The Effects of Generative AI on High Skilled Work: Evidence from Three Field Experiments with Software Developers
Zheyuan Cui, Mert Demirer, Sonia Jaffe, Leon Musolff, Sida Peng, Tobias Salz · SSRN Electronic Journal · 1 Jan 2024
Read original sourceOpen access
Where this sits in the debate
Direction of the finding as read from the abstract by an AI model (AI-written).
Key takeaway AI-written
Reports on three separate field experiments (at Microsoft, Accenture, and a Fortune 100 firm) measuring how GitHub Copilot access affects software developers' output, finding sizable increases in completed tasks per week that are largest for less-experienced programmers. As one of the largest employer-scale randomized trials of a generative AI coding tool to date, it adds enterprise-level evidence to the debate over whether AI narrows or widens skill gaps among workers.
About this publication
Preprint. A preprint — posted publicly before or without journal peer review; findings may still be revised.
- Published
- 2024 (34 months ago)
- Citations
- 53 (OpenAlex)
53 citations recorded by OpenAlex — too few 2024 papers in this collection for a fair comparison.
□ added to this collection in 2026.
Leon Musolff, Mert Demirer, Sida Peng, Sonia Jaffe, Tobias Salz, Zheyuan Cui signed the wemustactnow.ai letter · author affiliation matches seed institutions massachusetts institute of technology, princeton university, university of pennsylvania — part of why this paper is in the collection.
Related by theme, not necessarily by method or conclusion.
- GPTs are GPTs: An Early Look at the Labor Market Impact Potential of Large Language Models
arXiv · 2023 - Experimental evidence on the productivity effects of generative artificial intelligence
Science · 2023 - The simple macroeconomics of AI
Economic Policy · 2024 - TaskNet: Building Occupational Task Inventories from Worker Interviews at Scale
Stanford Digital Economy Lab · 2026
Overview assembled automatically from collected metadata. How this site is built.