Building a Flood Prediction Model provided lower secondary students with a gamified learning experience in artificial intelligence through the Market Marshes mission from the AI Quests series, developed in collaboration with Google Research and the Stanford Accelerator for Learning. In the role of researchers, students discovered that AI systems are not “black boxes,” but rather the result of a sequence of human decisions.
Through a simulation focused on solving a real social challenge — flood prediction — students explored the full AI lifecycle, from data collection and cleaning to training, testing, deployment, and the importance of human oversight. The workshop supported the development of digital competence CS3.4.12 in line with the DigComp 3.0 framework and helped participants understand the difference between predictive and generative AI models. Particular emphasis was placed on a human-centred approach and on human responsibility in final decision-making.