Planetary Intelligence
Will Marshall proposes Planetary Intelligence (PI), a new AI framework using real-time Earth observation data to address physical-world challenges in security, sustainability, and disaster response.
The useful question is what changes for users, developers or buyers, and whether the announcement stays industry context or becomes something people can actually use.
Will Marshall introduces Planetary Intelligence (PI), a concept for AI systems that understand the physical world in real-time by integrating sensor data such as satellite imagery, weather networks, and IoT devices. Unlike large language models (LLMs) that rely on text-based knowledge, PI leverages continuous visual and environmental data to provide actionable insights for decision-making in fields like agriculture, disaster response, and urban planning. The approach aims to bridge the gap between digital awareness and real-world action, enabling more informed and timely responses to urgent problems.
Planet Labs’ SuperDove satellites, alongside NASA’s Landsat program, have collected nearly 5,000 global images since 1972, creating a comprehensive visual record of Earth. This dataset, combined with real-time feeds from roads, ports, and weather stations, forms the foundation for Large Earth Models (LEMs). These models could analyze specific events, such as wildfires or floods, by cross-referencing satellite data with weather patterns, infrastructure, and historical records. For example, a fire chief could use an LEM to track a wildfire’s movement and plan evacuations in real-time, integrating live aerial reports and population data.
The applications of PI extend to sectors like insurance, civil governance, and maritime security. An insurance agent could query which properties are affected by a disaster and compare conditions to past events, while a government official could identify unpermitted construction by cross-referencing satellite imagery with permit records. Coast guards could monitor illegal fishing or sanctions evasion by detecting vessels with disabled transponders, generating verified evidence for enforcement actions. Planet is developing an AI app to make its global data archive queryable via natural language, enabling non-expert users to generate insights and reports at scale.
Planet Labs positions PI as a tool to enhance transparency and accountability, particularly in security and sustainability. By providing continuous, high-resolution monitoring of global activities—such as military movements or environmental changes—PI could reduce the risk of miscalculations in conflicts or crises. The framework aligns with Planet’s mission to use space-based data to drive better decisions, emphasizing the importance of real-time information in preventing escalation and enabling swift, informed action.