Google adds Polar Peaks neuroscience lesson to AI Quests

Google Research and the Stanford Accelerator for Learning launched Polar Peaks on October 26, adding neuroscience to the AI Quests series. The activities invite students aged 11 to 14 to take the role of researchers and explore how AI is developed. Google says more than 400,000 quests have been completed across the series, which has been localized into ten languages; that figure is a count of completed activities, not individual students.

In the new quest, in-game mentors guide students through scanning and mapping neurons to investigate a character’s memory loss. Tasks include cleaning data and testing a model, making human inspection part of the activity. The scientific setting gives students concrete decisions to examine as they compare their model’s behavior with the task they were trying to solve.

The announcement also distinguishes the lesson from teacher preparation. Training modules for the first two quests are accessible through the Google AI Educator Series. The dedicated module for Polar Peaks is due the following month, November. Google gives no exact day for that release.

Existing teacher materials show how the programme is intended to fit a classroom. Google for Education published its Market Marshes training page on September 1. That earlier quest puts students in a simulated community crisis involving floods. They work with environmental data and forecasting models rather than treating a prediction as a finished answer. The training focuses on helping a teacher explain those choices and the work behind an AI system.

The course describes lesson slides and a teacher guide that educators can use in discussion, followed by a short assessment for a completion badge. Optional professional-development credit is presented separately from the basic materials. These are features of an existing teacher-training programme, not measurements of what students learn from the new neuroscience quest. They also help explain why adding a game and preparing its classroom support can follow different schedules.

The science behind the new story has its own timeline. On September 3, Google Research and collaborators described a complete male fruit fly central nervous system map, including the brain and ventral nerve cord. It contains more than 166,000 neurons and 125 million synaptic connections. The work was led by HHMI Janelia with researchers in Cambridge and other collaborators, providing a biological wiring diagram that scientists can examine and compare with other maps.

The research process begins with thin tissue sections and electron-microscope images. Computing and AI help assemble those flat images into three-dimensional reconstructions of cells. Human experts at HHMI Janelia verified and annotated the resulting structures. The dataset can be explored through Neuroglancer, Google’s open-source visualization tool, connecting the automated reconstruction with inspection by researchers.

Google’s September account contrasts that scale with the roughly 86 billion neurons in a human brain. Researchers use smaller organisms to investigate nervous systems and compare male and female fly maps for differences and shared structures. Polar Peaks draws on that longer research effort, translating parts of the process into a student activity rather than reproducing an entire scientific reconstruction in a classroom.