Dutch students show growing enthusiasm for generative AI in education (phys.org)

Researchers at Utrecht University surveyed 410 Dutch computing students across 23 schools — from high schools to universities of applied sciences and research universities — and found broad, growing enthusiasm for generative AI in learning. Sixty percent said they use generative AI at least weekly; high school pupils were the most positive (≈70%), students at universities of applied sciences >57%, and university students just over 50%. Use cases vary by cohort: high schoolers mainly use AI for text tasks (translation, background research), while post‑secondary students lean on models for programming tasks such as writing code and debugging. The study is reported in the Proceedings of the 25th Koli Calling International Conference on Computing Education Research. The findings matter for the AI/ML and education communities because they show rapid, practical adoption of code‑oriented LLM tools and reveal divergent attitudes: many students welcome the productivity gains, but roughly half of university students worry generative AI can undermine deep learning by short‑circuiting problem solving. Key implications include the need to incorporate AI literacy into curricula (how models work, limitations, and risks), rethink assessment and pedagogy to preserve skill acquisition, and study long‑term impacts on learning. The authors plan follow‑ups to track evolving use patterns and whether students retain “learning by doing” versus becoming tool‑dependent.