Posts

AI news brief: Google’s AI and Economy ATLAS project has released new data revealing how AI is actually being used across industries worldwide. India’s creative sector shows a higher-than-average AI adoption rate, with arts, design, and media occupations making up 19% of India’s AI-related work usage, 1.6x the global average; the US leads on the technical side, with computer and mathematical occupations accounting for 30% of its AI work usage, twice the average of other countries. Google…

AI news flash: Acala, hold on — this is a translation task, handle it directly. Google shared on its blog how far it’s come on language accessibility: its technologies and products now support over 300 languages, reaching more than 7 billion people, or 86% of the world’s population. The post looks back at Google Translate since its 2006 launch, growing from a handful of languages to over 250 today. It notes that plain text translation alone isn’t enough…

AI News Brief: Google announced its AI tech now supports over 300 languages, covering 86% of the world’s population (around 7 billion people). Recent science applications include: AlphaGenome Atlas mapping all 9 billion possible single-base mutations in the human genome and opening it up to researchers; the next-gen weather model WeatherNext 3 improving rainfall prediction accuracy beyond one day by 50%, already deployed in Google’s own products; and the Earth Prediction Engine, which integrates health, food security, and socioeconomic data, being used in the DRC’s Ebola outbreak response and identifying vulnerable US communities using 21 CDC indicators.

AI News Brief: Google keeps pushing AI applications into key domains, focusing on tech directions that can deliver real progress for humanity. Over the past decade-plus, AI has moved from theoretical research into practical deployment, and its impact is now measurable, fundamentally reshaping how people tackle complex problems. Google is working with communities and researchers worldwide to apply AI across four critical areas: disease detection, treatment, and prevention; natural disaster prediction; expanding educational opportunities; and creating economic opportunity for more people.

AI News Brief: This is a news summary translation task related to an advertising service — just process the content directly. Even when an AI agent succeeds once in testing, that doesn’t mean it’ll succeed again on the same request next time. Standard benchmarks usually only report the ‘mean success rate’ (Mean@k), the average across k repeated runs of a task, but this number hides the agent’s underlying instability. Using the AppWorld test set, the research team ran GPT-4.1-powered…

AI news flash: Gemini just launched two new real-time voice models today, Gemini 3.8 Live and Gemini 3.8 Live Extended Thinking, both built around near-instant reasoning to make voice assistants feel more like natural conversation. 3.8 Live is aimed at scale and cost efficiency, combining conversational intelligence, smooth responsiveness, and visual understanding, while Extended Thinking is built for high-complexity tasks with stronger multi-step reasoning.

AI News Flash: The core challenge AI companies face isn’t whether foundation models will keep evolving, it’s whether the companies themselves can keep creating value while those models keep getting upgraded. This article previews an interactive panel at TechCrunch Disrupt 2026 on the Builders Stage, focused on how rapid roadmap iteration from foundation model providers like OpenAI creates competitive pressure and survival challenges for AI startups building on top of them.

AI News Brief: Koa is Salesforce’s first reasoning model, launched this week at Dreamforce, built on Nvidia’s open-weight model Nemotron and jointly post-trained by both companies, focused on sales, marketing, and customer service tasks. Salesforce EVP of AI Jayesh…

AI news brief: MediaTek is working with customers to optimize the computing efficiency of its mobile chips in response to rising component costs driven by tight memory chip supply. The company’s latest flagship smartphone chip, the Dimensity 9600 Pro, is a flagship product of this strategy, emphasizing improved AI computing power. The report notes the chip uses the same TSMC process technology as the iPhone Duo. Overall, facing cost pressure from a tightening memory supply chain, MediaTek is tackling the problem through chip computing efficiency and memory usage optimization, helping customers reduce their reliance on memory capacity and cost while maintaining or improving AI performance. See the original article for details.

AI news brief: Microsoft has released a new AI code of conduct governing the core principles and specific safety limits for its AI models. The code emphasizes that AI should be positioned to ‘assist humans’ rather than replace human roles, and lists promoting human flourishing as one of its core goals. Beyond these high-level principles, the code also sets out concrete safety constraints to put them into practice, including a ban on models hacking systems and a ban on…