See every side of every news story
Published loading...Updated

Google DeepMind’s AlphaEarth AI Model Maps the Planet Like a “Virtual Satellite”

GLOBAL, JUL 30 – AlphaEarth processes petabytes of satellite images to map ecosystems with 23.9% greater accuracy than previous AI models, aiding classification of rainforests, deserts, and wetlands.

  • Wednesday saw the release by Google DeepMind of AlphaEarth Foundations, an AI model that functions like a `virtual satellite` to map land and coastal waters using integrated Earth observation data.
  • Before AlphaEarth, Google DeepMind had faced challenges with data overload and inconsistency in planetary mapping efforts, says Chris Brown.
  • In performance tests, AlphaEarth uses its embeddings technique to generate ecosystem data at 10-by-10-meter resolution, achieving 23.9% higher accuracy, Google said.
  • Google released the Satellite Embedding dataset in Google Earth Engine, tested by over 50 organizations over the past year, including the UN FAO and Harvard Forest.
  • These insights, from AlphaEarth, can inform decisions on food security, deforestation, urban expansion, and water resources, with access to over one trillion data points annually, Google said.
Insights by Ground AI
Does this summary seem wrong?

13 Articles

Think freely.Subscribe and get full access to Ground NewsSubscriptions start at $9.99/yearSubscribe

Bias Distribution

  • 100% of the sources lean Left
100% Left

Factuality 

To view factuality data please Upgrade to Premium

Ownership

To view ownership data please Upgrade to Vantage

Monitoreamos broke the news in on Wednesday, July 30, 2025.
Sources are mostly out of (0)