claude-sonnet-5	2/5	Some friction exists from theft/loss-prevention concerns, customer preference for human interaction, and cash-handling regulations, but no licensing or legal requirement mandates a human perform this task.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated POS and self-checkout systems operate at a tiny fraction of cashier labor cost per transaction, easily an order of magnitude cheaper when amortized across high transaction volumes.
Cost vs. human wage	claude-sonnet-5	4/5	Once kiosk/automation hardware is installed, per-transaction cost is far below a cashier's wage, though upfront capital costs and maintenance reduce the ratio slightly versus pure software tasks.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature POS systems and self-checkout terminals perform receipt issuance, change calculation, and basic refund processing reliably in production at scale across retail worldwide. These systems are proven and widely deployed.
Technical feasibility today	claude-sonnet-5	4/5	Self-checkout kiosks, automated cash-recycling machines, and e-commerce refund systems reliably perform this task at scale in production today across retail and grocery.

Post charges against guests' or patients' accounts.

CI 75–86 · exposure 80 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Hospitality, healthcare, and retail sectors have been systematically automating charge posting for decades; the shift from manual posting to digital and now AI-integrated billing is deeply embedded in industry operations.
Sector adoption velocity	claude-sonnet-5	4/5	Hospitality and healthcare administrative billing have already broadly adopted automated charge-posting systems as standard practice, though full end-to-end automation varies by setting.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI assists by flagging discrepancies, suggesting charge categories, and accelerating the posting workflow, but human oversight remains valuable for disputes and exceptions, making this moderately augmentative rather than fully transformative.
Augmentation potential	claude-sonnet-5	3/5	Even where automated, cashiers still often verify, correct, or manually enter charges not captured automatically, so AI/software assistance meaningfully speeds up but doesn't fully replace human involvement in all cases.
Task automatability	claude-haiku-4-5-20251001	4/5	Posting charges to accounts is largely data-entry and record-matching work that AI systems can automate end-to-end with minimal human intervention. Current POS and billing systems already perform much of this automatically; AI could complete the full workflow (charge capture, account lookup, posting, verification) with ≥50% time savings at equal quality in most hospitality and healthcare settings.
Task automatability	claude-sonnet-5	4/5	Posting charges to guest or patient accounts is a structured, rules-based data entry task that integrated POS/PMS/billing systems can already perform automatically once transactions occur, meeting the time-saving threshold for most instances.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Light regulatory oversight of the posting mechanism itself, though audit trails and payment governance exist. Healthcare and payment-processing environments require compliance, but automation of charge posting is already standard practice and legally permissible with appropriate controls.
Adoption barriers	claude-sonnet-5	2/5	Some oversight is expected for billing accuracy and dispute handling, and healthcare billing has compliance requirements, but there is no licensing requirement specifically for posting charges.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Once integrated into billing infrastructure, the marginal cost per charge posted is negligible (near-zero inference and processing), far cheaper than paying a cashier or billing clerk to manually post each transaction.
Cost vs. human wage	claude-sonnet-5	4/5	Automated charge-posting via integrated software is far cheaper per transaction than manual entry by a cashier, though system integration and maintenance costs are nontrivial.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Mature, production-deployed systems (integrated billing software, POS platforms, healthcare billing systems) reliably perform charge posting at scale across thousands of organizations. This task is already routinely automated in real-world deployments.
Technical feasibility today	claude-sonnet-5	4/5	Hotel PMS, hospital billing systems, and POS software already auto-post charges in production at scale, though exceptions, adjustments, and disputed charges still often require human review.

Compile and maintain non-monetary reports and records.

CI 67–92 · exposure 78 · augmentation 63 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail and food-service sectors are actively deploying inventory and transaction automation; POS system integration with backend reporting is now mainstream, indicating fast, deep adoption in digitized retail environments.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is a moderately digitized sector with widespread POS adoption, but many small and mid-size retailers still rely on manual or semi-manual record-keeping practices, so adoption is uneven.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can meaningfully assist cashiers by auto-populating records, flagging discrepancies, and generating summary reports, which raises clerk productivity even when humans remain responsible for verification and sign-off.
Augmentation potential	claude-sonnet-5	4/5	AI-enabled inventory and reporting tools significantly reduce time cashiers spend compiling records, letting them focus on customer-facing duties while software handles data aggregation.
Task automatability	claude-haiku-4-5-20251001	5/5	Compiling and maintaining non-monetary reports and records is largely data entry and organization work that current AI systems can fully automate. RPA and document-processing AI can extract, categorize, and structure transaction and inventory records with minimal human intervention, easily meeting the 50% time-saving threshold.
Task automatability	claude-sonnet-5	4/5	Compiling and maintaining routine non-monetary records (inventory logs, transaction summaries, shift reports) is a structured data task that off-the-shelf software and AI-assisted systems can largely automate with modest setup.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few regulatory or licensing barriers exist for automating non-monetary record maintenance; the main friction is organizational (legacy systems, staff training, change management) rather than legal or compliance-based.
Adoption barriers	claude-sonnet-5	1/5	No licensing, liability, or human-contact requirements attach to compiling internal non-monetary records; retailers can freely automate this without regulatory or professional constraints.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Inference and automation of record compilation costs pennies per transaction and requires minimal ongoing oversight compared to the fully-loaded wage of a cashier spending time on this clerical task, yielding a substantial cost advantage.
Cost vs. human wage	claude-sonnet-5	4/5	Automated reporting via existing POS/inventory software is very cheap per unit of output compared to paying a cashier's time to compile records manually.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed products (RPA platforms, OCR systems, database automation tools) reliably perform record compilation and maintenance in retail and hospitality settings today. Minor gaps exist in handling ambiguous or handwritten inputs, but production systems handle the majority of non-monetary reporting at scale.
Technical feasibility today	claude-sonnet-5	3/5	POS systems and retail software already auto-generate many reports, but full replacement of ad hoc record compilation still requires configuration and occasional manual correction, so reliability varies by store/system.

Sell tickets and other items to customers.

CI 66–75 · exposure 67 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Ticketing venues (cinemas, transit, attractions) and retail have deployed kiosks and mobile/app-based purchasing widely; major venues report rapid substitution of cashier roles. Adoption is uneven by sector size but measurably deep in high-transaction environments.
Sector adoption velocity	claude-sonnet-5	4/5	Retail, transit, and entertainment sectors have rapidly adopted self-checkout and automated ticketing over the past decade, representing fast, visible displacement.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI assists minimally in the core task of selling tickets—mostly flagging inventory or fraud alerts to a human operator. The task is more naturally replaced than augmented, so augmentation value is limited.
Augmentation potential	claude-sonnet-5	3/5	AI-powered POS systems, recommendation prompts, and inventory alerts assist cashiers in upselling and processing transactions faster, though the task itself is simple enough that augmentation value is moderate.
Task automatability	claude-haiku-4-5-20251001	3/5	Current AI/automated systems can handle transaction processing, payment acceptance, and basic inventory tracking with significant setup, but the full customer interaction—handling special requests, managing refunds, addressing complaints—requires human judgment roughly half the time. Kiosks exist but typically require human fallback for non-standard transactions.
Task automatability	claude-sonnet-5	4/5	Ticket and item sales are largely transactional and already handled by self-checkout kiosks, vending machines, and online ticketing systems that meet or exceed the 50% time-saving bar.
Adoption barriers	claude-haiku-4-5-20251001	2/5	No legal requirement mandates a licensed human cashier; the main barriers are customer preference for human contact and organizational inertia in some venues, but these are weak and eroding as self-service normalizes.
Adoption barriers	claude-sonnet-5	2/5	Some friction exists from customer preference for human interaction, accessibility needs, and occasional cash-handling regulations, but no licensing or legal requirement mandates a human seller.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	A kiosk or self-service terminal has high upfront capital cost but minimal per-transaction operating cost; amortized over volume, the cost per ticket sold is substantially lower than a cashier's fully-loaded wage, especially in high-throughput venues.
Cost vs. human wage	claude-sonnet-5	4/5	Automated kiosks and online sales systems have low marginal cost per transaction compared to a cashier's wage, though hardware and maintenance costs keep it from being a full order of magnitude cheaper in all contexts.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Automated ticketing and point-of-sale systems are deployed in production across venues, airports, and retail worldwide; self-checkout and ticket kiosks operate reliably for routine transactions at scale. Error rates remain elevated for edge cases but the core transaction flow is mature and operational.
Technical feasibility today	claude-sonnet-5	4/5	Self-service kiosks, automated ticket machines, and e-commerce checkout systems are widely deployed in production across retail, transit, and entertainment venues today.

Cash checks for customers.

CI 60–81 · exposure 80 · augmentation 50 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	5/5	Financial services have aggressively adopted automated check processing for decades; mobile deposit, ATM check capture, and digital payment alternatives have displaced in-person check cashing at high velocity across the sector.
Sector adoption velocity	claude-sonnet-5	2/5	Retail and cash-handling sectors adopt automation unevenly and slowly compared to information/finance sectors, with many small retailers still relying on manual cashiers for this task.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI assists cashiers by flagging suspicious checks, verifying signatures, and processing routine deposits faster, but the task itself is being displaced rather than augmented as digital payments rise.
Augmentation potential	claude-sonnet-5	3/5	AI-assisted fraud detection and ID verification tools can help cashiers process checks faster and more securely, though the core interaction remains largely procedural rather than judgment-intensive.
Task automatability	claude-haiku-4-5-20251001	4/5	Check processing can be largely automated through image capture, OCR, signature verification, and fraud detection systems. Modern mobile deposit and check scanning eliminate most manual handling, though final authorization typically requires human oversight, preventing a full 5.
Task automatability	claude-sonnet-5	4/5	Check cashing is largely a rules-based verification and transaction process that can be handled by kiosks, ATMs, or automated check-cashing machines with high reliability, though some edge cases (fraud judgment, ID verification) still need human oversight.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Banking regulations, anti-fraud requirements, and liability rules create meaningful legal and compliance barriers; banks must follow federal check clearing standards and maintain human oversight for suspicious or large transactions.
Adoption barriers	claude-sonnet-5	3/5	Financial transactions involving cash and identity verification carry regulatory requirements (KYC, anti-fraud, occasional dispute resolution) that create moderate friction, though not requiring a licensed professional.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	Automated check processing costs pennies per transaction while a cashier's fully-loaded wage makes manual check cashing economically uncompetitive, easily an order of magnitude difference at scale.
Cost vs. human wage	claude-sonnet-5	4/5	Once installed, automated check-cashing machines process transactions at a fraction of the marginal labor cost per transaction compared to a cashier's wage, though upfront hardware costs exist.
Technical feasibility today	claude-haiku-4-5-20251001	5/5	Deployed products (mobile check deposits, automated clearing house systems, bank ATMs with check readers) reliably process millions of checks daily in production across major financial institutions.
Technical feasibility today	claude-sonnet-5	4/5	Automated check-cashing kiosks and ATMs already perform this reliably in production at grocery stores, check-cashing chains, and banks, though not universally deployed at every register.

Answer incoming phone calls.

CI 59–81 · exposure 62 · augmentation 50 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Retail adoption of AI phone systems is growing but uneven; many small cashier-heavy businesses still rely on humans, though larger chains are piloting voice agents. Widespread production deployment remains slower than in finance or tech sectors.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is a moderately digitized sector; basic call automation (IVR, chatbots) is common but full replacement of cashier phone duties is uneven and often still routed to staff.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI can assist cashiers by screening and routing calls, summarizing caller intent, or suggesting relevant information before transfer. This augments cashier productivity on phone-heavy shifts but does not transform the core task on its own.
Augmentation potential	claude-sonnet-5	3/5	AI can pre-screen, route, or provide info for common calls, reducing interruptions for cashiers, though it doesn't fully eliminate the need for a human to step in for many calls.
Task automatability	claude-haiku-4-5-20251001	4/5	AI voice systems (IVR, conversational agents) can handle a large portion of incoming calls end-to-end for routine inquiries, transfers, or information provision, achieving significant time savings. However, complex or escalated calls may still require human intervention, preventing a full 5 rating.
Task automatability	claude-sonnet-5	3/5	Voice AI/IVR systems can handle routine incoming calls (store hours, order status) but complex or ambiguous customer queries in a retail cashier context still require human handling for full quality parity.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few legal or regulatory barriers mandate human phone answering in most retail contexts. Minimal liability risk for simple call routing or information provision; customer preference for human contact is the main friction, not regulatory.
Adoption barriers	claude-sonnet-5	1/5	Answering phone calls at a retail register has no licensing or legal requirement for human performance, and customers generally accept automated systems for basic inquiries.
Cost vs. human wage	claude-haiku-4-5-20251001	5/5	AI inference and integration costs for call handling are orders of magnitude cheaper than paying a cashier's wage for the equivalent task, especially at volume. A single AI system handles thousands of calls daily with minimal marginal cost.
Cost vs. human wage	claude-sonnet-5	4/5	Automated phone-answering systems are cheap to run per call compared to paying a cashier's time to answer phones, especially at scale.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Deployed AI phone systems (e.g., Google Duplex, customer service chatbots integrated with phone lines) are in production at scale, handling millions of calls reliably. Minor gaps remain in edge cases and emotionally complex scenarios, justifying a 4 rather than 5.
Technical feasibility today	claude-sonnet-5	3/5	AI phone answering products (IVR, conversational voice bots) are deployed widely in retail and customer service, but often with narrow scripts and frequent handoffs to humans for anything nonstandard.

Issue trading stamps, and redeem food stamps and coupons.

CI 44–75 · exposure 50 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	Retail has rapidly deployed self-checkout and automated coupon scanning over the past decade; many chains now operate self-service terminals at scale, reflecting fast adoption in digitized retail environments.
Sector adoption velocity	claude-sonnet-5	4/5	Retail and grocery sectors have rapidly adopted self-checkout and automated payment processing over the past decade, though full replacement of cashiers remains partial.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-assisted coupon validation and fraud detection can improve a cashier's accuracy and speed, but the core task remains straightforward enough that augmentation provides incremental rather than transformative gains.
Augmentation potential	claude-sonnet-5	3/5	AI-enabled point-of-sale systems assist cashiers with coupon validation, stamp redemption tracking, and fraud detection, improving speed and accuracy.
Task automatability	claude-haiku-4-5-20251001	2/5	Issuing stamps and redeeming coupons involve some automation potential (barcode scanning, inventory tracking), but the task also requires judgment calls on coupon validity, customer interaction, and exception handling that current AI cannot fully resolve end-to-end without human oversight.
Task automatability	claude-sonnet-5	4/5	Self-checkout systems and automated payment/coupon processing already handle most of this transactional work, though physical scanning and stamp handling still require some human or hardware intervention.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal legal barriers exist, but organizational inertia around customer preference for manned checkout, need for fraud detection, and retailer liability for coupon acceptance create moderate friction to full automation.
Adoption barriers	claude-sonnet-5	2/5	Some regulatory requirements exist around SNAP/EBT verification and coupon fraud prevention, but these are already handled by certified automated systems in many stores.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	Self-checkout and integrated coupon-scanning systems have become cost-competitive with cashier labor in some retail contexts, but integration complexity and exception handling keep the total cost roughly comparable to paying a cashier.
Cost vs. human wage	claude-sonnet-5	4/5	Automated kiosks and scanning systems cost far less per transaction than a staffed cashier once installed, though upfront hardware and maintenance costs are nontrivial.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Self-checkout and point-of-sale systems can handle basic coupon scanning, but redemption verification and handling invalid/expired coupons remains primarily manual in production systems; no deployed product fully automates this task reliably.
Technical feasibility today	claude-sonnet-5	4/5	Self-checkout kiosks, barcode scanners, and coupon/EBT processing systems are widely deployed in production at grocery and retail chains today.

Bag, box, wrap, or gift-wrap merchandise, and prepare packages for shipment.

CI 26–89 · exposure 53 · augmentation 13 · importance 4.1/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	4/5	E-commerce fulfillment, third-party logistics, and large retailers are rapidly deploying automated packaging systems in production (Amazon, Shopify-connected warehouses). Adoption is fastest in high-digitization, high-volume sectors; traditional small retail adoption lags.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting self-checkout and automation in inventory/POS systems, but physical packaging/bagging automation via robots is rare and slow to deploy.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI-assisted tools (picking optimization, wrapping guidance) offer modest support to human packers, but the core task is mechanical and does not benefit substantially from AI augmentation; replacement is far more relevant than assistance.
Augmentation potential	claude-sonnet-5	1/5	AI offers essentially no meaningful assistance to a human performing physical bagging, boxing, or wrapping of merchandise.
Task automatability	claude-haiku-4-5-20251001	5/5	Bagging, boxing, wrapping, and packaging are repetitive, rule-based physical operations where robotic systems and computer vision have demonstrated over 50% time savings at equal quality in e-commerce and logistics settings (Amazon, Alibaba warehouses). While full end-to-end automation faces challenges with variable item sizes, current systems handle standard scenarios reliably.
Task automatability	claude-sonnet-5	2/5	This is a physical manipulation task requiring dexterity to handle diverse items, materials, and packaging; current AI (robotics) cannot reliably do this end-to-end at equal quality with major time savings., though self-checkout has shifted some of this to customers rather than AI.'
Adoption barriers	claude-haiku-4-5-20251001	1/5	No licensing, regulatory oversight, or legal requirement mandates a human perform bagging and wrapping. Customer preference for human contact is minimal for this task, and liability risks from automated packaging are low, creating minimal adoption friction.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human perform this, but physical infrastructure, checkout layout, and customer expectations create moderate organizational friction.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Industrial packaging robots cost significantly less per unit throughput than human labor once amortized over large volumes, though integration and maintenance add overhead. In high-velocity distribution centers, the cost ratio heavily favors automation; in low-volume retail it may be marginal.
Cost vs. human wage	claude-sonnet-5	1/5	Robotic manipulation systems capable of this task are far more expensive to acquire, integrate, and maintain than paying a cashier's wage for the same output.
Technical feasibility today	claude-haiku-4-5-20251001	4/5	Robotic bagging and packaging systems are deployed in production at major logistics and retail hubs (e-commerce fulfillment centers), though they typically handle standardized items; performance degrades on irregular items. Most traditional retail cashier stations lack this automation today, but the technology is proven at scale in high-volume environments.
Technical feasibility today	claude-sonnet-5	1/5	No deployed robotic products reliably bag, box, or gift-wrap varied retail merchandise in production; this remains research-stage for general dexterous manipulation.

Answer customers' questions, and provide information on procedures or policies.

CI 47–59 · exposure 42 · augmentation 63 · importance 4.4/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Many large retailers have deployed chatbots or self-service kiosks for FAQs, but adoption remains uneven. Small retailers lag, and many implementations are narrow-scope pilots rather than full replacement of cashier inquiry-handling.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a mixed-digitization sector; self-service and kiosk adoption is growing but slower than white-collar sectors, with many small and mid-size retailers still relying heavily on human cashiers for this task.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI assistants can significantly augment cashiers by providing instant access to policy summaries, return procedures, and promotional details, enabling faster and more confident customer responses without replacing the human entirely.
Augmentation potential	claude-sonnet-5	3/5	AI-powered knowledge bases, translation tools, and quick-reference systems can help cashiers answer policy questions faster and more accurately, providing moderate productivity gains.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can answer scripted, FAQ-like questions about standard policies, but struggles with nuanced customer contexts, exceptions, and real-time procedural knowledge specific to a store's operations. Most questions require human judgment or access to live systems.
Task automatability	claude-sonnet-5	3/5	Simple, repetitive policy/procedure questions (return policy, store hours, payment methods) can be handled by chatbots or kiosk AI, but many questions require real-time contextual judgment or physical assistance that reduces full automation potential.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Few legal barriers exist; retailers can deploy self-service info systems without licensing. Main frictions are customer preference for human contact and retailer liability for incorrect information, but these are soft rather than regulatory barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but customer preference for human interaction and in-person contact expectations at checkout create moderate friction against pure automation.
Cost vs. human wage	claude-haiku-4-5-20251001	4/5	Inference cost for a chatbot is very low (fractions of a cent per query), while a human cashier's loaded wage is $20–30/hour. Even accounting for oversight and integration overhead, AI is substantially cheaper per answer provided.
Cost vs. human wage	claude-sonnet-5	4/5	Automated FAQ/chat systems are far cheaper per interaction than paying a cashier's wage for the same information delivery, though integration and oversight costs offset some savings.
Technical feasibility today	claude-haiku-4-5-20251001	3/5	Chatbots and conversational AI exist in production at some retailers, but they frequently fail on out-of-scope questions, require human escalation, and struggle with accent/dialect variation. Reliability remains material but uneven.
Technical feasibility today	claude-sonnet-5	3/5	Retail chatbots and self-checkout help systems are deployed today (e.g., Walmart, Amazon kiosks) but still have material error rates and narrow scope, often escalating to a human for anything beyond basic queries.

Count money in cash drawers at the beginning of shifts to ensure that amounts are correct and that there is adequate change.

CI 25–69 · exposure 50 · augmentation 63 · importance 4.6/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While some large retailers have invested in coin/bill counters, widespread adoption of automated cash-counting remains limited. Many small-to-medium retailers still use manual counting, and digital payment trends reduce (not increase) the total cash-handling volume, dampening adoption incentives.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is a moderately digitized sector; larger chains have adopted automated cash handling but many small businesses still rely on manual counting, giving a middling adoption pace.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Automated bill and coin counters do assist cashiers by eliminating manual counting labor, reducing human error, and speeding up the verification process. However, the human remains responsible for reconciling discrepancies and ensuring drawer correctness, so the augmentation is partial and procedurally bounded.
Augmentation potential	claude-sonnet-5	4/5	Automated counters and POS systems significantly speed up and improve accuracy of the counting process while cashiers still oversee and verify the results.
Task automatability	claude-haiku-4-5-20251001	2/5	While coin/bill counting can be partially automated with hardware (coin counters, bill counters), the end-to-end task of verifying drawer correctness and assessing change adequacy requires human judgment about discrepancies and reconciliation. Current AI systems lack the dexterity and real-world integration to reliably handle physical cash sorting and counting at scale without significant setup, falling well short of 50% time savings at equal quality.
Task automatability	claude-sonnet-5	4/5	Counting cash is a well-defined, repetitive numeric task that automated cash-management systems and smart safes can perform with high accuracy, saving significant time versus manual counting.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Significant barriers exist: banks and retailers have strict audit and reconciliation requirements, cash handling involves fiduciary responsibility, and most retail operations have compliance protocols requiring a human to sign off on drawer counts. Regulatory and liability requirements mean that humans typically must verify and take responsibility for accuracy.
Adoption barriers	claude-sonnet-5	2/5	There are minor organizational and trust barriers (verifying machine accuracy, occasional manual audits) but no licensing or legal requirement mandating a human count cash.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Dedicated cash-counting hardware is expensive to purchase and maintain, and integration with existing POS systems adds overhead. The cost per transaction is likely comparable to or higher than a cashier's wage for the time spent, especially when accounting for equipment capital and maintenance.
Cost vs. human wage	claude-sonnet-5	3/5	Automated cash counting hardware has upfront capital and maintenance costs that can rival labor costs for small operations, though at scale it becomes cheaper than repeated manual counts.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some specialized cash-counting machines exist in retail, but they require manual feeding and human oversight of results. No deployed AI agent system performs this full task reliably in production; existing solutions are narrow hardware counters, not intelligent systems that verify drawer correctness or make decisions about change adequacy.
Technical feasibility today	claude-sonnet-5	4/5	Deployed products like smart safes, automated coin/bill counters, and POS-integrated cash drawer systems already perform this reliably in many retail and banking environments today.

Process merchandise returns and exchanges.

CI 30–55 · exposure 38 · augmentation 50 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail adoption of autonomous return processing remains limited despite some kiosk pilots; most high-street and e-commerce retailers still rely primarily on human cashiers or customer service staff for returns, with AI playing only a supportive role in existing workflows.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is adopting self-checkout and automated return technology steadily, but adoption is uneven across chains and store sizes, with many still relying primarily on staffed counters.
Augmentation potential	claude-haiku-4-5-20251001	3/5	POS systems with AI-assisted fraud detection, refund policy lookups, and inventory integration can meaningfully assist cashiers in processing returns faster and more accurately, though the human remains essential for inspection and dispute resolution.
Augmentation potential	claude-sonnet-5	3/5	POS systems and integrated inventory/return software assist cashiers by automating lookup, refund calculation, and policy checks, improving speed and accuracy while humans still handle exceptions and customer interaction.
Task automatability	claude-haiku-4-5-20251001	2/5	While modern POS systems can handle refund calculation and inventory updates automatically, the task requires physical inspection of returned items, judgment about condition and eligibility for return, and handling of customer disputes—most of which still require human discretion and cannot be fully automated end-to-end at 50% time savings with current AI.
Task automatability	claude-sonnet-5	3/5	Self-checkout kiosks and automated return systems can handle straightforward returns, but exceptions, fraud checks, and judgment calls on damaged goods or policy edge cases still require human intervention. Roughly half the volume could be automated with existing kiosk/POS technology.
Adoption barriers	claude-haiku-4-5-20251001	3/5	Retailers face customer preference for human interaction during returns, potential liability if automated systems reject valid returns incorrectly, and organizational friction around managing exceptions; however, no strict legal licensing barrier prevents some automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but some friction exists from fraud-prevention needs, customer preference for human interaction during disputes, and store policies requiring manager overrides for exceptions.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of a fully automated return kiosk or robotic inspection system (hardware, maintenance, integration) exceeds the labor cost of a cashier processing returns in most retail environments, especially accounting for exception handling and customer service overhead.
Cost vs. human wage	claude-sonnet-5	3/5	Kiosk hardware, software licensing, and maintenance costs are non-trivial compared to a low-wage cashier, though at high transaction volume the automation can approach cost parity or better.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Current AI can process straightforward refunds via integrated POS systems, but reliable deployed solutions for autonomous assessment of return eligibility (inspecting damaged goods, verifying receipts, handling exceptions) remain limited; most production systems still require cashier oversight and judgment.
Technical feasibility today	claude-sonnet-5	3/5	Retailers deploy self-service return kiosks and automated refund systems in production (e.g., Amazon, big-box retailers), but these systems have narrow scope and often escalate complex cases to staff, so reliability is uneven across contexts.

Assist customers by providing information and resolving their complaints.

CI 30–41 · exposure 25 · augmentation 63 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	3/5	Large retailers and e-commerce have deployed AI-assisted customer service systems, but adoption remains mixed; many small retailers and physical cashier environments still rely primarily on human staff. Pilots are common, but genuine end-to-end AI complaint resolution in production remains limited.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a physical, moderately digitized sector; self-service technology adoption is growing but complaint handling remains largely human-staffed with slow structural change.
Augmentation potential	claude-haiku-4-5-20251001	4/5	AI can draft responses, surface relevant policies, and flag priority issues, substantially raising human cashier productivity in handling routine inquiries and complaints. This assistive augmentation is already present in many retail environments via integrated help systems.
Augmentation potential	claude-sonnet-5	3/5	AI-powered POS systems and chat support can supply cashiers with information, scripts, and policy lookups that speed up complaint resolution, offering meaningful but partial assistance.
Task automatability	claude-haiku-4-5-20251001	2/5	AI can provide basic informational responses and suggest standard complaint resolutions, but the empathetic presence, nuanced judgment, and ability to handle emotionally complex customer interactions at equal quality remain difficult. Current systems often fail to resolve non-standard complaints or adapt to customer tone authentically.
Task automatability	claude-sonnet-5	2/5	Basic informational queries can be handled by chatbots, but genuine complaint resolution requires judgment, empathy, and often exception-making (refunds, policy overrides) that current AI cannot fully execute end-to-end in a physical retail context.4
Adoption barriers	claude-haiku-4-5-20251001	2/5	Customer preference for human interaction and brand reputation risk create meaningful friction, and some high-touch complaints carry legal liability (damage claims, regulatory disputes) requiring human judgment or sign-off. However, no formal licensing or legal mandate requires human complaint resolution.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement, but customer preference for human interaction during complaints, liability for refund decisions, and in-person service expectations create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	3/5	Deployed AI customer service costs (infrastructure, training, oversight) are comparable to entry-level cashier wages per transaction or interaction, especially when accounting for integration and monitoring to catch poor resolutions.
Cost vs. human wage	claude-sonnet-5	2/5	Deploying conversational AI/kiosks has real hardware and integration costs, and complex complaints still require human escalation, so blended cost savings versus a cashier's wage are modest.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Chatbots and AI assistants exist in retail but typically handle only straightforward FAQs and refund policies; they frequently escalate complex complaints to humans or produce generic, unsatisfying responses. No production system reliably resolves diverse customer complaints without human oversight.
Technical feasibility today	claude-sonnet-5	2/5	Self-checkout kiosks and retail chatbots exist but complaint resolution at point-of-sale is still overwhelmingly handled by human cashiers or escalated to managers; no mature product handles this reliably in-store.

Help customers find the location of products.

CI 30–40 · exposure 25 · augmentation 50 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While retail is digitizing, most cashiers still handle product-finding verbally or by directing customers, and few stores have deployed reliable AI systems for this task. Adoption remains limited to larger chains with advanced inventory infrastructure.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a lower-digitization, high-physical-presence sector where AI adoption for basic wayfinding assistance remains limited and mostly pilot-stage.
Augmentation potential	claude-haiku-4-5-20251001	3/5	Cashiers could be augmented by quick-lookup AI tools (e.g., product-location queries to a searchable system at the register), improving their ability to answer customer questions, though this assistance is incremental rather than transformative.
Augmentation potential	claude-sonnet-5	3/5	Store apps, digital maps, and AI chat assistants can help cashiers or customers locate products faster, offering moderate but not transformative assistance.
Task automatability	claude-haiku-4-5-20251001	2/5	Current AI could potentially answer some product-location queries via text/voice interfaces if integrated with store inventory systems, but cashiers typically must navigate complex store layouts, handle real-time inventory changes, and interact with customers in unstructured ways. End-to-end automation that matches human efficiency is not reliably deployable at scale today.
Task automatability	claude-sonnet-5	2/5	Directing customers to product locations requires physical presence, store-specific spatial knowledge, and real-time perception that current general AI cannot perform end-to-end in a physical retail environment.'
Adoption barriers	claude-haiku-4-5-20251001	3/5	Retail organizations can adopt product-locating AI without legal restriction, but physical store presence, customer preference for human help, and integration friction with legacy inventory systems create moderate adoption friction.
Adoption barriers	claude-sonnet-5	1/5	No licensing, legal, or regulatory barriers prevent using signage, apps, or AI directional aids for this task.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Integration costs (inventory system connectivity, AI service fees, hardware/kiosks, maintenance) plus ongoing oversight are substantial. For a task currently performed by low-wage cashiers as part of their role, the all-in automation cost per query likely exceeds the human labor equivalent.
Cost vs. human wage	claude-sonnet-5	2/5	Deploying kiosks, apps, or in-store AI wayfinding systems requires hardware, mapping, and maintenance costs that often exceed the marginal cost of a cashier answering a quick question.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some retailers deploy kiosks or mobile apps with product location data, but these require manual setup, struggle with real-time accuracy, and lack the contextual understanding humans bring. No mature, production-proven solution reliably handles the variability and customer interaction requirements of this task.
Technical feasibility today	claude-sonnet-5	2/5	Some store apps and kiosks with wayfinding or chatbot features exist, but they are narrow, error-prone, and not a full substitute for a cashier physically pointing customers to aisles.

Greet customers entering establishments.

CI 5–60 · exposure 25 · augmentation 25 · importance 4.7/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail adoption of AI for greeting remains negligible; the sector has not deployed such systems at meaningful scale, and current adoption patterns show continued reliance on human greeters.
Sector adoption velocity	claude-sonnet-5	3/5	Retail is adopting self-service and automated kiosks at a moderate pace, with self-checkout widespread but full greeting automation less prioritized as a standalone feature.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI offers minimal assistance to a human greeting task; the action is inherently human-to-human and requires presence, so digital tools provide limited enhancement value.
Augmentation potential	claude-sonnet-5	2/5	AI offers minimal augmentation for this micro-task since greeting is quick and doesn't require significant cognitive assistance for a human cashier.
Task automatability	claude-haiku-4-5-20251001	1/5	Greeting customers requires live human presence and genuine interpersonal engagement, which current AI systems cannot deliver in physical retail settings. While chatbots can simulate greetings digitally, the task specifically involves in-person interaction at a physical establishment entrance.
Task automatability	claude-sonnet-5	3/5	Simple verbal greeting can be automated via kiosks, self-checkout prompts, or voice systems, but full physical presence and adaptive social interaction still require setup and hardware beyond pure software AI.rat.io benefit is limited by the triviality of the task itself.rateduration.i.e it saves little actual time since greeting takes seconds already.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong adoption barriers exist: customer preference for human interaction, brand image concerns, liability for physical systems in high-traffic areas, and organizational reluctance to remove the human touch from first customer contact.
Adoption barriers	claude-sonnet-5	1/5	No licensing, legal, or liability barriers prevent automating a simple greeting; many stores already use automated entry chimes or self-checkout prompts.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The capital, maintenance, and integration costs of autonomous greeting systems far exceed the wage of a cashier or door greeter, with no demonstrated ROI in actual deployments.
Cost vs. human wage	claude-sonnet-5	3/5	Kiosks and automated greeting systems have upfront hardware/software costs but low marginal cost per interaction, roughly comparable to the minimal wage cost of a few seconds of cashier time.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs live customer greeting in retail environments; robotic or AI systems attempting this lack naturalness, contextual sensitivity, and genuine social presence that customers expect and that drive retail engagement.
Technical feasibility today	claude-sonnet-5	3/5	Automated greeting systems (self-checkout kiosks, digital signage with voice prompts) exist and are deployed in retail, but they are narrow-scope and don't replicate genuine human greeting interaction quality.

Stock shelves, sort and reshelve returned items, and mark prices on items and shelves.

CI 29–35 · exposure 25 · augmentation 25 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Despite pilot projects, shelf-stocking automation remains rare in production across retail. Most stores still rely on human cashiers and stock associates; adoption is slow due to high capex, technical immaturity, and the cost-benefit threshold not yet crossed at scale.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a low-to-moderate digitization sector for physical tasks; while self-checkout and electronic pricing are spreading, robotic shelf-stocking remains largely pilot-stage.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Barcode scanners and mobile price-checking tools offer modest assistance, but AI augmentation is minimal; the task is largely manual and procedural, with limited opportunity for AI to substantially raise human productivity while remaining in the loop.
Augmentation potential	claude-sonnet-5	2/5	AI-based inventory and pricing systems can inform workers where and how to restock or price items, offering some assistance, but the physical execution remains manual and unaided by AI directly.
Task automatability	claude-haiku-4-5-20251001	2/5	While some components—like barcode scanning and price marking—could be partially automated, the task requires physical manipulation in unstructured retail environments (sorting, shelving on variable shelves, handling diverse items) where current robots lack dexterity and speed. End-to-end automation with 50% time savings is not yet demonstrable with off-the-shelf systems.
Task automatability	claude-sonnet-5	2/5	Stocking shelves, sorting returns, and marking prices requires physical manipulation of items in variable store layouts, which current AI systems cannot perform end-to-end without robotic hardware that is not yet widely deployed for this purpose.dark
Adoption barriers	claude-haiku-4-5-20251001	2/5	Retail chains have flexibility to deploy automation, but customer preference for human interaction, union rules in some chains, and the low wage of cashiers reduce urgency. No legal requirement mandates a human, but no strong regulatory barrier blocks automation either.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal barriers exist, but physical store environments, variable product types, and existing human labor infrastructure create moderate organizational friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current robotic shelf-stocking solutions remain capital-intensive and slow relative to human cashiers; the installed cost, maintenance, and integration overhead far exceed the wage cost of a part-time or full-time cashier performing this task.
Cost vs. human wage	claude-sonnet-5	2/5	Robotic stocking solutions require significant capital investment, maintenance, and integration costs that generally exceed the low wage cost of cashier labor for this task in most retail settings.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Shelf-stocking robots exist in research and limited pilot deployments (e.g., Capgemini/Walmart trials), but they struggle with real-world variability, item diversity, and speed. No mature, production-scale product reliably handles the full task across standard retail environments today.
Technical feasibility today	claude-sonnet-5	2/5	Some automated pricing (electronic shelf labels) and warehouse robots exist, but reliable general-purpose shelf-stocking and returns-sorting robots are not deployed at scale in typical retail stores today.

Monitor checkout stations to ensure they have adequate cash available and are staffed appropriately.

CI 25–35 · exposure 25 · augmentation 50 · importance 4.5/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	While large retailers have adopted POS and workforce management dashboards, adoption of autonomous monitoring and decisioning remains limited. Most chains still rely on human shift managers for these oversight functions; uptake of AI-driven automation for this specific task is still in pilot or early deployment phases rather than mainstream production.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is adopting AI for inventory and self-checkout monitoring, but supervisory staffing/cash oversight functions remain largely human-run with slow uptake of full automation.
Augmentation potential	claude-haiku-4-5-20251001	3/5	AI-powered dashboards and alerts (cash level tracking, predictive staffing need estimation based on traffic patterns) genuinely assist human managers by surfacing data they would otherwise have to manually check, reducing time spent on routine monitoring and flagging exceptions for faster response.
Augmentation potential	claude-sonnet-5	3/5	AI-powered dashboards and queue-length/camera analytics can alert supervisors to low cash or staffing gaps, improving their responsiveness even though humans still make and execute decisions.
Task automatability	claude-haiku-4-5-20251001	2/5	While AI could monitor cash levels via data feeds and staffing dashboards, the task requires real-time judgment about what constitutes 'adequate' cash (context-dependent on transaction patterns, time of day, inventory) and appropriate staffing levels (balancing customer flow, break schedules, skill mix). Current systems can flag low balances but cannot reliably replace the human decision-making needed for operational adjustments.
Task automatability	claude-sonnet-5	2/5	This is a supervisory, physical-presence task involving real-time staffing decisions and cash reconciliation across a physical space, which current AI cannot fully perform end-to-end without human oversight and action.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Retail management chains typically require a human manager or supervisor to be physically present and legally accountable for cash handling, inventory control, and staff scheduling. Store policies, loss-prevention protocols, and labor law often mandate human sign-off on cash reconciliation and staffing decisions, creating a regulatory and operational barrier to full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing requirement, but organizational trust in automated staffing/cash oversight and liability for cash handling errors create moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	Existing monitoring software (POS integration, workforce analytics) requires setup, licensing, and ongoing human oversight to act on alerts. For a routine supervisory task often handled by a single shift manager, the infrastructure cost is not yet dramatically cheaper than the occasional human attention the role already receives.
Cost vs. human wage	claude-sonnet-5	2/5	Sensor/POS-based monitoring tools exist but require integration, hardware, and human decision-making on top, so total cost is not dramatically lower than having a supervisor glance at stations.
Technical feasibility today	claude-haiku-4-5-20251001	2/5	Some POS and workforce management systems offer alerts for low cash or understaffing, but these are narrow components of the full monitoring task and require substantial manual review and override by managers. No end-to-end deployed product reliably handles the contextual judgment of whether monitored conditions actually warrant intervention.
Technical feasibility today	claude-sonnet-5	2/5	Some retail systems provide dashboards flagging low cash drawers or queue lengths, but the actual monitoring and staffing decisions still require a human supervisor to interpret and act.

Maintain clean and orderly checkout areas, and complete other general cleaning duties, such as mopping floors and emptying trash cans.

CI 15–24 · exposure 8 · augmentation 0 · importance 4.0/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail remains a laggard sector for physical automation. Most checkout areas rely on human cleaners; robotic adoption in this specific context is negligible, with pilots rare and production deployment nearly nonexistent.
Sector adoption velocity	claude-sonnet-5	1/5	Retail cleaning automation (robotic mopping etc.) sees minimal deployment in checkout areas; this sub-task lags far behind office/digital task automation trends.
Augmentation potential	claude-haiku-4-5-20251001	1/5	Current AI systems offer no meaningful assistance for floor mopping, trash disposal, or organizing checkout areas. These are physical tasks where AI augmentation is not applicable.
Augmentation potential	claude-sonnet-5	1/5	AI offers essentially no productivity assistance for manual cleaning tasks like mopping or emptying trash cans.
Task automatability	claude-haiku-4-5-20251001	2/5	While robots exist for some cleaning tasks, mopping floors and emptying trash in active checkout areas require navigation, object manipulation, and adaptation to variable layouts—capabilities current AI systems lack reliably. Maintaining 50% time savings at equal quality across the full task remains infeasible today.
Task automatability	claude-sonnet-5	1/5	This is a physical manual labor task (mopping, tidying, trash disposal) that requires embodied manipulation in unstructured environments, which current AI (software-based) cannot perform at all.
Adoption barriers	claude-haiku-4-5-20251001	2/5	Minimal legal barriers exist, but physical and operational friction is high: navigation in crowded retail spaces, safety liability, integration with existing store infrastructure, and organizational preference for human workers who perform multiple checkout tasks all slow substitution.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human for cleaning, but practical barriers like the need for dexterous mobile robots in cluttered public spaces limit substitution.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Robotic systems capable of mopping and trash removal cost tens of thousands of dollars, require maintenance and integration, and operate slowly. The all-in cost far exceeds the loaded wage of a cashier performing light cleaning duties.
Cost vs. human wage	claude-sonnet-5	1/5	Physical cleaning robots capable of mopping and trash handling in dynamic retail environments are expensive capital purchases plus maintenance, far exceeding the marginal cost of a cashier's incidental cleaning duties.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs this end-to-end task (floor mopping, trash disposal, organizing checkout areas) in real retail environments. Cleaning robots exist but are narrow, slow, and require controlled environments—not production solutions in active checkout zones.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product performs general physical cleaning of retail checkout areas; commercial cleaning robots exist only for narrow floor-cleaning niches and are not integrated into cashier workflows.

Assist with duties in other areas of the store, such as monitoring fitting rooms or bagging and carrying out customers' items.

CI 10–26 · exposure 8 · augmentation 13 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	2/5	Retail automation has focused on self-checkout and backend logistics rather than cashier-adjacent floor duties. Adoption of AI for this specific task mix is minimal; most stores still use human staff for these roles.
Sector adoption velocity	claude-sonnet-5	1/5	Retail floor physical assistance tasks show negligible AI/robotic adoption; retail automation efforts focus on checkout and inventory, not physical customer assistance and monitoring.
Augmentation potential	claude-haiku-4-5-20251001	2/5	Current AI tools offer minimal assistance with fitting-room monitoring or bagging tasks. Computer vision for inventory tracking is emerging but not yet reliable enough to materially augment a cashier's productivity in practice.
Augmentation potential	claude-sonnet-5	1/5	AI tools offer essentially no assistance for physically monitoring fitting rooms or carrying/bagging items, as these are manual, situational, in-person tasks.
Task automatability	claude-haiku-4-5-20251001	2/5	While bagging and basic inventory monitoring of fitting rooms have some automatable elements, the task requires physical manipulation, customer interaction, and contextual judgment that current AI agents cannot reliably perform end-to-end. The carrying/physical handling component is entirely beyond current capabilities, preventing the 50% time-saving threshold.
Task automatability	claude-sonnet-5	1/5	This task requires physical presence, mobility, and manipulation of physical items (bagging, carrying, monitoring physical spaces), which current AI systems cannot perform without embodied robotics that are not generally deployed for this purpose.
Adoption barriers	claude-haiku-4-5-20251001	3/5	While there are no strict legal barriers preventing automation, retail stores value human customer service and presence on the floor, and the physical labor component creates organizational inertia. However, these are soft rather than hard barriers.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal barriers exist, but the physical nature of the task creates a practical barrier since no scalable automated alternative exists yet.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	Current automation for this task (if it existed) would require expensive robotics and computer vision integration that far exceeds the loaded wage of a part-time cashier assistant, making it economically infeasible at scale.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute for the physical labor involved, so any comparison would require robotics with costs far exceeding low-wage cashier labor for this task.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed product reliably performs the full scope of this task (fitting-room monitoring, bagging, and carrying) in retail environments. Robotics for item handling exist only in narrow, controlled settings and are not in general retail deployment.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product performs fitting room monitoring or physical bagging/carrying of retail items at scale; this remains a physical labor task outside AI's current capability envelope.

Offer customers carry-out service at the completion of transactions.

CI 5–28 · exposure 8 · augmentation 13 · importance 3.9/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail has been slow to adopt AI for checkout-floor customer interactions; most automation focuses on self-checkout kiosks or backend inventory, not on automating the offer of services that require social judgment and customer rapport.
Sector adoption velocity	claude-sonnet-5	1/5	Retail cashiering is a low-digitization, physical-labor sector where AI/robotic adoption for carry-out services is essentially nonexistent in production.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist by alerting cashiers to cues that a customer might benefit from carry-out service, but the core task—reading customer intent and making a personalized offer—remains fundamentally human and benefits only marginally from AI support.
Augmentation potential	claude-sonnet-5	1/5	AI offers no meaningful assistance to a human performing this specific physical carry-out task.
Task automatability	claude-haiku-4-5-20251001	2/5	Offering carry-out service requires recognizing transaction completion, verbal or gestural interaction with customers, and contextual judgment about customer needs. Current AI systems cannot reliably perform the full interaction end-to-end; this task involves social presence and real-time responsiveness that AI cannot replicate in physical retail environments today.
Task automatability	claude-sonnet-5	1/5	This is a physical task requiring lifting and carrying items to a customer's vehicle or location, which current AI systems cannot perform; only robotics could address it, and that is not deployed for this purpose.'
Adoption barriers	claude-haiku-4-5-20251001	4/5	Strong barriers exist: customers strongly prefer human interaction at checkout, retail employers value the personal touch for service differentiation, and legal liability for mishandling transactions or customer injury creates organizational friction against full automation.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal requirement mandates a human specifically, but the physical nature and customer service expectation create practical barriers to automation without robotics.
Cost vs. human wage	claude-haiku-4-5-20251001	2/5	The cost of deploying a system capable of detecting transaction completion, engaging customers, and offering carry-out service would be substantial relative to the wage of a cashier performing this brief interaction—which is a low-marginal-cost add-on to their existing role.
Cost vs. human wage	claude-sonnet-5	1/5	There is no AI substitute performing this physical task, so any AI cost comparison is moot; a human remains the only viable option and thus cheaper by default.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably performs this interpersonal customer-service task in production cashier settings. While chatbots exist, they cannot physically interact with customers at checkout or assess when to offer this service in real time.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product offers physical carry-out assistance to customers; this remains a purely human physical service task.

Request information or assistance, using paging systems.

CI 5–26 · exposure 8 · augmentation 0 · importance 4.3/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail remains a laggard sector for AI adoption, and paging-based communication is a low-tech, low-frequency task that generates no pressure for automation. There is no evidence of pilot or production deployments targeting this specific function.
Sector adoption velocity	claude-sonnet-5	2/5	Retail cashiering is a physical, low-digitization environment where AI adoption for such micro-tasks is minimal and not a current focus of automation efforts.
Augmentation potential	claude-haiku-4-5-20251001	1/5	AI offers no meaningful assistance for the simple act of requesting help via a paging system; the task is straightforward communication that requires only human judgment about when help is needed, which AI cannot reliably predict.
Augmentation potential	claude-sonnet-5	1/5	AI offers negligible assistance for this simple, situational communication task that is already fast and low-cost for a human to perform manually.
Task automatability	claude-haiku-4-5-20251001	1/5	This task requires initiating human communication via paging systems to solicit help or information. It is fundamentally an interpersonal request action that does not reduce to data processing, and current AI cannot independently operate paging systems or meaningfully replace the human judgment of when and whom to contact.
Task automatability	claude-sonnet-5	2/5	This is a brief, situational physical/verbal action (calling for a manager or price check via intercom) that requires human presence at the register and real-world triggering events, limiting end-to-end automation.led by AI.
Adoption barriers	claude-haiku-4-5-20251001	4/5	Store management systems and paging infrastructure are typically proprietary and access-controlled. Additionally, human judgment about whether and when to request assistance remains valuable, and organizations prefer human accountability for communication on the shop floor.
Adoption barriers	claude-sonnet-5	2/5	No licensing or legal barrier exists, but the task requires physical presence and integration with store paging hardware, creating practical friction against automation.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The task is trivial in human cost (a few seconds of a cashier's time) and requires no specialized training. Any AI implementation would require system integration, maintenance, and oversight that would far exceed the minimal labor cost of occasional paging requests.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this specific micro-task, so cost comparison favors the human cashier who already performs it as an incidental part of the job.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product today can autonomously use paging systems to request assistance in a retail cashier context. This would require real-time integration with proprietary store systems and contextual judgment about what information or assistance is needed.
Technical feasibility today	claude-sonnet-5	1/5	No deployed AI product autonomously requests in-person assistance via paging systems at a retail register; this remains a manual, human-initiated action tied to physical store operations.

Supervise others and provide on-the-job training.

CI 0–13 · exposure 0 · augmentation 38 · importance 4.2/5 · click for rater detail

Panel ratings per dimension for this task
Dimension	Model	Rating	Rationale
Sector adoption velocity	claude-haiku-4-5-20251001	1/5	Retail sectors have minimal adoption of AI-driven supervision; most workers remain supervised by human managers and team leads, reflecting both regulatory and cultural resistance to algorithmic management.
Sector adoption velocity	claude-sonnet-5	2/5	Retail is a low-digitization sector with slow AI adoption for management functions, though e-learning tools are used for training content delivery.
Augmentation potential	claude-haiku-4-5-20251001	2/5	AI could assist supervisors with scheduling analysis or performance data summaries, but current systems offer limited augmentation for the core tasks of mentoring, real-time feedback, and personnel decision-making.
Augmentation potential	claude-sonnet-5	3/5	AI can assist with training materials, checklists, and performance tracking, but the core supervisory and mentoring interaction still relies on the human trainer.
Task automatability	claude-haiku-4-5-20251001	1/5	Supervising others and providing on-the-job training inherently require real-time human judgment, empathy, and adaptive coaching based on individual worker performance and learning styles. Current AI systems cannot reliably perform these interpersonal and evaluative functions end-to-end.
Task automatability	claude-sonnet-5	1/5	Supervising staff and delivering hands-on training requires interpersonal judgment, motivation, and real-time coaching that current AI cannot perform end-to-end.
Adoption barriers	claude-haiku-4-5-20251001	5/5	Labor law, management liability, and employment regulation create legal and organizational requirements that a human supervisor or manager must perform this function; firms cannot delegate core supervision to AI alone without legal risk.
Adoption barriers	claude-sonnet-5	3/5	No licensing requirement exists, but organizational structure and accountability norms mean supervisory roles are typically retained by humans, creating moderate friction.
Cost vs. human wage	claude-haiku-4-5-20251001	1/5	The cost of AI systems capable of any meaningful supervision (including cameras, NLP, and oversight infrastructure) plus human review would exceed the cost of a human supervisor in most retail contexts.
Cost vs. human wage	claude-sonnet-5	1/5	There is no viable AI substitute performing this task, so cost comparison favors the human by default.
Technical feasibility today	claude-haiku-4-5-20251001	1/5	No deployed AI product reliably supervises workers or delivers personalized on-the-job training in cashier environments today. While AI can assist with training content delivery, it cannot replace human supervisors who must make personnel decisions and provide adaptive coaching.
Technical feasibility today	claude-sonnet-5	1/5	No deployed product autonomously supervises retail employees or conducts on-the-job training; this remains a human management function.

Related occupations — Sales & Related

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.