Large language models reduce public knowledge sharing on online Q&A platforms
PNAS Nexus · 02.09.2024 · 3 (9)
DOI: 10.1093/pnasnexus/pgae400 ↗Kurz gefasst
Fachartikel in der Zeitschrift PNAS Nexus, online erschienen am 11.09.2024. Verfasst von R Maria Del Rio-Chanona und 2 weiteren Autor:innen. Beteiligte Einrichtung: Department of Computer Science, University College London, London WC1E 6EA, United Kingdom. Die Arbeit wurde bisher 52-mal zitiert (Crossref) und nennt 69 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Hungarian National Scientific Fund.
Abstract (Originalsprache)
Large language models (LLMs) are a potential substitute for human-generated data and knowledge resources. This substitution, however, can present a significant problem for the training data needed to develop future models if it leads to a reduction of human-generated content. In this work, we document a reduction in activity on Stack Overflow coinciding with the release of ChatGPT, a popular LLM. To test whether this reduction in activity is specific to the introduction of this LLM, we use counterfactuals involving similar human-generated knowledge resources that should not be affected by the introduction of ChatGPT to such extent. Within 6 months of ChatGPT’s release, activity on Stack Overflow decreased by 25% relative to its Russian and Chinese counterparts, where access to ChatGPT is limited, and to similar forums for mathematics, where ChatGPT is less capable. We interpret this estimate as a lower bound of the true impact of ChatGPT on Stack Overflow. The decline is larger for posts related to the most widely used programming languages. We find no significant change in post quality, measured by peer feedback, and observe similar decreases in content creation by more and less experienced users alike. Thus, LLMs are not only displacing duplicate, low-quality, or beginner-level content. Our findings suggest that the rapid adoption of LLMs reduces the production of public data needed to train them, with significant consequences.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Computer Science, University College London, London WC1E 6EA, United Kingdom.
- Bennett Institute for Public Policy, University of Cambridge, Cambridge CB3 9DT, United Kingdom.
- Complexity Science Hub Vienna, Vienna 1080, Austria.
- Faculty of Economics, LMU Munich, Munich 80539, Germany.
- Sanofi Hungary, Budapest 1138, Hungary.
- Department of Network Science, Corvinus University of Budapest, Budapest 1093, Hungary.
ORCID-Profile: R Maria Del Rio-Chanona, Nadzeya Laurentsyeva, Johannes Wachs
Förderung
- Hungarian National Scientific Fund
Zitieren
APA
Del Rio-Chanona, R. M., Laurentsyeva, N., & Wachs, J. (2024). Large language models reduce public knowledge sharing on online Q&A platforms. PNAS Nexus, 3(9). https://doi.org/10.1093/pnasnexus/pgae400
Vancouver
Del Rio-Chanona RM, Laurentsyeva N, Wachs J. Large language models reduce public knowledge sharing on online Q&A platforms. PNAS Nexus. 2024;3(9). doi:10.1093/pnasnexus/pgae400
BibTeX
@article{DelRioChanona2024,
author = {Del Rio-Chanona, R Maria and Laurentsyeva, Nadzeya and Wachs, Johannes},
title = {Large language models reduce public knowledge sharing on online Q&A platforms},
journal = {PNAS Nexus},
year = {2024},
volume = {3},
number = {9},
doi = {10.1093/pnasnexus/pgae400}
}
EUVIMED