AI through a social-science lens: Rethinking learning, teaching, and research [Keynote address]

Vazquez-Calvo, B. (2025, July 4). AI through a social-science lens: Rethinking learning, teaching, and research[Keynote address]. Eighteenth Global Studies Conference & Twentieth International Conference on Interdisciplinary Social Sciences, University of Málaga, Málaga, Spain. https://doi.org/10.5281/zenodo.15853715 AI Through a Social-Science Lens: Rethinking Learning, Teaching, and Research Boris Vazquez-Calvo [email protected] University of Seville Artificial Intelligence (AI) tools have slipped quietly from research labs into everyday classrooms: they now rewrite the texts our students read, grade essays in seconds, and forecast when attention will drift. Faced with systems this powerful—and this fast-moving—social-science educators, learners, and researchers need more than curiosity; they need a clear interpretive map. This keynote provides that map by looking at AI through a three-part lens—learning, teaching, and researching—and by turning headline statistics and empirical studies into plain-language insights that matter on the ground. For example, in a survey of 1,611 Japanese university students of Spanish, 67 % had used generative-AI tools in the past month and 60 % used them weekly or daily for comprehension, translation, and writing support (Aroz et al., 2025)—evidence that AI has become routine, not experimental, indicating paths for growth and teaching intervention. Additionally, a meta-analysis of 50 intelligent-tutoring-system trials showed an average gain of 0.66 SD (Kulik & Fletcher, 2016)—roughly the jump from the 50th to the 75th percentile. However, the same review shows much smaller gains (≈ 0.13 SD) on independent, standardized tests; the big effects (≈ 0.73 SD) appear when the test is written by the designers themselves, meaning that AI can be powerful, yet results shrink when you step outside the tutor’s homework sheet. The COST (Content–Others–Self–Tasks) (Zhu et al., 2023) framework and a socio-cognitive approach to attitude (Hallajow, 2018) help pinpoint where and what is really improving learning-wise. A 2024 survey-model of 393 Malaysian university lecturers revealed that the more anxious teachers felt about AI, the less willing they were to experiment with it—each one-point rise on the anxiety scale cut “adoption readiness” by almost half a point. Crucially, readiness itself proved the major swing factor: it accounted for 67 % of the overall change in teachers’ positive attitude toward AI-based assessment (Shahid et al., 2024). Put simply, boosting teachers’ practical confidence does far more to win acceptance than any top-down policy circular or shiny new gadget, whereas unchecked anxiety stalls innovation before it starts. A qualitative discourse analysis with 17 teachers worldwide revealed common worries about plagiarism yet a cautious willingness to experiment (Paz-López et al., under review). Together these findings quantify—and humanize—the pedagogical lag between fast-moving student adoption and hesitant teacher uptake. A review of 146 AI-in-higher-education papers applauded the field’s technical ingenuity—better prediction models, smarter chatbots—but also warned that most studies stop at the algorithm and rarely ask “so what for teaching?” or “who benefits?” (Zawacki-Richter et al., 2019). In other words, research output is booming, yet its links to real pedagogy and social justice remain thin. The equity gap widens when we look at language coverage: fewer than 2 % of the planet’s 7,000-plus languages appear in Natural Language Processing (NLP) corpora (Joshi et al., 2020), so today’s AI “polyglot” still ignores 98 % of the world’s linguistic diversity. UNESCO’s Recommendation on the Ethics of Artificial Intelligence(2021, p. 10-13) responds by setting minimum guardrails— transparency, human oversight, bias and stereotyping checks, environmental duty of care—while recent AI-literacy frameworks distil four actionable competencies: know what AI is, use it wisely, evaluate its outputs, and act ethically (Ng et al., 2021). Taken together, the message is clear: impressive code is not enough; without pedagogy and inclusion, AI risks reinforcing the very inequities it promises to solve. The evidence sends a clear but urgent signal: AI capability is sprinting; AI literacy is jogging. Unless educators and researchers learn to explain these systems, critique their blind spots, and co-design them for equity, the technology’s promise will harden into new barriers. The keynote closes with three immediately actionable steps—teacher “sandbox” workshops to melt anxiety, multilingual data drives to shrink the 2 % language gap, and ethics-by-design checklists for every pilot research or classroom project Together, these steps form a practical roadmap for making minds and machines thrive. References Aroz, A., Hirose, H., Nishimura, K., & Cassany, D. (2025). Inteligencia artificial generativa para aprender español: Prácticas y valoraciones. Cuadernos CANELA, 36, 1–22. Hallajow, N. (2018). Identity and Attitude: Eternal Conflict or Harmonious Coexistence. Journal of Social Sciences, 14(1), 43–54. https://doi.org/10.3844/jssp.2018.43.54 Joshi, P., Santy, S., Budhiraja, A., Bali, K., & Choudhury, M. (2020). The state and fate of linguistic diversity and inclusion in the NLP world. In Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics (pp. 6282–6293). Association for Computational Linguistics. https://doi.org/10.18653/v1/2020.acl-main.560 Kulik, J. A., & Fletcher, J. D. (2016). Effectiveness of intelligent tutoring systems: A meta-analytic review. Review of Educational Research, 86(1), 42–78. https://doi.org/10.3102/0034654315581420 Ng, D. T. K., Chai, C. S., & Low, E. L. (2021). Conceptualising AI literacy: An exploratory review. Computers & Education: Artificial Intelligence, 2, Article 100041. https://doi.org/10.1016/j.caeai.2021.100041 Paz-López, A., Shafirova, L., & Vazquez-Calvo, B. (under review). A discourse analysis of language teachers’ attitudes toward Artificial Intelligence. Shahid, M. K., Zia, T., Bangfan, L., Iqbal, Z., & Ahmad, F. (2024). Exploring the relationship of psychological factors and adoption readiness in determining university teachers’ attitude on AI-based assessment systems. The International Journal of Management Education, 22(2), 100967. https://doi.org/10.1016/j.ijme.2024.100967 UNESCO. (2021). Recommendation on the Ethics of Artificial Intelligence. https://unesdoc.unesco.org/ark:/48223/pf0000380455 Zawacki-Richter, O., Marín, V. I., Bond, M., & Gouverneur, F. (2019). Systematic review of research on artificial-intelligence applications in higher education. International Journal of Educational Technology in Higher Education, 16, Article 39. https://doi.org/10.1186/s41239-019-0171-0 Zhu, C., Meng, S., Luo, J., Li, T., & Wang, M. (2023). How to harness the potential of ChatGPT in education? Knowledge Management & E-Learning: An International Journal, 15(2), 133–152. https://doi.org/10.34105/j.kmel.2023.15.008