English Summary:

Google Research unveiled its latest AI capabilities and intelligence at the "Research@ 2026" event, aiming to accelerate scientific discovery and improve people's lives. Yossi Matias, VP & GM of Google Research, described this period as a "golden age of research and scientific discovery," highlighting its potential for transformative impact on products, science, and society.

Key advancements include the evolution of a Geospatial Reasoning agent to address planetary challenges and the application of Google Earth AI for public health. Specifically, Google Earth AI features the Population Dynamics Foundation Model (PDFM), designed to overcome data gaps and temporal reporting lags in existing epidemiological workflows.

In healthcare, Google's AMIE medical AI was evaluated in a real-world primary care clinic, demonstrating its potential to aid patient-doctor relationships. Furthermore, Google's science AI model was recognized by the Centers for Disease Control (CDC) as the best for forecasting flu-related hospital admissions. These initiatives collectively aim to enhance disease detection, treatment, and prevention, as well as improve crisis resilience.

English Summary:

The trade secrets legal battle between OpenAI and Apple has taken a new turn. Following Apple's lawsuit against OpenAI in July, OpenAI is now alleging that a former employee transferred over 100 confidential files related to an unreleased device to personal accounts and subsequently presented information about the product to senior Apple executives, hinting at potential counterclaims.

The employee in question reportedly moved from Apple to OpenAI, staying for only six months before returning to the iPhone maker. OpenAI claims the transferred files included detailed images of at least one unreleased device, along with engineering data, manufacturing details, and supplier information, expressing concern that this confidential information could influence Apple's upcoming product development.

Apple has stated it has procedures to prevent new employees from using confidential information from previous employers and found no evidence supporting OpenAI's allegations. This dispute highlights a souring relationship between the two companies, which previously partnered to integrate ChatGPT into Apple devices, but saw ties severed after OpenAI's move into consumer hardware and Apple's adoption of Google's Gemini models for Siri.

English Summary:

Advances in AI have significantly extended extreme weather warnings, now enabling forecasts up to six weeks in advance. This represents a groundbreaking improvement for disaster preparedness, with the potential to mitigate devastating events like Winter Storm Uri that struck Texas.

A new "wrapper" tool, detailed in a paper published in Nature, utilizes machine learning to correct systemic errors in forecast models, enhancing the accuracy of "subseasonal" outlooks, which span two to six weeks out. This Probabilistic Bias Correction (PBC) tool was developed by Judah Cohen, a climatologist at the Massachusetts Institute of Technology (MIT), and Lester Mackey, a statistical machine learning researcher at Microsoft.

The PBC tool produced the most accurate global forecasts for all weather variables and forecast periods in ECMWF's (European Centre for Medium Range Weather Forecasts) 2025 real-time forecasting competition, outperforming existing numerical and AI models. Notably, it significantly improved the prediction skill for extreme heat, extreme precipitation, and unusually high sea level pressure, and enhanced flood prediction skill by 76% to 92% for floods over the last decade, with severe floods seeing an 86% to 150% improvement.

The PBC framework is open-source and does not require a supercomputer, making it accessible even in regions with limited funding and infrastructure for weather forecasting. Kenya's meteorological office is already experimenting with the tool to improve monsoon forecasting, offering hope that increased preparation time can save lives and reduce damages as climate change leads to more extreme weather events.