TechStuff NASA and AI: Decoding Our Universe
Jan 13, 2026
Join Philip Thigo, Kenya's Special Technology Envoy, Kevin Murphy from NASA, and Juan Bernabe Moreno of IBM, as they dive into the integration of AI and satellite data for environmental impact. Thigo shares insights on Kenya's ambitious 15 billion tree campaign and how AI is used for effective flood response. Murphy discusses NASA's open-data policies and the powerful Prithvi model for analyzing Earth observation data. Moreno elaborates on innovative applications predicting weather phenomena and tracking harmful algal blooms, showcasing a future where tech meets climate action.
AI Snips
Chapters
Transcript
Episode notes
NASA's Massive Open Data Treasure
- NASA collects enormous, free, open geospatial datasets from dozens of satellites that anyone can use.
- These datasets enable unexpected discoveries like mapping groundwater and glacier mass via gravity measurements.
Transformers Learn Earth's Physical Patterns
- IBM and NASA adapted transformer-based foundation models to learn Earth systems from satellite data.
- The models learned physical patterns and generalized well, enabling faster, more accurate analysis.
Prithvi's Origin And Meteorologist Reaction
- IBM and NASA launched Prithvi, a geospatial foundation model trained on Landsat and Sentinel-2 data.
- Prithvi expanded into weather and climate modeling and impressed NASA meteorologists with its capabilities.