This vision paper argues that AI should be taught through authentic scientific inquiry rather than as a standalone topic. It frames computer vision, clustering, and generative modeling as pedagogically bounded scientific instruments embedded in observing, analyzing, and modeling. Each instrument should preserve its core scientific function while exposing students to a specific point of critical reflection about how AI can help or mislead inquiry. The authors also discuss agentic AI spanning the full inquiry process, but argue that students should first develop foundational understanding of scientific practices and AI instruments.
No heat snapshots are available in the last 24 hours.