Google turns real AI research into a classroom game

Google Research and the Stanford Accelerator for Learning just launched AI Quests, an immersive, game-based learning platform designed to teach AI literacy to middle schoolers.

Targeting students aged 11 to 14, the interactive quests put kids in the shoes of Google scientists navigating fantasy worlds to tackle real-world problems. The project debuted today and is freely available to educators worldwide.

Solving real-world crises in a fantasy game

Guided by an in-game mentor named Professor Skye, students jump straight into the fundamentals of machine learning without having to write code.

The first playable quest is based directly on Google Research's real-world Flood Forecasting initiative. Students must gather rainfall and river flow metrics, evaluate data quality, and train and test a model to help villagers predict upcoming floods.

Along the way, kids learn firsthand how data quality and human judgment shape what an AI outputs, getting a front-row seat to the concept of garbage in, garbage out.

At the end of the run, students unlock a video message from the actual Google engineers behind the flood tech, explaining how the science works in the field.

Eye exams, brain maps, and ready-made lesson plans

Google confirmed that two more quests are slated to roll out in the coming months.

One upcoming module will focus on diabetic retinopathy, a condition that can cause blindness, showing how AI assists in medical screenings. The second pulls from Google’s connectomics work, guiding students through mapping and understanding the wiring of the human brain.

To make classroom adoption as painless as possible, each module comes with full lesson plans, teacher guides, and pre- and post-game classroom activities.

Global education initiatives are already picking it up. Experience AI, a program co-developed by Google DeepMind and the Raspberry Pi Foundation, is embedding AI Quests into its worldwide curriculum, alongside Stanford’s CRAFT platform.

Why it matters

Most current efforts around student AI education tend to swing between two extremes: overly technical Python tutorials or superficial lessons on how to prompt a chatbot to write a history essay. Google and Stanford took a far more practical middle path.

For teachers and young learners, AI Quests addresses the real core of AI literacy. Students do not need to memorize neural network math to understand that biased or sloppy data produces bad decisions. Delivering these lessons inside a gamified quest makes the abstract concept of model training tangible for a middle schooler, while ready-to-use teacher guides lower the barrier for non-technical educators.

On a strategic level, this is classic long-term brand building by Google. At a time when generative AI faces fierce scrutiny over copyright, plagiarism, and energy use, Google is deliberately framing AI around climate resilience, healthcare, and neuroscience. By giving schools turnkey tools centered on these humanitarian applications, Google is actively shaping how the next generation perceives the role of artificial intelligence in society.