Loft Orbital's Earth-observation satellite YAM-9 ran Google DeepMind's open-weight Gemma 3 model in orbit, demonstrating autonomous classification and summarization of satellite imagery in response to natural-language queries. Enabled through a software harness built by NASA's Jet Propulsion Laboratory (JPL), participants describe it as the first publicly disclosed case of a VLM (Vision-Language Model) running in space.
Continue reading
The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in✓ Signed in — this article isn’t included in your current plan.