Introducing Google Earth AI & AlphaEarth Foundations
And why they matter to educators
A constant refrain is that artificial intelligence is fundamentally reshaping the world around us. It’s now also cataloging the globe in unprecedented ways. Yesterday, we learned of a foundational shift in how humans can gather, understand, and derive meaning from geospatial information.
A central player is ‘Google Earth AI’, a new and significant AI-driven transformation, and its AlphaEarth Foundations, a new “virtual satellite” model that can see through clouds and map intricate earth details. Both were introduced yesterday. Their unveilings fell under the radar in the face of new NotebookLM and ChatGPT features (more on that tomorrow). Still, they have important implications for educators and students across STEM, history, geography, and other disciplines.
AlphaEarth Foundations: The Virtual Satellite
Let’s start with AlphaEarth Foundations. The model takes a ton (“petabytes”) of data from disparate Earth observations and weaves them all together. It then condenses this vast amount of information into what’s called an “embedding." Essentially, an embedding creates compact and uniform summary for every 10-by-10-meter square of the planet's surface and coastal waters. These "embeddings" are easy for computers to process, thus providing a rich, detailed and consistent picture of the planet's evolution.
AlphaEarth Foundations capabilities are pretty impressive:
Seeing Through the Clouds: It can map farmland development in places where persistent cloud cover traditionally made agricultural monitoring incredibly difficult.
Unprecedented Detail: It renders detailed surface features in challenging areas like Antarctica, which are notoriously challenging for conventional satellites. It can even reveal subtle agricultural land-use changes that conventional satellite imagery often misses.
Efficiency and Accuracy: This model is very efficient, requiring 16 times less storage than comparable AI systems. It's also the first Earth observation visualization approach to support continuous time, meaning it can create a continuous view of a location.
Google is making its massive collection of AlphaEarth embeddings available through “Google Earth Engine.” Over 50 organizations, including the UN Food and Agriculture Organization and Stanford University, already use this system. The technology currently powers flood and wildfire alerts in Google Search and Maps and in the future, it could be combined with Gemini for even more insights.
Freely available come August 8! Please check this newsletter next week for more details.
Google Earth AI
Google Earth AI is an umbrella of sorts for Google’s geospatial initiatives. Google has a collection of different AI models, datasets, and tools designed to tackle critical global needs, from environmental monitoring and disaster prediction to urban planning and enhancing map accuracy. AlphaEarth Foundations, for instance, is a specific, new AI model that sits under that Google Earth AI umbrella. So, when we talk about AI-powered weather forecasts, flood predictions, or generative AI grounding Google Maps, these are part of the larger Google Earth AI initiative.
The aforementioned Google Earth Engine, a core part of the Google Earth ecosystem, allows other parts of Google Earth AI or users to leverage it. In essence, Google Earth Engine (which has been around for decades) provides the robust platform, and Google Earth AI, through AlphaEarth Foundations, provides the intelligent, refined data, creating a tremendous resource for dynamic, inquiry-based learning across STEM, history, geography, and beyond.
Schools and universities can access Google Earth Engine’s powerful “Satellite Embedding dataset.” The excellent news is that it's available free of charge for academic and teaching purposes. (Select "Register a Noncommercial or Commercial Cloud project" and "Unpaid Usage" for Education.) More good news is that Google Earth Engine and the AlphaEarth Foundations dataset are designed to make complex geospatial data usable for even middle and high school students.
The Pedagogical Power of Google Earth AI
Google Earth AI is impressive. But what does it mean for educators? Well, for one, it means moving beyond static maps and textbook examples to dynamic, real-world engagement, nurturing a deeper, more active understanding of our planet and its complex systems.
STEM & Data Literacy: From Code to Climate Models
Google Earth AI provides a treasure trove of real-world, dynamic data for analysis, modeling, and problem-solving in science, technology, engineering, and mathematics.
Environmental Science: Students can analyze historical and real-time flood data or wildfire boundaries to understand climate change impacts. They could even simulate mitigation strategies, turning abstract concepts into tangible investigations.
Data Science & Programming: Google Earth AI offers opportunities for students to engage with the concepts of AI, data visualization, and predictive analytics, even without coding. For advanced students, it can be a gateway to understanding agentic workflows and large-scale data integration.
Engineering & Urban Planning: Students could analyze congestion patterns in their own communities, propose traffic signal optimizations, or even design hypothetical evacuation routes, using real data and AI-powered simulations. In other words, Google Earth AI can foster practical problem-solving and an understanding of complex systems design.
History & Geography: Reimagining Place and Time
The accuracy provided by Google Maps means historical and geographical explorations become rich and dynamic.
Historical Analysis: Students can use accurate, current geospatial data to explore how landscapes have changed over time, analyze demographic shifts, or map the spread of historical events. One use is to trace the path of a historical migration and overlay it with environmental data to understand its challenges.
Contemporary Geography: Understanding natural hazards becomes visceral. Students can track real-time weather patterns, analyze flood risks in vulnerable regions, or monitor wildfire progression. This fosters a deeper understanding of human-environment interaction, resource management, and global challenges.
Global Citizenship: Google has put a focus on data-scarce regions in Africa for weather and flood forecasting. This provides case studies for discussions on global equity, sustainable development, and how AI can address real-world humanitarian challenges.
Conclusion
The integration of advanced AI into Google Earth AI and its related geospatial platforms is an impressive leap forward in our capacity to understand, predict, and interact with the physical world. For the educational landscape, this means moving beyond theoretical learning to hands-on engagement with real-world data and challenges. Educators now have powerful new avenues to foster critical thinking, data literacy, interdisciplinary problem-solving, and a deeper sense of global citizenship in their students.
Any suggestions to improve this newsletter? Please message me or leave a comment below!