US EditionGoogle AI Model Maps World in 10-Meter Squares for Machines
GLOBAL, JUL 30 – AlphaEarth Foundations reduces mapping errors by nearly 24% and uses 16 times less storage, enabling governments and organizations to monitor environmental changes with improved accuracy and speed.
- On July 30, 2025, Google DeepMind announced AlphaEarth Foundations, an AI model mapping Earth in 10-meter squares to aid scientific analysis.
- This launch responds to decades of challenges managing vast satellite data and scarce ground-truth labels that limit high-resolution mapping.
- AlphaEarth Foundations integrates diverse Earth observation data into annual global embeddings used by over 50 organizations including MapBiomas for ecosystem monitoring.
- The model reduces error rates by nearly 24% and requires 16 times less storage than prior methods, supporting detailed tracking of land cover, crops, and evapotranspiration.
- This technology promises to advance environmental management, enabling real-time, continuous mapping that supports governments, researchers, and conservation efforts worldwide.
Insights by Ground AI
Does this summary seem wrong?
18 Articles
18 Articles
Google DeepMind says its new AI can map the entire planet with unprecedented accuracy
Google DeepMind unveils AlphaEarth Foundations, an AI system that processes satellite data 16x more efficiently to create detailed Earth maps for tracking deforestation, climate change, and environmental shifts.
·San Francisco, United States
Read Full Article
The large volumes of information presented in the new tool provide dozens of different public sources, including optical satellite images, radar, laser mapping in 3D and climate simulations.
Europa SurReposted by 
El Día de Córdoba
El Día de CórdobaGoogle's new DeepMind tool opens a new era for planetary observation: it unifies petabytes of satellite data to generate compact, accurate, and consistent digital representations of the Earth. Amazon will pay $20 million a year.
Coverage Details
Total News Sources18
Leaning Left1Leaning Right0Center2Last UpdatedBias Distribution67% Center
Bias Distribution
- 67% of the sources are Center
67% Center
L 33%
C 67%
Factuality
To view factuality data please Upgrade to Premium
