Google Earth Gets an AI Chatbot to Help Chart the Climate Crisis (www.wired.com)

Google has added a conversational layer to Google Earth by integrating its AlphaEarth Foundations satellite-data model with Gemini-powered geospatial reasoning, letting users query terabytes of imagery and environmental layers just like a chatbot. The system fuses historical satellite imagery, surface temperature, elevation contours, air-pollution and population datasets, plus weather and other models, to surface targeted answers—e.g., “find algae blooms”—and generate lists, maps and time‑series that reveal changes over years. These capabilities build on Google Earth’s professional tiers (Professional at $75/month, Professional Advanced at $150/month), while core time-lapse features remain free. For the AI/ML community this demonstrates a practical multimodal fusion of large language/model reasoning and high-volume geospatial data to accelerate climate monitoring and disaster forecasting. Key technical implications include scaling geospatial LLM inference over petabyte-class imagery, aligning heterogeneous models (imagery, weather, population) for causal or predictive insights, and exposing the need for explainability, bias assessment, and latency-sensitive updating. The rollout highlights opportunities for AI-driven environmental intelligence—identifying vulnerable communities, tracking policy impacts, and improving response—but also raises questions about access, model transparency and the trade-offs between AI’s energy footprint and its climate benefits.