AI Quests

From climate change to food security, AI has the potential to address some of humanity’s biggest challenges. But, for this to happen, we must inspire a new generation of AI changemakers and prepare the middle schoolers of today to change lives through responsible AI deployment. Fusing cutting-edge Google Research projects with pedagogical expertise from Stanford University, AI Quests is a fully immersive, interactive learning experience that shows students the wonder of what’s possible.

Research shows students’ performance can improve 50% when a subject is gamified. Spurred on by this insight, we collaborated extremely closely with Google Research and the Stanford team to develop an endlessly exciting world of playable quests, each inviting students to follow in the footsteps of real Google researchers by exploring AI technology and learning to deploy it responsibly.

Every aspect of our simulated planet was designed to let students see the impact of their actions. From the rich variety of environments to the weird and wonderful creatures who inhabit them, we wanted students to experience firsthand how AI can augment human decisions, as well as understanding the importance of ethical, informed use.

Each mission is introduced by a primary character, with their own deeply developed backstory, regional history and lore. In Market Marshes, we meet Luna, a part-amphibian part-moss-like creature born from her marshland environment. Working with Stanford, we created Luna and her story specifically to encourage empathy from the students, which in turn helps them to build a genuine sense of mission and desire to help. This emotional involvement leads to more engaged learning, a better understanding of the subject, and real inspiration to consider how else AI might be able to solve real-world problems.

We used Google AI throughout the development process, from environment and character creation to voiceovers. The environments combine Gemini-generated imagery with a layer of custom 3D assets, modelled and optimised before being exported to WebGL. By instancing key objects across the scenes, we kept the performance high and the file size low. Each model was brought to life with our custom toon shader in WebGL pulling colour directly from the vertex and colour attribute data. An additional dimension that was crucial to building an immersive experience was sound. We carefully crafted and layered both ambient sound and soundtracks to bring our environments to life; adding spot sounds to offer extra guidance and feedback.

By bringing together Google’s deep technical expertise and Stanford’s knowledge of learning processes, we’ve been able to create something truly original: a platform that democratises innovation. A place that doesn’t just educate students on the uses of AI, but gives them space to imagine new frontiers of advancement. Frontiers they’ll one day, hopefully, be a part of.