Elamur Research⌕
☰
home / papers / 2026 / Which Skills Debate Reaches the Public? Comparing Scientific Literature and Media Coverage of AI and LLM Skill Impacts (2022–2025)

Which Skills Debate Reaches the Public? Comparing Scientific Literature and Media Coverage of AI and LLM Skill Impacts (2022–2025)

Paper · 2026 · Maud Reveilhac, Gerold Schneider, Bohdan Trembovelskyi et al.

Summary

The paper examines how scientific findings on LLMs and human skills reach the Swiss public, and what changes along the way. The authors compare English-language literature reviews on LLM effects in education and the workplace with German- and French-language Swiss news coverage from Swissdox, both covering 2022–2025. Their reproducible pipeline combines BERTopic modeling with conceptual mapping. A translation-based pivot makes the multilingual media material comparable with the English reviews. To describe how scientific framings are recontextualized, the authors introduce a three-way scheme: themes can be amplified, simplified, or absent in the media. The two arenas diverge sharply. In the reviews, pedagogy and differentiated skill concepts dominate. The press favors a broad, product-centred AI-skills narrative focused on capabilities, job loss and replacement, while AI literacy, metacognition, reflective use and pedagogical integration receive little attention. The authors argue that skill change should be studied as a public-opinion issue, because narrow media framing can limit democratic and policy debate.

Key points

  • Corpora: 246 review documents (168 on education, 78 on work) and 4,610 Swissdox passages (2,915 German, 1,695 French). The passages were cut from 6,062 German and 3,478 French articles by splitting at the first full stop after every 200 tokens and keeping only query-matching passages.
  • Topic shares: One pedagogical topic covers most of the education-review corpus, and the dominant workplace-review topic holds 51.28%. In the media, a generic AI-skills topic takes 52.99%, while the education-centred topic reaches only 1.52%.
  • Temporal drift in media framing: Coverage moves from technological novelty (2022–2023) to use and application (2023–2024), then to labour-market and societal consequences (2024–2025).
  • ChatGPT as shorthand: The press often uses "ChatGPT" to stand for AI in general, which reinforces a product-centred, disruption-oriented framing.
  • Limitations: Corpus comparability and cross-lingual alignment are imperfect, the study has no direct public-opinion measures, and it does not apply dedicated time-aware methods (for instance, topic models that track change over time). The work is funded by TA-SWISS.

Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.

 Fields, topics and concepts

📚 Shelves 2The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 works

🧭 Topics 1Research subjects and languages the work is about.Language models & pretraining · 3,320 works

💡 Concepts 1Key ideas and methods explained on our concept pages.Topic modeling · 47 works

Cite

Cite as Maud Reveilhac, Gerold Schneider, Bohdan Trembovelskyi, Nisha Yadav, Vlada Druta, Aurelia Tamò-Larrieux, Simon Mayer, Clement Guitton and Joshua C. Yang. 2026. Which Skills Debate Reaches the Public? Comparing Scientific Literature and Media Coverage of AI and LLM Skill Impacts (2022–2025). Proceedings of the 11th Edition of the Swiss Text Analytics Conference.

The paper itself, not this page. BibTeX from the library's record of the work.

@inproceedings{reveilhac2026which,
  title = {{Which Skills Debate Reaches the Public? Comparing Scientific Literature and Media Coverage of AI and LLM Skill Impacts (2022–2025)}},
  author = {Maud Reveilhac, Gerold Schneider, Bohdan Trembovelskyi, Nisha Yadav, Vlada Druta, Aurelia Tamò-Larrieux, Simon Mayer, Clement Guitton and Joshua C. Yang},
  booktitle = {Proceedings of the 11th Edition of the Swiss Text Analytics Conference},
  year = {2026},
  url = {https://aclanthology.org/2026.swisstext-1.16/}
}

References

26 references, 4 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Daron Acemoglu and David Autor. 2011. Skills, tasks and technologies: Implications for employment and earnings. In David Card and Orley Ashenfelter, editors, Handbook of Labor Economics, volume 4, pages 1043–1171. Elsevier. find
  2. Jay Alammar and Maarten Grootendorst. 2024. Hands-On Large Language Models: Language Understanding and Generation. O’Reilly Media. find
  3. Dimo Angelov. 2020. Top2Vec: Distributed representations of topics. Computing Research Repository, arXiv:2008.09470.
  4. Svein Arnesen, Tanja S. Broderstad, James S. Fishkin, Mads P. Johannesson, and Alice Siu. 2025. Knowledge and support for AI in the public sector: A deliberative poll experiment. AI & Society, 40(5):3573– 3589. find
  5. Paul R. Brewer, John Bingaman, Albert Paintsil, David C. Wilson, and Will Dawson. 2022. Media use, interpersonal communication, and attitudes toward artificial intelligence. Science Communication, 44(5):632–658. find
  6. Erik Brynjolfsson, Danielle Li, and Lindsey Raymond. 2025. Generative AI at work. The Quarterly Journal of Economics, 140(2):889–942. find
  7. Tony Clear, Åsa Cajander, Alison Clear, Roger McDermott, Mats Daniels, Anna Fjelkner, Arto Hellas, Päivi Kinnunen, Juho Leinonen, James Prather, Arnold Pears, Emma Saarinen, and Claudia Szabo. 2025. AI integration in the IT professional workplace: A scoping review and interview study with implicatio… find
  8. Kibona Daniel, Mussa M. Msambwa, and Zhe Wen. 2025. Can generative AI revolutionise academic skills development in higher education? a systematic literature review. European Journal of Education. Advance online publication. find
  9. Michail Giannakos, Roger Azevedo, Peter Brusilovsky, Dragan Gašević, Inge Molenaar, Xavier Ochoa, and 1 others. 2025. The promise and challenges of generative AI in education. Behaviour & Information Technology, 44(11):2518–2544. find
  10. Pawel Gmyrek, Janine Berg, and Daniel Bescond. 2023. Generative AI and jobs: A global analysis of potential effects on job quantity and quality. Working Paper 96, International Labour Organization. find
  11. Maarten Grootendorst. 2022. BERTopic: Neural topic modeling with a class-based TF-IDF procedure. Computing Research Repository, arXiv:2203.05794.
  12. Mathieu Jacomy, Tommaso Venturini, Sebastien Heymann, and Mathieu Bastian. 2014. Forceatlas2, a continuous graph layout algorithm for handy network visualization designed for the gephi software. PLOS ONE, 9(6):e98679. find
  13. Enkelejda Kasneci, Katharina Sessler, Stefan Küchemann, Maria Bannert, Darya Dementieva, Frank Fischer, Ulrike Gasser, Georg Groh, Stephan Günnemann, Eyke Hüllermeier, Stephan Krusche, Gitta Kutyniok, Tomer Michaeli, Christian Nerdel, Friedhelm Pfeiffer, Oleksandra Poquet, Michael Sailer, Albrecht S… — ChatGPT for Good? On Opportunities and Challenges of Large Language Models for Education
  14. Philipp Lorenz-Spreen, Lisa Oswald, Stephan Lewandowsky, and Ralph Hertwig. 2023. A systematic review of worldwide causal and correlational evidence on digital media and democracy. Nature Human Behaviour, 7:74–101. find
  15. David McClure. 2016. textplot. https://github.com/davidmcclure/textplot. GitHub repository. ↗
  16. Dong Nguyen and Ekaterina Hekman. 2024. The news framing of artificial intelligence: a critical exploration of how media discourses make sense of automation. AI & Society, 39:437–451. find
  17. Nils Reimers and Iryna Gurevych. 2019. Sentence-BERT: Sentence embeddings using siamese BERT-networks. In Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing, pages 3982–3992. Association for…
  18. Nader Salari, Mojtaba Beiromvand, Ali Hosseinian-Far, Masoumeh Mohammadi, Bahareh Fatahi, and Siti Shohaimi. 2025. Impacts of generative artificial intelligence on the future of labor market: A systematic review. Smart Analytics, Artificial Intelligence and Sustainable Performance, 1(1):100067. find
  19. Sidra Sidra and Christine Mason. 2024. Reconceptualizing AI literacy: The importance of metacognitive thinking in an artificial intelligence-enabled workforce. In 2024 IEEE International Conference on Teaching, Assessment and Learning for Engineering (TALE). IEEE. find
  20. Brandon M. Stewart, Justin Grimmer, and Margaret E. Roberts. 2022. Text as Data: A New Framework for Machine Learning and the Social Sciences. Princeton University Press, Princeton, NJ. find
  21. Jörg Tiedemann, Mikko Aulamo, Daria Bakshandaeva, Michele Boggia, Stig-Arne Grönroos, Tommi Nieminen, Alessandro Raganato Yves Scherrer, Raul Vazquez, and Sami Virpioja. 2023. Democratizing neural machine translation with OPUS-MT. Language Resources and Evaluation, (58):713–755. find
  22. Jiaqi Wang and Wei Fan. 2025. The effect of Chat-GPT on students’ learning performance, learning perception, and higher-order thinking: Insights from a meta-analysis. Humanities and Social Sciences Communications, 12:744. find
  23. Marta Wieczorek, Mohammad Hosseini, and Bert Gordijn. 2025. Unpacking the ethics of using AI in primary and secondary education: A systematic literature review. AI and Ethics. Advance online publication. find
  24. Allison Woodruff, Renee Shelby, Patrick G. Kelley, Shiri Rousso-Schindler, Jeffrey Smith-Loud, and Lauren Wilcox. 2024. How knowledge workers think generative AI will (not) transform their industries. In Proceedings of the 2024 CHI Conference on Human Factors in Computing Systems, pages 1–26. Associ… find
  25. Lixiang Yan, Lei Sha, Ling Zhao, Yi Li, Roberto Martinez-Maldonado, Guanliang Chen, Xiaoming Li, Yiqi Jin, and Dragan Gašević. 2024. Practical and ethical challenges of large language models in education: A systematic scoping review. British Journal of Educational Technology, 55(1):90–112. find
  26. Shuo Zhang, P. Ganapathy Prasad, and Noah L. Schroeder. 2025. Learning about AI: A systematic review of reviews on AI literacy. Journal of Educational Computing Research. Advance online publication. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/reveilhac-2026-which-skills-debate-reaches-public/