Google is leveraging LLMs to convert unstructured historical news reports into quantitative data, enabling better prediction of flash floods in data-scarce regions.
Why it matters
This application demonstrates how AI, specifically Large Language Models, can unlock valuable insights from vast amounts of unstructured historical data. By converting qualitative text into quantitative data, AI can significantly improve the accuracy and lead time of disaster predictions, such as flash floods, in regions where traditional sensor data is limited. This has direct implications for disaster preparedness, public safety, and resource allocation.
Google is using smart computer programs (AI) to read old news stories about floods. By turning the words into numbers, the AI can help predict when and where new floods might happen, even if there aren't many sensors in that area.
Cerebras Systems and Amazon Web Services are partnering to integrate Cerebras' AI chips into AWS data centers, aiming to accelerate AI applications.
Read on Economic Times Tech →BE Semiconductor Industries (Besi), a key player in advanced chip packaging technology for AI chips, is reportedly fielding takeover interest from major equipment makers like Lam Research and Applied Materials due to surging demand.
Read on Economic Times Tech →Peacock is integrating AI-driven video experiences, vertical clips, and mobile games to boost user engagement and growth.
Read on TechCrunch →