Go to SwissText 2026 Conference homepageWhich Skills Debate Reaches the Public? Comparing Scientific Literature and Media Coverage of AI and LLM Skill Impacts (2022–2025)
Track: Scientific Track
Keywords: large language models, skill transformation, media framing, education and work, BERTopic, conceptual maps
TL;DR: Across 4,600+ media texts and 246 reviews, we show that nuanced skill concepts dominate science but are largely absent in media, which focuses on AI capability and job risk.
Abstract: As large language models (LLMs) rapidly enter classrooms and workplaces, the public discourse about their impact on human skills has significant consequences for public understanding, policy, and education. We analyze this discourse through a comparison of the scientific literature and Swiss multilingual media coverage in German and French on how LLMs affect skills in education and work from 2022 to 2025. Using a reproducible pipeline that combines conceptual mapping and BERTopic modeling and proposing and implementing a tripartite analytical framework– amplified, simplified, absent– for examining how scientific framings are recontextualized in public discourse, we find a sharp divergence in thematic structure. We use these three categories to show that media coverage amplifies concrete use cases, simplifies differentiated scientific debates about skills, and leaves several central scientific themes less visible. In education reviews, a single pedagogical topic accounts for the majority of the corpus; in workplace reviews, it remains dominant at 51.28%. In Swiss media, however, the dominant topic is a broad and generalized AI-skills discourse (52.99%), while the education-centered topic accounts for only 1.52%. Conceptual maps show in greater detail that media coverage foregrounds AI capabilities, job loss, and replacement, while giving limited attention to themes central in the literature, including AI literacy, reflective use, metacognition, and pedagogical integration.
Submission Number: 44
Loading