{ "version": "V7", "generated_at": "2026-05-15T08:44:09.356Z", "entry_count": 25, "domain_counts": { "exposure": 7, "tasks": 8, "complementarity": 1, "validation": 7, "mobility": 1, "forecast": 5, "uncertainty": 1, "productivity": 4, "augmentation": 4, "measurement": 11, "context": 7 }, "role_counts": { "active_core": 15, "validation": 1, "candidate_v5": 2, "supporting_context": 7 }, "entries": [ { "key": "anthropic_economic_index_2026", "title": "Anthropic Economic Index: New building blocks for understanding AI use", "authors": [ "Anthropic" ], "year": 2026, "published_at": "2026-01-15", "publisher": "Anthropic", "url": "https://www.anthropic.com/economic-futures/", "type": "report", "domains": [ "exposure", "tasks", "measurement" ], "role": "active_core", "status": "active", "used_for": [ "observed occupation exposure source", "usage gap framing", "task evidence design" ], "source_keys": [ "anthropic_economic_index_2026" ], "claim_ids": [ "deterministic_no_llm_core", "reliability_weighted_exposure_ensemble" ], "summary": "Adds observed AI-usage evidence to the exposure stack and motivates the repo's task-primitives sidecar.", "limitations": "Observed Claude usage is not a full labour-market census and is still a platform-specific measure.", "repo_notes": "Active live source in the audited exposure ensemble and a major input to the future task-native direction." }, { "key": "anthropic_labor_market_impacts_2026", "title": "Labor market impacts of AI: A new measure and early evidence", "authors": [ "Maxim Massenkoff", "Peter McCrory" ], "year": 2026, "published_at": "2026-03-05", "publisher": "Anthropic", "url": "https://www.anthropic.com/research/labor-market-impacts", "type": "report", "domains": [ "tasks", "validation", "forecast" ], "role": "active_core", "status": "active", "used_for": [ "observed exposure framing", "task-native shadow model", "near-term impact interpretation" ], "source_keys": [], "claim_ids": [], "summary": "Separates theoretical capability from observed exposure and emphasizes that early labour effects remain limited.", "limitations": "Uses US outcome data and a platform-linked usage measure, so it still needs Singapore-specific interpretation.", "repo_notes": "Primary candidate reference for promoting the shadow model beyond readiness-only governance." }, { "key": "onet_database_2024", "title": "O*NET Database 30.2", "authors": [ "O*NET Resource Center" ], "year": 2026, "published_at": "2026", "publisher": "O*NET / U.S. Department of Labor", "url": "https://www.onetcenter.org/database.html", "type": "dataset", "domains": [ "tasks", "context", "measurement" ], "role": "supporting_context", "status": "supporting", "used_for": [ "task context", "technology-skill context", "job-zone education proxy", "task-primitives matching" ], "source_keys": [ "onet_occupation_data", "onet_task_statements", "onet_technology_skills", "onet_job_zones" ], "claim_ids": [ "onet_task_and_technology_context" ], "summary": "Provides the task statements, technology skills, and job-zone context used for explanation and task matching.", "limitations": "O*NET is US-based and enters the Singapore product mainly as explanatory or crosswalk context.", "repo_notes": "Not a direct structural-score input, but essential for the supporting task layer and future task-native scoring." }, { "key": "metr_time_horizons_2026", "title": "Task-Completion Time Horizons of Frontier AI Models", "authors": [ "METR" ], "year": 2026, "published_at": "2026-03-03", "publisher": "METR", "url": "https://metr.org/time-horizons/", "type": "report", "domains": [ "forecast", "measurement" ], "role": "candidate_v5", "status": "candidate", "used_for": [ "capability-horizon calibration" ], "source_keys": [], "claim_ids": [], "summary": "Tracks the current task-duration horizons of frontier models and offers a capability input for scenario calibration.", "limitations": "The benchmark is model-centric and software-task heavy, so it is not a direct occupation impact measure.", "repo_notes": "Useful for the forecast layer, not for the core structural score." }, { "key": "dillon_etal_2025", "title": "AI, Productivity, and Work Quality", "authors": [ "Erica Dillon", "et al." ], "year": 2025, "published_at": "2025", "publisher": "NBER", "url": "https://www.nber.org/papers/w33795", "doi": "10.3386/w33795", "type": "working_paper", "domains": [ "productivity", "augmentation" ], "role": "active_core", "status": "active", "used_for": [ "augmentation calibration priors", "work-quality tradeoff framing" ], "source_keys": [], "claim_ids": [], "summary": "Adds evidence that AI can change both output quantity and work quality, reinforcing the need for occupation-specific augmentation priors.", "limitations": "Experimental and workflow-specific evidence still needs careful translation into occupation-level scoring.", "repo_notes": "Useful as a V5.1 calibration reference rather than as a direct V4.x score ingredient." }, { "key": "hampole_etal_2025", "title": "Artificial Intelligence and the Labor Market", "authors": [ "Menaka Hampole", "Dimitris Papanikolaou", "Lawrence D.W. Schmidt", "Bryan Seegmiller" ], "year": 2025, "published_at": "2025-02", "publisher": "NBER", "url": "https://www.nber.org/papers/w33509", "doi": "10.3386/w33509", "type": "working_paper", "domains": [ "tasks", "validation", "measurement" ], "role": "active_core", "status": "active", "used_for": [ "task concentration buffer", "task-native demand interpretation" ], "source_keys": [], "claim_ids": [], "summary": "Shows that mean exposure and concentration of exposure in a few tasks can have different labour-demand implications.", "limitations": "The repo uses a simplified shadow-model concentration buffer rather than the paper's full firm-task empirical setting.", "repo_notes": "Primary scientific justification for publishing task exposure concentration as its own field and using it in the shadow model." }, { "key": "bls_skills_data_2025", "title": "BLS Skills Data", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2025, "published_at": "2025", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/emp/data/skills-data.htm", "type": "dataset", "domains": [ "tasks", "measurement", "context" ], "role": "supporting_context", "status": "supporting", "used_for": [ "US skills context", "task explanation", "occupation comparison" ], "source_keys": [ "bls_skills_data_2025" ], "claim_ids": [], "summary": "Official 17-skill summary for occupations, built on O*NET inputs and BLS projections occupations.", "limitations": "It is a summarized skills view, not a direct task-by-task automation score.", "repo_notes": "Best public structured skill overlay for the US country layer." }, { "key": "bls_cps_demographics_2025", "title": "CPS Demographics by Occupation", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2025, "published_at": "2025", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/cps/cpsoccind.htm", "type": "dataset", "domains": [ "measurement", "context", "validation" ], "role": "supporting_context", "status": "supporting", "used_for": [ "US worker profile", "demographic transition context", "equity analysis" ], "source_keys": [ "bls_cps_demographics_2025" ], "claim_ids": [], "summary": "Occupation-linked demographic and labor-force context drawn from the Current Population Survey.", "limitations": "Occupation/industry classification changes reduce perfect longitudinal comparability.", "repo_notes": "Useful for showing who is exposed and for planning transition context, not for changing the structural pressure score." }, { "key": "ilo_genai_exposure_2025", "title": "Generative AI and Jobs: A Refined Global Index of Occupational Exposure", "authors": [ "ILO" ], "year": 2025, "published_at": "2025", "publisher": "International Labour Organization", "url": "https://www.ilo.org/publications/generative-ai-and-jobs-refined-global-index-occupational-exposure", "type": "report", "domains": [ "exposure" ], "role": "active_core", "status": "active", "used_for": [ "ISCO-aligned exposure source" ], "source_keys": [ "ilo_genai_2025" ], "claim_ids": [ "reliability_weighted_exposure_ensemble" ], "summary": "Adds a recent global occupational exposure measure aligned to international occupation codes.", "limitations": "Still a global exposure measure rather than a Singapore outcome model.", "repo_notes": "Included because its ISCO alignment improves crosswalk robustness for the ensemble." }, { "key": "anthropic_work_transforming_2025", "title": "How AI is transforming work at Anthropic", "authors": [ "Anthropic" ], "year": 2025, "published_at": "2025-12-02", "publisher": "Anthropic", "url": "https://www.anthropic.com/research/how-ai-is-transforming-work-at-anthropic", "type": "report", "domains": [ "tasks", "productivity", "augmentation", "context" ], "role": "supporting_context", "status": "supporting", "used_for": [ "software-work case study", "observed adoption context", "learning and iteration framing" ], "source_keys": [], "claim_ids": [], "summary": "Shows how AI changes engineering and research workflows inside Anthropic, adding a concrete adoption case study.", "limitations": "Internal workplace evidence is highly specific to Anthropic and should not be generalized into a country labour market without adjustment.", "repo_notes": "Useful as a contextual adoption case study and a reminder that augmentation can change task mix, learning speed, and workflow design." }, { "key": "humlum_vestergaard_2025", "title": "Large Language Models, Small Labor Market Effects", "authors": [ "Anders Humlum", "Emilie Vestergaard" ], "year": 2025, "published_at": "2025-05", "publisher": "NBER", "url": "https://www.nber.org/papers/w33777", "doi": "10.3386/w33777", "type": "working_paper", "domains": [ "forecast", "validation" ], "role": "active_core", "status": "active", "used_for": [ "near-term realised-risk shrinkage" ], "source_keys": [], "claim_ids": [], "summary": "Finds small early labour-market effects from chatbot adoption despite meaningful task restructuring, supporting conservative near-term risk shrinkage.", "limitations": "The study is Denmark-specific and examines early effects; longer-run displacement remains unresolved.", "repo_notes": "Supports the repo's choice to keep structural risk separate from near-term or realised-risk interpretations." }, { "key": "coyle_poquiz_2025", "title": "Making AI Count: The Next Measurement Frontier", "authors": [ "Diane Coyle", "John Lourenze S. Poquiz" ], "year": 2025, "published_at": "2025-10", "publisher": "NBER", "url": "https://www.nber.org/papers/w34330", "doi": "10.3386/w34330", "type": "working_paper", "domains": [ "measurement", "uncertainty" ], "role": "candidate_v5", "status": "candidate", "used_for": [ "measurement philosophy", "future uncertainty design" ], "source_keys": [], "claim_ids": [], "summary": "Argues that AI measurement should become more granular, task-based, and outcome-focused than current official statistics.", "limitations": "This is a measurement agenda rather than an occupation-scoring formula.", "repo_notes": "Best reference for the repo's longer-run shift from heuristic confidence toward richer uncertainty and task-native measurement." }, { "key": "dellacqua_etal_2025", "title": "Navigating the Jagged Technological Frontier", "authors": [ "Fabrizio Dell'Acqua", "et al." ], "year": 2025, "published_at": "2025", "publisher": "NBER", "url": "https://www.nber.org/papers/w33641", "doi": "10.3386/w33641", "type": "working_paper", "domains": [ "productivity", "augmentation" ], "role": "active_core", "status": "active", "used_for": [ "augmentation heterogeneity", "workflow calibration" ], "source_keys": [], "claim_ids": [], "summary": "Highlights that AI gains are jagged across tasks and expertise levels rather than smooth across a whole occupation.", "limitations": "Provides strong augmentation evidence but still within constrained experimental settings.", "repo_notes": "Supports the repo direction of modelling augmentation as its own construct rather than as a mirror image of automation." }, { "key": "bls_oews_2025", "title": "Occupational Employment and Wage Statistics (OEWS)", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2025, "published_at": "2025", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/oes/", "type": "dataset", "domains": [ "measurement", "context", "validation" ], "role": "supporting_context", "status": "supporting", "used_for": [ "US wage context", "US wage calibration", "occupation-level earnings narratives" ], "source_keys": [ "bls_oews_2025" ], "claim_ids": [], "summary": "Official occupation-level wage and employment statistics for the United States.", "limitations": "OEWS is US-specific and must not be interpreted as a global wage distribution.", "repo_notes": "Best public source for the US wage layer and the richest local wage context currently available." }, { "key": "bls_ooh_2025", "title": "Occupational Outlook Handbook / Occupation Finder", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2025, "published_at": "2025", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/ooh/occupation-finder.htm", "type": "report", "domains": [ "context", "forecast" ], "role": "supporting_context", "status": "supporting", "used_for": [ "US narrative context", "occupation descriptions", "work environment copy" ], "source_keys": [ "bls_ooh_2025" ], "claim_ids": [], "summary": "Official occupation profiles used for narrative, education, and work-environment context.", "limitations": "Designed for career guidance rather than AI-risk scoring.", "repo_notes": "Adds the human-readable occupation narrative layer for the US market." }, { "key": "bls_ors_2025", "title": "Occupational Requirements Survey (ORS)", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2025, "published_at": "2025", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/ors/about-overview.htm", "type": "dataset", "domains": [ "measurement", "context", "validation" ], "role": "supporting_context", "status": "supporting", "used_for": [ "US bottleneck context", "physical and cognitive demand overlay", "transition friction calibration" ], "source_keys": [ "bls_ors_2025" ], "claim_ids": [], "summary": "Official occupation requirements survey covering the physical, cognitive, and preparation demands of work.", "limitations": "ORS is not a labour-market outcome series; it is a requirements survey used for bottleneck and friction context.", "repo_notes": "The best current official analogue for transition friction and human-bottleneck validation in the US layer." }, { "key": "bick_blandin_deming_2025", "title": "The Rapid Adoption of Generative AI", "authors": [ "Alexander Bick", "Adam Blandin", "David J. Deming" ], "year": 2025, "published_at": "2024-09", "publisher": "NBER", "url": "https://www.nber.org/papers/w32966", "doi": "10.3386/w32966", "type": "working_paper", "domains": [ "forecast", "measurement" ], "role": "active_core", "status": "active", "used_for": [ "near-term adoption calibration" ], "source_keys": [], "claim_ids": [], "summary": "Documents that workplace generative-AI adoption is fast, supporting a separate adoption layer in forecast models.", "limitations": "Adoption speed alone does not identify realised labour displacement or productivity effects.", "repo_notes": "Used to justify separating structural pressure from near-term realised-risk proxies." }, { "key": "imf_occupational_mobility_2024", "title": "Exposure to Artificial Intelligence and Occupational Mobility: A Cross-country Analysis", "authors": [ "IMF staff" ], "year": 2024, "published_at": "2024-06-07", "publisher": "IMF Working Paper", "url": "https://www.imf.org/en/publications/wp/issues/2024/06/07/exposure-to-artificial-intelligence-and-occupational-mobility-a-cross-country-analysis-549989", "type": "working_paper", "domains": [ "mobility" ], "role": "active_core", "status": "active", "used_for": [ "future empirical transition model" ], "source_keys": [], "claim_ids": [], "summary": "Suggests that mobility responses to AI pressure follow structured pathways rather than generic occupational distance rules.", "limitations": "Cross-country evidence informs the transition design, but the repo still needs a Singapore-specific mobility dataset.", "repo_notes": "Guides the schema for observed transition priors and future ranking logic in the transition layer." }, { "key": "bls_occupational_projections_2024_2034", "title": "US BLS Employment Projections 2024-2034", "authors": [ "U.S. Bureau of Labor Statistics" ], "year": 2024, "published_at": "2024", "publisher": "U.S. Bureau of Labor Statistics", "url": "https://www.bls.gov/emp/tables/occupational-projections-and-characteristics.htm", "type": "dataset", "domains": [ "validation", "measurement" ], "role": "validation", "status": "active", "used_for": [ "cross-country convergent validation", "employment proxy wage-pool basis" ], "source_keys": [ "bls_projections_2024_2034" ], "claim_ids": [ "bls_proxy_for_wage_pool", "bls_cross_check_directional_only", "confidence_and_mapping_are_calibrated_directionally", "occupation_family_validation_directional" ], "summary": "Provides the external benchmark used for convergent validation and for the BLS-weighted proxy employment field.", "limitations": "This is US evidence and should not be interpreted as Singapore occupation outcome truth.", "repo_notes": "Lives in the repo as both a validation benchmark and a clearly labeled external proxy." }, { "key": "brynjolfsson_li_raymond_2023", "title": "Generative AI at Work", "authors": [ "Erik Brynjolfsson", "Danielle Li", "Lindsey Raymond" ], "year": 2023, "published_at": "2023-04", "publisher": "NBER", "url": "https://www.nber.org/papers/w31161", "doi": "10.3386/w31161", "type": "working_paper", "domains": [ "productivity", "augmentation" ], "role": "active_core", "status": "active", "used_for": [ "augmentation calibration priors" ], "source_keys": [], "claim_ids": [], "summary": "Shows large heterogeneous productivity effects from AI assistance in a specific workflow, supporting separate augmentation modelling.", "limitations": "The effect size is workflow-specific and should not be generalized into a universal augmentation constant.", "repo_notes": "Candidate reference for replacing a single structural augmentation heuristic with workflow-aware priors." }, { "key": "openai_gpts_are_gpts_2023", "title": "GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models", "authors": [ "Tyna Eloundou", "Sam Manning", "Pamela Mishkin", "Daniel Rock" ], "year": 2023, "published_at": "2023-03-17", "publisher": "OpenAI", "url": "https://openai.com/index/gpts-are-gpts/", "type": "article", "domains": [ "exposure", "tasks" ], "role": "active_core", "status": "active", "used_for": [ "LLM-specific task exposure framing", "candidate V5 task-native design" ], "source_keys": [], "claim_ids": [], "summary": "Frames LLM exposure around task feasibility and time-saving potential rather than broad automation narratives.", "limitations": "The paper is early and US-oriented, and it does not by itself provide Singapore labour-market calibration.", "repo_notes": "Used as a reference for the repo's future task-native model direction, not as a direct live source key." }, { "key": "eloundou_etal_2023", "title": "GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models", "authors": [ "Tyna Eloundou", "Sam Manning", "Pamela Mishkin", "Daniel Rock" ], "year": 2023, "published_at": "2023-03", "publisher": "arXiv / OpenAI", "url": "https://arxiv.org/abs/2303.10130", "type": "paper", "domains": [ "exposure" ], "role": "active_core", "status": "active", "used_for": [ "LLM exposure source" ], "source_keys": [ "eloundou_gpt_exposure_2023" ], "claim_ids": [ "reliability_weighted_exposure_ensemble" ], "summary": "Provides the GPT-oriented exposure source used as one leg of the live exposure ensemble.", "limitations": "Like other exposure indices, it measures capability overlap rather than realised displacement.", "repo_notes": "Kept in the live ensemble because it adds an LLM-specific construct not covered by AIOE alone." }, { "key": "pizzinelli_etal_2023", "title": "Labor Market Exposure to AI: Cross-country Differences and Distributional Implications", "authors": [ "Carolina Pizzinelli", "et al." ], "year": 2023, "published_at": "2023-10-04", "publisher": "IMF Working Paper", "url": "https://www.imf.org/en/Publications/WP/Issues/2023/10/04/Labor-Market-Exposure-to-AI-Cross-country-Differences-and-Distributional-Implications-539656", "type": "working_paper", "domains": [ "complementarity", "exposure" ], "role": "active_core", "status": "active", "used_for": [ "human bottleneck layer", "complementarity framing" ], "source_keys": [ "pizzinelli_theta_2023" ], "claim_ids": [ "deterministic_no_llm_core", "structural_pressure_not_prediction" ], "summary": "Provides the complementarity framework that the repo operationalises as the human bottleneck layer.", "limitations": "Designed as a broad labour-exposure framing rather than a Singapore occupation-level calibrated outcome model.", "repo_notes": "Directly tied to the theta-based bottleneck implementation in the current scorer." }, { "key": "felten_raj_seamans_2021", "title": "Occupational, industry, and geographic exposure to artificial intelligence: A novel dataset and its potential uses", "authors": [ "Edward Felten", "Manav Raj", "Robert Seamans" ], "year": 2021, "published_at": "2021", "publisher": "Strategic Management Journal", "url": "https://collaborate.princeton.edu/en/publications/occupational-industry-and-geographic-exposure-to-artificial-intel/", "type": "paper", "domains": [ "exposure" ], "role": "active_core", "status": "active", "used_for": [ "AIOE exposure source", "occupation-level exposure baseline" ], "source_keys": [ "aioe_2021" ], "claim_ids": [ "deterministic_no_llm_core", "reliability_weighted_exposure_ensemble", "structural_pressure_not_prediction" ], "summary": "Provides the published AIOE occupation exposure dataset used as a baseline source in the ensemble.", "limitations": "Measures theoretical exposure rather than realised AI use or job outcomes.", "repo_notes": "Tied directly to the live AIOE source key and still used in the canonical exposure ensemble." }, { "key": "felten_raj_seamans_2018", "title": "A Method to Link Advances in Artificial Intelligence to Occupational Abilities", "authors": [ "Edward Felten", "Manav Raj", "Robert Seamans" ], "year": 2018, "published_at": "2018", "publisher": "AEA Papers and Proceedings", "url": "https://www.aeaweb.org/articles?id=10.1257%2Fpandp.20181021", "doi": "10.1257/pandp.20181021", "type": "paper", "domains": [ "exposure", "tasks" ], "role": "active_core", "status": "active", "used_for": [ "task-to-occupation exposure framing", "methodology background" ], "source_keys": [], "claim_ids": [ "structural_pressure_not_prediction" ], "summary": "Introduces the task-ability linkage approach that underpins modern AI-exposure measurement.", "limitations": "Provides the conceptual bridge from AI capabilities to work content, but not current observed usage or Singapore-specific outcomes.", "repo_notes": "Referenced as the foundational exposure framework behind later AIOE-style occupation measures." } ] }