Planetary Intelligence
The article discusses the concept of Large Earth Models (LEMs), a new type of machine intelligence that integrates real-time physical world data. It contrasts LEMs with large language models (LLMs), highlighting the limitations of LLMs in understanding real-world events. The author emphasizes the importance of satellite data and other sensory inputs in creating a more comprehensive understanding of our environment.
- ▪Large Earth Models (LEMs) aim to provide real-time insights into the physical world, unlike large language models (LLMs) which are limited to textual data.
- ▪Satellite data, such as that from NASA's Landsat and Planet's SuperDove, serves as a foundational element for LEMs, offering continuous visual memory of the Earth.
- ▪LEMs could provide specific information about events like floods, comparing current conditions to historical data, enhancing our understanding of environmental changes.
Opening excerpt (first ~120 words) tap to expand
Planetary IntelligenceWill MarshallMay 27, 2026ShareEverything, everywhere, all at once.In this essay, I introduce an idea for a new type of machine intelligence that understands our physical world in real-time – a powerful expansion of AI’s capabilities. I talk about implications for people across economic, security, and sustainability domains. And I speculate about humanity's place in the cosmos, and how giving AI sensors may help humans and machines coexist safely and survive the Great Filter.Photo by NASAI. The Models That Are BlindThe AI large language models that have captivated the world have consumed the written record of human civilization: every article, essay, book, and conversation that humanity has committed to the internet.
…
Excerpt limited to ~120 words for fair-use compliance. The full article is at Hacker News (Newest).
Discussion
0 commentsMore from Hacker News (Newest)
