# AI+Education Summit 2026: AI Quests – When Learning Sciences Meet Product Design toward AI Literacy Source: https://www.youtube.com/watch?v=qOZ16hAtnuI Recap page: https://rapidrecap.app/video/qOZ16hAtnuI Generated: 2026-02-19T18:42:30.811+00:00 --- ## Quick Overview The AI+Education Summit 2026 panel, featuring Alon Harris, Victor Lee, and moderator Ronit Levavi Morad, detailed the development of "AI Quests," an educational initiative designed to teach AI literacy by translating complex research into age-appropriate, gamified lesson plans that focus on human agency and ethical considerations. **Key Points:** - The "AI Quests" initiative translates complex AI research into gamified, age-appropriate lesson plans to foster AI literacy in teens. - The panel introduced three main quests: "Market Marshes" (Flood Forecasting), "Dusky Dunes" (Blindness Prevention), and "Polar Peaks" (Mapping the Brain). - The pedagogical approach emphasizes human agency, requiring students to make decisions on data cleaning, model training, and understanding ethical implications. - The Flood Forecasting Quest ("Market Marshes") involves teaching students about AI models that predict river flooding, explicitly showing the steps from understanding the problem to assessing flood safety. - The Blindness Prevention Quest involves students using AI-powered eye scans to diagnose diabetic retinopathy, mirroring real-world tools and demands. - A key pedagogical principle is balancing technical understanding with human context, ensuring students understand the consequences of AI actions (e.g., poor data leading to inaccurate forecasts). - The initiative is a collaboration between Google Research and the Stanford Accelerator for Learning (SAL). ![Screenshot at 00:05: The title slide for the panel, "AI Quests: When Learning Sciences Meet Product Design Toward AI Literacy," immediately frames the discussion around merging educational theory with practical AI learning tools.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-00-05.jpg) **Context:** This video captures a panel discussion from the 2026 AI+Education Summit, focusing on the "AI Quests" project, a collaboration between Google Research and the Stanford Accelerator for Learning (SAL). The panel, moderated by Ronit Levavi Morad, included Alon Harris (Google Research) and Victor Lee (Stanford/SAL), who discussed how they bridge the gap between cutting-edge AI research and K-12 education by creating interactive, gamified learning experiences aimed at building AI literacy. ## Detailed Analysis The panel introduced "AI Quests," an initiative built on the premise that humans can initiate and design AI applications to solve major global challenges. The core goal is to move students from passive consumption to active participation in AI, emphasizing enduring understanding, ethics, and human agency. The project is structured around three specific quests mirroring real-world research domains: Flood Forecasting (Market Marshes), Blindness Prevention, and Mapping the Brain (Polar Peaks). Alon Harris highlighted the importance of avoiding the mistake of focusing only on technical skills, stressing that students must understand the societal impact and ethical considerations, such as data bias. Victor Lee emphasized the iterative design process, where students learn by doing—collecting data, cleaning it, training models, and testing outcomes—and seeing the consequences of their decisions. The team consciously designed the experience to ensure students understand the limitations (like relying on potentially inaccurate forecasts) and the importance of human oversight, contrasting the game experience with the real-world rigor required in both science and product development. The panel concluded by thanking the audience and providing the URL for access: research.google/ai-quests. ### Panel Introduction - Ronit Levavi Morad introduces Alon Harris and Victor Lee, emphasizing the cross-disciplinary synergy between Google Research and Stanford Accelerator for Learning to promote AI literacy. ### The AI Quests Concept - The project aims to translate real-world AI research into engaging, gamified experiences, moving students from passive users to active critical thinkers regarding AI's role in addressing challenges like flood forecasting and disease diagnosis. ### Flood Forecasting Quest ('Market Marshes') - This quest models the real-world effort of using AI to predict floods, requiring students to complete tasks like data collection, cleaning, training a model, and testing accuracy against historical data. ### Blindness Prevention Quest - This quest mirrors the real-world challenge of diagnosing diabetic retinopathy using AI analysis of eye scans, demanding students understand data fidelity and accuracy criteria. ### Mapping the Brain Quest ('Polar Peaks') - This quest focuses on neuroscience, challenging students to understand complex biological systems and how AI models are used to map neurons and synapses, highlighting the need for expertise beyond basic coding. ### Pedagogical Approach - Key learning principles include embedded feedback/productive failure, leveraging self-explanation (using 'Learning Tickets'), situating learning in real-world contexts, and guiding students with pedagogical agents (like Professor Skye and Luna). ![Screenshot at 00:05: Title slide for the 2026 AI+Education Summit panel discussing AI Quests.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-00-05.jpg) ![Screenshot at 01:04: Professor Skye, an AI expert/mentor, welcomes the user to the 'AI Quests Onboarding' experience, outlining the mission to help locals with real-world issues like flood prediction.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-01-04.jpg) ![Screenshot at 01:30: Gameplay screenshot showing the 'Select data chips' interface, demonstrating the selection of 'Rainfall' and 'River Flow' data to train an AI model.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-01-30.jpg) ![Screenshot at 10:41: Slide summarizing the three main research challenges addressed by AI Quests: Flood Forecasting, Blindness Prevention, and Mapping the Brain, along with their respective quests.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-10-41.jpg) ![Screenshot at 21:31: Slide detailing the four core pedagogical approaches: Embedded feedback/productive failure, Leveraging self-explanation, Guiding with Pedagogical Agents, and Situated learning.](https://ss.rapidrecap.app/screens/qOZ16hAtnuI/00-21-31.jpg)