Pith. sign in

REVIEW 27 cited by

GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2303.10130 v5 pith:37EWBKXN submitted 2023-03-17 econ.GN cs.AIcs.CYq-fin.EC

GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models

classification econ.GN cs.AIcs.CYq-fin.EC
keywords llmsgptssoftwaretaskscapabilitiesllm-poweredmodelseconomic
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
read the original abstract

We investigate the potential implications of large language models (LLMs), such as Generative Pre-trained Transformers (GPTs), on the U.S. labor market, focusing on the increased capabilities arising from LLM-powered software compared to LLMs on their own. Using a new rubric, we assess occupations based on their alignment with LLM capabilities, integrating both human expertise and GPT-4 classifications. Our findings reveal that around 80% of the U.S. workforce could have at least 10% of their work tasks affected by the introduction of LLMs, while approximately 19% of workers may see at least 50% of their tasks impacted. We do not make predictions about the development or adoption timeline of such LLMs. The projected effects span all wage levels, with higher-income jobs potentially facing greater exposure to LLM capabilities and LLM-powered software. Significantly, these impacts are not restricted to industries with higher recent productivity growth. Our analysis suggests that, with access to an LLM, about 15% of all worker tasks in the US could be completed significantly faster at the same level of quality. When incorporating software and tooling built on top of LLMs, this share increases to between 47 and 56% of all tasks. This finding implies that LLM-powered software will have a substantial effect on scaling the economic impacts of the underlying models. We conclude that LLMs such as GPTs exhibit traits of general-purpose technologies, indicating that they could have considerable economic, social, and policy implications.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 27 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution

    cs.CL 2023-09 unverdicted novelty 8.0

    Promptbreeder evolves both task prompts and the mutation prompts that improve them using LLMs, outperforming Chain-of-Thought and Plan-and-Solve on arithmetic and commonsense reasoning benchmarks.

  2. JobBench: Aligning Agent Work With Human Will

    cs.AI 2026-05 unverdicted novelty 7.0

    JobBench is a new benchmark with 130 occupational tasks where the best of 36 tested AI models achieves only 45.9% success.

  3. Bounded by Risk, Not Capability: Quantifying AI Occupational Substitution Rates via a Tech-Risk Dual-Factor Model

    cs.CY 2026-04 conditional novelty 6.5

    AI occupational substitution is bounded more by commercial risk and liability than technical capability, producing high OAI for data scientists and near-zero for physical trades.

  4. Mixture of Cognitive Experts in Large Vision-Language Models

    cs.CV 2026-07 conditional novelty 6.0

    Routing CV experts into atomic evidence then Bloom-staged verbalization improves LVLM benchmarks and yields measurable query-conditioned reasoning traces.

  5. Design and Report Benchmarks for Knowledge Work

    cs.AI 2026-05 unverdicted novelty 6.0

    Proposes a three-step benchmark design method (define work activity, specify tested setting, score work product) derived from work studies and O*NET, demonstrated via three case analyses.

  6. Generative AI and the Productivity Divide: Human-AI Complementarities in Education

    cs.AI 2026-05 unverdicted novelty 6.0

    Randomized trial shows LLM assistance boosts average performance in technical self-study but increases inequality based on AI interaction skills, which scaffolding can reduce.

  7. Will AI Agents Free Us From Meaningless Work? A Human-Centered Analysis

    cs.CY 2026-05 unverdicted novelty 6.0

    Task-level survey data shows perceived bullshitness predicts desire for AI delegation and reduced oversight needs.

  8. LLMs Corrupt Your Documents When You Delegate

    cs.CL 2026-04 unverdicted novelty 6.0

    LLMs corrupt an average of 25% of document content during long delegated editing workflows across 52 domains, even frontier models, and agentic tools do not mitigate the issue.

  9. Bounded by Risk, Not Capability: Quantifying AI Occupational Substitution Rates via a Tech-Risk Dual-Factor Model

    cs.CY 2026-04 unverdicted novelty 6.0

    AI job substitution rates are limited by business risks such as liability and compliance rather than technical capability alone, resulting in high exposure for cognitive roles like data scientists and resilience for p...

  10. Training for Technology: Adoption and Productive Use of Generative AI in Legal Analysis

    cs.CY 2026-03 conditional novelty 6.0

    Brief training raised law students' optional LLM adoption from 26% to 41% and exam grade points by 0.27 relative to untrained access, while untrained access was unproductive.

  11. AI-Augmented Surveys: Leveraging Large Language Models and Surveys for Opinion Prediction

    cs.CL 2023-05 unverdicted novelty 6.0

    LLM embeddings enable strong retrodiction of masked GSS opinions via cross-validation and external validation but only modest performance on entirely unasked opinions.

  12. ChemCrow: Augmenting large-language models with chemistry tools

    physics.chem-ph 2023-04 conditional novelty 6.0

    ChemCrow augments LLMs with 18 expert chemistry tools to autonomously plan and execute syntheses and guide molecular discoveries in organic synthesis, drug discovery, and materials design.

  13. Characterizing initial human-AI proof formalization workflows

    cs.AI 2026-06 unverdicted novelty 5.0

    A controlled user study and qualitative survey find that AI assistance raises formalization accuracy for math proofs, with users flexibly combining multiple tools while retaining oversight.

  14. Who Prices Cognitive Labor in the Age of Agents? Compute-Anchored Wages

    cs.AI 2026-05 unverdicted novelty 5.0

    AI agents convert compute capital into cognitive labor units, so on substitutable tasks the competitive human wage is bounded above by relative productivity times compute intensity times the rental rate of compute.

  15. The AI Skills Shift: Mapping Skill Obsolescence, Emergence, and Transition Pathways in the LLM Era

    cs.CL 2026-04 unverdicted novelty 5.0

    Benchmarking four LLMs on O*NET skills yields SAFI scores showing mathematics and programming as most automatable while active listening and reading comprehension are least, with 78.7% of real AI interactions being au...

  16. The economic alignment problem of artificial intelligence

    econ.GN 2026-02 unverdicted novelty 5.0

    AI risks arise from growth-oriented economies, and post-growth concepts such as satisficing, the Doughnut model, and resource caps can reduce those risks while prioritizing tool-like AI over agentic systems.

  17. "If You're Very Clever, No One Knows You've Used It": The Social Dynamics of Developing Generative AI Literacy in the Workplace

    cs.HC 2026-02 accept novelty 5.0

    Hiding generative AI use to signal expertise reduces knowledge sharing and transparency among workplace colleagues.

  18. Inteligencia artificial y empleo en Espa\~na: una aproximaci\'on territorial y de g\'enero a la exposici\'on laboral

    cs.CY 2025-12 unverdicted novelty 5.0

    A sector-based AI exposure matrix applied to Spanish provincial employment data reveals stable higher exposure in metropolitan service regions and a consistent gender gap with female workers more exposed.

  19. Vibe Coding in Product Teams: Reconfiguring AI-Assisted Workflows, Prototyping, and Collaboration

    cs.HC 2025-09 accept novelty 5.0

    Interviews reveal a four-stage vibe coding workflow that accelerates prototyping while introducing tensions between quick efficiency and reflective design intention, plus asymmetries in trust and ownership.

  20. From Model Design to Organizational Design: Complexity Redistribution and Trade-Offs in Generative AI

    cs.CY 2025-06 unverdicted novelty 5.0

    LLMs relocate rather than eliminate trade-offs among generality, accuracy, and simplicity, shifting complexity to infrastructure, compliance, and expertise and redefining competitive advantage around managing that shift.

  21. GPT-4o System Card

    cs.CL 2024-10 unverdicted novelty 5.0

    GPT-4o is OpenAI's end-to-end multimodal model with human-like audio latency, improved non-English text performance, stronger vision and audio understanding, and accompanying safety evaluations.

  22. Who Benefits from AI? Self-Selection, Skill Gap, and the Hidden Costs of AI Feedback

    econ.GN 2024-09 unverdicted novelty 5.0

    Chess platform data shows self-selection by skilled users into AI feedback masks true effects, widens skill gaps, and causally reduces intellectual diversity via 42 natural experiments.

  23. StarCoder: may the source be with you!

    cs.CL 2023-05 accept novelty 5.0

    StarCoderBase matches or beats OpenAI's code-cushman-001 on multi-language code benchmarks; the Python-fine-tuned StarCoder reaches 40% pass@1 on HumanEval while retaining other-language performance.

  24. CodePori: Large-Scale System for Autonomous Software Development Using Multi-Agent Technology

    cs.SE 2024-02 unverdicted novelty 4.0

    CodePori is a multi-agent LLM system for code generation whose participant evaluation identifies practical challenges like memory limits and hallucinations missed by binary benchmarks.

  25. ASE-26: a curriculum for agentic software engineering as a discipline

    cs.CY 2026-05 unverdicted novelty 3.0

    ASE-26 is a proposed undergraduate curriculum for agentic software engineering organized around an evolutionary spiral of intent and build, with 21 modules and pedagogical commitments for agent-co-produced work.

  26. Who Prices Cognitive Labor in the Age of Agents? Compute-Anchored Wages

    cs.AI 2026-05 unverdicted novelty 3.0

    AI agents are a compute-to-cognitive-labor technology, bounding the competitive wage for substitutable human cognitive tasks by the rental rate of compute capital.

  27. Measuring the Occupation-Level Impact of AbbVie Intelligence: AI Applicability Analysis, 2024-2025

    cs.CY 2026-05 unverdicted novelty 2.0

    Empirical analysis of AbbVie internal AI usage finds +10.0% and +6.68% gains in occupation-level applicability scores after platform release and learning summit respectively.