An Earth observation satellite autonomously identified areas of interest using a vision-language model in orbit for the first time. The Yam-9 spacecraft, built by Loft Orbital, ran NASA JPL's software powered by Google DeepMind's Gemma 3 to classify sensor data from natural language queries. This reduces the need to download raw data for human analysis.
Tap to vote and see what everyone thinks.
Summary by ByteBrief
Pokemon Go Data Not Shared With Drone Firm