What Can Machines Learn and What Does It Mean for Occupations and the Economy?
Open accessThe DOI matches its Crossref record; 3 of the 3 references it cites were found in a registry or an index.
Abstract
Advances in machine learning (ML) are poised to transform numerous occupations and industries. This raises the question of which tasks will be most affected by ML. We apply the rubric evaluating task potential for ML in Brynjolfsson and Mitchell (2017) to build measures of “Suitability for Machine Learning” (SML) and apply it to 18,156 tasks in O*NET. We find that (i) ML affects different occupations than earlier automation waves; (ii) most occupations include at least some SML tasks; (iii) few occupations are fully automatable using ML; and (iv) realizing the potential of ML usually requires redesign of job task content.
Overview
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
In the citing paper
“Most occupations in most industries have at least some tasks that could be replaced by AI, yet currently there is no occupation in which all tasks could be replaced88.”
From Understanding and shaping the future of work with self-determination theory (2022) · cites this paper as [88]
In the citing paper
“The variance explained by these earlier efforts (Acemoglu and Autor, 2011a;Frey and Osborne, 2017;Brynjolfsson et al., 2018;Felten et al., 2018;Webb, 2020;Brynjolfsson et al., 2023), along with wage controls, ranges from 60 to 72%, indicating that 28 to 40% of the variation in our AI exposure measure remains unaccounted for by previous technology exposure measurements.”
From GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models (2023) · cites this paper as Brynjolfsson et al., 2018
In the citing paper
“There is evidence that the use of LLM to perform job tasks can free up time for higher value activities (38–41).”
From Large language model exposure and precarious occupations: Unpacking relationships in the Canadian labor force (2026) · cites this paper as [40]
In the citing paper
“Machine learning models can flag anomalies in high-volume screening scans-tasks with well-defined inputs and verifiable outputsrepresenting automating exposure (Brynjolfsson et al., 2018).”
From AI, Output, and Employment (2026) · cites this paper as Brynjolfsson et al., 2018
Citations
Cites
The Skill Content of Recent Technological Change: An Empirical Exploration
Autor, D. H., Levy, F., & Murnane, R. J. (2003) · The Quarterly Journal of Economics · doi:10.1162/003355303322552801 · Cited by 6,205
The growth of low-skill service jobs and the polarization of the US labor market
Autor, D. H., & Dorn, D. (2013) · American Economic Review · doi:10.1257/aer.103.5.1553 · Cited by 3,824 · cited 2×
“2 Past automation using explicit rules or manually written computer algorithms to automate tasks has had a significant impact on productivity and the workforce ( Acemoglu and Autor 2011; Autor and Dorn 2013; Autor, Levy, and Murnane 2003 .”
MethodWhat can machine learning do? Workforce implications
Brynjolfsson, E., & Mitchell, T. (2017) · Science · doi:10.1126/science.aap8062 · Cited by 1,131 · cited 9×
“Furthermore, tasks within jobs typically show considerable variability in SML, while few ( if any jobs can be fully automated using ML. Machine learning technology can transform many jobs in the economy, but full automation will be less significant than the reengineering of processes and the reorganization of tasks.”
BackgroundMethodInfluentialSelf-citation
Research works printed but not linked: 1
- Autor David H, 2014, Reevaluating Labor Market Dynamics, Federal Reserve Bank of St. Louis Economic Policy Proceedings, 129
Cited by
Items 801–844 of 844
The Two Axes That Sort AI Use: Task-Level Evidence from Platform Usage LogsPreprint
verschuere, B., & Cameron, A. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7408558
Beyond AI Integration: Reconstructing Higher Education for the Age of Artificial Intelligence
Khawaldeh, J. (2026) · PhilPapers (PhilPapers Foundation) · doi:10.5281/zenodo.22693389
When Everyone Has an Excavator: How AI Changes Comparative Advantage in Accounting ResearchPreprint
Iqbal, A. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7447659
ORQA: An Occupation-Realistic Question and Answer Framework for LLM Professional KnowledgePreprint
Krishnan, S., Chang, S. Y., & Nagaraj, A. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2609.12366
“Existing economic work measures exposure, adoption, or productivity effects of AI in the labor market [Brynjolfsson et al., 2018, Eloundou et al., 2023, Handa et al., 2025, Chatterji et al., 2025, Brynjolfsson et al., 2025, Dell’Acqua et al., 2026].”
West, J., & Cosimano, M. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7462099
Determinants of AI acceptance in the workplace: Evidence from eleven European countries
Stefkovics, Á., & Orbán, F. (2026) · Computers in Human Behavior Reports · doi:10.1016/j.chbr.2026.101313
Precarious Augmentation in the Age of AI: Task Bottlenecks and Organizational TippingPreprint
Kendall, J. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7464582
Artificial intelligence, global value chains, and biased technological progress
Zhao, J., 王明琳, Shu, X., & Wang, Y. (2026) · International Review of Economics & Finance · doi:10.1016/j.iref.2026.105826
Frontier technology adopters and the aggregate decline of routine jobs
Arntz, M., Genz, S., Gregory, T., Lehmer, F., & Zierahn, U. (2026) · Journal of Economic Behavior & Organization · doi:10.1016/j.jebo.2026.107732 · Cited by 1
How Does Intelligent Manufacturing Improve Labour Income Share: A Quasi-Natural Experiment in China
Wang, Q., Ren, H., & Hu, N. (2026) · SAGE Open · doi:10.1177/21582440261491494
Perceptions of artificial intelligence bias and gender inclusivity: toward a gendered divergence
Au, A. (2026) · Information Communication & Society · doi:10.1080/1369118x.2026.2737402
Volarević, M., Krasić, S. O., & Vučković, M. (2026) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-9190-8.ch006
Yapay Zekânın İşgücü Verimliliği ve İstihdam Üzerindeki Etkileri
CİNEL, E. A. (2026) · Politik Ekonomik Kuram · doi:10.30586/pek.1949021
Ruilova, J. J. C., & Matute-Petroche, J. (2025) · Sinergia Académica · doi:10.51736/sa826
Maulana, J., Prasetiyo, J., & Mulyadi, M. (2026) · Jurnal Manajemen & Pendidikan [JUMANDIK] · doi:10.58174/resv5g27
Liu, J. (2026) · Journal of risk and financial management · doi:10.3390/jrfm19100755
Robasto, D., & Natali, S. (2024) · Institutional Research Information System University of Turin (University of Turin)
Grove, W. J. (2022) · UWC Research Repository (University of the Western Cape)
Hew, T. H., Rasiah, R., & Xie, N. (2026) · Frontiers in Public Health · doi:10.3389/fpubh.2026.1929386
Évolutions de l’intelligence artificielle au travail et collaborations humain-machine
كوتيه, آ.-م., سو, ز., خنور, ص., Côté, A.-M., Zan, S. U., & Khennour, S. (2025) · سلسلة أروقة العلوم
How Much Work Can AI Agents Do? Task Value Inside the AI Time-Horizon FrontierPreprint
Harris, T. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7570818
Fowosere, O. (2026) · International Journal of Research and Innovation in Social Science · doi:10.47772/ijriss.2026.100900260
Complementary Software and Occupational GenAI ExposurePreprint
Arslan, B. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7571619
Artificial Intelligence, Worker Reallocation, and Occupational NetworksPreprint
Johansson, S. F. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7616188
Analyzing the Effects of GDP per Working-Age Capita on AI Usage
Khandelwal, D. (2026) · Western Undergraduate Economics Review
Western Undergraduate Economics Review Annual 2026
(2026) · Western Undergraduate Economics Review
Gupta, R., & Deshpande, M. (2026) · Acta Innovations · doi:10.62441/actainnovations.vi.1999
Ekuma, K. J., & Uba, C. D. (2026) · Cogent Education · doi:10.1080/2331186x.2026.2745299
Artificial Intelligence Exposure, Tacit Experience and Occupational Wage GrowthPreprint
尚泽忠 (2026) · Research Square · doi:10.21203/rs.3.rs-11067023/v1
A Research Agenda for the Economics of Transformative AI
Brynjolfsson, E., Korinek, A., & Agrawal, A. (2026) · Communications of the ACM · doi:10.1145/3769433
“11,12,15 While prior work used the U.S. Department of Labor's O*Net taxonomy of about 18,000 tasks, future work could apply natural language processing to classify hundreds of millions of job postings and résumés, creating a much more fine-grained and dynamic task taxonomy.”
Self-citationThe Delegation Divide: An Open Index of AI-Agent Delegability across Occupations and CountriesPreprint
Maya, C. E., & Londoño, S. J. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.30752
“Webb (2020) measures overlap between patent texts and task descriptions, and Brynjolfsson et al. (2018) score each task’s suitability for machine learning.”
TaskNet: Building Occupational Task Inventories from Worker Interviews at ScalePreprint
Wang, J., Haupt, A. A., Robinson, I. A., Lewis, P., Brynjolfsson, E., & Yang, D. (2026) · SSRN Electronic Journal · doi:10.2139/ssrn.7670419
Self-citationHalder, S. R., Mukherjee, A., Tripathi, A., Bansal, M., Sridevi, R., & Seniasamy, R. (2026) · Research Square · doi:10.21203/rs.3.rs-11152053/v1
Papu, S. M. (2026) · Nelson Mandela University Research Repository
INTEGRATION OF GENERATIVE AI SKILLS IN JOB REQUIREMENTS: EVIDENCE FROM JOB ADVERTISEMENTS IN GEORGIA
Khukhunaishvili, N., & Zhghenti, T. (2026) · Baltic Journal of Economic Studies · doi:10.30525/2256-0742/2026-12-5-47-57
Harder Than the Task: From Technical to Organizational Difficulty in AI BenchmarkingPreprint
Barres, V., Barton, A., Foss, N. J., & Nielsen, H. M. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2611.03494
Measuring the prevalence and direction of innovation: a pilot approach using LLMs
Garg, P., Lambert, P. J., Martin, R., & Medvedev, D. (2026) · London School of Economics and Political Science Research Online (London School of Economics and Political Science) · doi:10.21953/researchonline.lse.ac.uk.00141830
TaskNet: Building Occupational Task Inventories from Worker Interviews at Scale
Wang, J., Haupt, A., Robinson, I. A., Lewis, P., Brynjolfsson, E., & Yang, D. (2026) · National Bureau of Economic Research · doi:10.3386/w35987 · Cited by 1
Self-citationAge, Education, and Occupational Gradients in Artificial Intelligence Anxiety
Zimmermann, K. F., Karabulut, G., Keklik, Z., & Altıntaş, K. (2026) · doi:10.53330/ggjn5216
Fox, K. D., & Deaton, J. E. (2026) · Research Square · doi:10.21203/rs.3.rs-11653360/v1
Metadata
- Authors
- Brynjolfsson, E., Mitchell, T., & Rock, D.
- Affiliations
- Massachusetts Institute of Technology (US)
- Carnegie Mellon University (US)
- Venue
- AEA Papers and Proceedings
- Publisher
- American Economic Association
- Published
- 2018-05-01
- Volume
- 108
- Pages
- 43-47
- ISSN
- 2574-0768 (ISSN-L)
- 2574-0776
- OpenAlex
- W2804936679
- Cited by
- 695
- References
- 2
- Citations by year
- 2026: 145 · 2025: 131 · 2024: 110 · 2023: 81 · 2022: 73 · 2021: 61 · 2020: 46 · 2019: 35 · 2018: 11
- FWCI
- 58.15
- Type
- journal article
- Language
- en
- Open access
- green
- orcid.org
- Licence
- cc-by-nc
- Locations
- AEA Papers and Proceedings (publishedVersion)
- RePEc: Research Papers in Economics (submittedVersion)
- DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc)
- Topics
- Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Labor market dynamics and wage inequality
- Social Sciences › Sociology and Political Science › Digital Economy and Work Transformation
- Social Sciences › Decision Sciences › Management Science and Operations Research › Impact of AI and Big Data on Business and Society
- Keywords
- machine learning, automation, occupation, O*NET, work tasks, labor market impact, job redesign
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.