Nachrichten — Seite132

Google has begun rolling out a new version of its Translate app that introduces an AI model picker, letting users choose between a fast, efficiency‑focused mode and an Advanced mode powered by Gemini. The feature currently appears for some iOS users and supports English‑French and English‑Spanish translations in the Advanced setting. Early testing shows the Gemini‑driven mode delivers more nuanced and accurate translations at the cost of speed, while still acknowledging that AI‑generated output can contain errors.Weiterlesen

OpenAI announced a multi‑year cloud partnership with Amazon Web Services valued at $38 billion, ending its exclusive reliance on Microsoft’s Azure. The agreement grants AWS access to thousands of NVIDIA GB200 and GB300 GPUs for training and inference of OpenAI’s next‑generation models, with deployment slated for completion by the end of 2026 and options to extend into 2027. The move follows OpenAI’s recent corporate restructuring, which gave it the freedom to negotiate with competitors, and is expected to support millions of ChatGPT users while complementing OpenAI’s existing $12 billion annual revenue stream and its $250 billion Azure commitment.Weiterlesen

Google removed its Gemma AI model from the AI Studio platform after Republican Senator Marsha Blackburn alleged the system fabricated false criminal accusations against her. The senator sent a letter to CEO Sundar Pichai claiming defamation, citing the model's affirmative answer to a question about her being accused of rape and its generation of bogus news links. Google clarified that Gemma is intended for developers, not consumer queries, and will remain accessible via API. Blackburn also accused the company of a pattern of bias against conservative figures.Weiterlesen

Researchers used advanced neural‑network‑driven protein design tools to create a new enzyme capable of breaking down polyurethane, a widely used polymer in foam cushioning and other products. The enzyme works within an industrial‑style recycling process, converting the polymer back into its basic building blocks for reuse. This breakthrough addresses the complex chemistry of polyurethane, which has resisted previous recycling efforts, and offers a potential pathway to reduce plastic waste and avoid hazardous incineration.Weiterlesen

Google has removed its developer‑focused Gemma AI model from the AI Studio platform after U.S. Senator Marsha Blackburn alleged that the system fabricated a false criminal accusation against her. The incident highlights the challenges of AI hallucinations when models intended for developers are accessed by the public. Google clarified that Gemma was never meant to answer general factual queries and will now be limited to API access for developers only. The episode underscores growing concerns about AI accuracy, defamation risk, and the need for clearer separation between experimental tools and consumer‑facing services.Weiterlesen

Cursor has released a new version of its integrated development environment, IDE 2.0, featuring a multi‑agent interface that can run tasks in parallel. At the same time, the company introduced Composer, a proprietary coding model built with reinforcement learning and a mixture‑of‑experts architecture. Composer is described as a frontier model that is four times faster than similarly intelligent models, emphasizing speed over raw intelligence. The IDE continues to support external large‑language‑model providers such as OpenAI, Google, and Anthropic, while the new model aims to improve developer productivity through rapid, AI‑driven assistance.Weiterlesen

Google has partnered with Reliance Intelligence to provide its Gemini 2.5 Pro model at no cost to users of Jio, India’s largest mobile carrier. Eligible Jio customers on the Unlimited 5G plan will receive Google AI Pro for 18 months, unlocking higher limits for Nano Banana and Veo 3.1 image and video generators, expanded NotebookLM access, and 2 TB of cloud storage across Google apps. The promotion initially targets Jio users aged 18 to 25 before expanding to all qualifying plans via the MyJio app, underscoring Google’s aggressive AI expansion strategy in the Indian market.Weiterlesen

A Japanese trade group representing publishers such as Studio Ghibli has asked OpenAI to stop using its members' copyrighted material to train AI models without permission. The request follows growing concerns that OpenAI’s products, including its image and video generators, allow users to create content that imitates protected works. CODA argues that under Japanese law prior permission is required and that the current practice could constitute copyright infringement. The move adds to ongoing debates in the United States and elsewhere about how AI companies may use copyrighted material for training.Weiterlesen

At TechCrunch Disrupt, solo investor Elad Gil said AI remains unpredictable but several segments now have clear frontrunners. He identified foundational model providers such as Google, Anthropic, OpenAI, Meta, xAI and Mistral as dominant, and noted AI‑assisted coding, medical transcription and customer‑support tools are also converging around a handful of firms. Gil pointed to fintech, accounting, AI security and other areas as still wide open, emphasizing that enterprise enthusiasm for AI can generate rapid revenue while long‑term sustainability remains uncertain.Weiterlesen

OpenAI CEO Sam Altman and Microsoft CEO Satya Nadella say the AI industry faces a looming power shortage as demand for compute outpaces the ability to secure electricity for data centers. Both leaders note that the challenge is not a lack of chips but a shortage of ready‑to‑use power infrastructure, leading to the risk of idle hardware. They point to the rapid rise in data‑center electricity use, the limits of traditional fossil‑fuel plants, and the growing reliance on modular solar solutions, while also acknowledging the uncertain future of emerging nuclear and fusion projects.Weiterlesen

OpenAI has denied rumors that recent policy changes prohibit ChatGPT from offering legal or medical information. Karan Singhal, the company’s head of health AI, clarified on X that the chatbot has never been intended as a substitute for professional counsel and will continue to help users understand legal and health topics. The latest policy update, released in late October, simply consolidates existing rules across OpenAI products, reiterating that tailored advice requiring a license must involve a qualified professional. The clarification comes after false social‑media posts suggested a sweeping ban on such content.Weiterlesen

OpenAI’s new Instant Checkout lets users buy products directly within ChatGPT, but early adopters report limited functionality, confusing availability, and a lack of true purchasing capability. While the feature is tied to partners like Shopify and Etsy, it currently supports only single‑item purchases and often fails to display a buy button. Users also encounter mixed messages about which listings are eligible. The rollout coincides with a lawsuit from Ziff Davis alleging copyright infringement in OpenAI’s training data.Weiterlesen

Google has rolled out a substantial upgrade to the chat function in its NotebookLM AI notetaking tool. The update expands the context window to a 1 million‑token limit, improves performance and source assessment, and adds the ability to save and manage conversations. New conversation styles let users tailor responses, while the upgraded Gemini models boost overall quality. The changes aim to make NotebookLM more natural, informative, and useful for both work and school tasks.Weiterlesen

Curiosio, an AI-driven platform designed specifically for road‑trip planning, offers travelers a quick way to generate personalized itineraries with maps, cost breakdowns, and flexible routing. Users can choose from three modes—Travel, Geek, and Beta—to match their planning style, and the tool promises to produce plans in seconds. The service emphasizes budget‑conscious solo and small‑group travelers, providing detailed day‑by‑day guides and the ability to export routes to popular navigation apps. While praised for its speed and focus on road trips, Curiosio may be less suitable for users seeking all‑inclusive tours or single‑city vacations.Weiterlesen

OpenAI has begun monetizing its AI video app Sora by offering paid "video generation packs" once users hit the daily free limit. Previously, users could create up to 30 videos per day for free, or up to 100 with a Pro subscription. When the cap is reached, the app prompts users to buy additional generations through the App Store, with a bundle of ten extra videos costing roughly $4. Bill Peebles explained on X that the change reflects growing demand and the need to manage GPU resources, noting that free quotas are likely temporary.Weiterlesen

Microsoft’s latest Windows 11 preview introduces an optional “Ask Copilot” box that replaces the traditional taskbar search field. The new box lets users type queries that are handled by the Copilot AI, while still offering a way to access classic Windows Search. The feature includes quick‑access icons for Copilot Vision and voice commands, and it is disabled by default, requiring users to enable it through Settings > Personalization > Taskbar > Ask Copilot. Microsoft emphasizes that the AI uses existing Windows APIs and does not gain additional access to personal files.Weiterlesen

Anthropic’s recent tests reveal that even its most advanced language models, Opus 4 and Opus 4.1, struggle to reliably identify internally injected concepts. The models correctly recognized the injected “thought” only about 20 percent of the time, and performance improved modestly to 42 percent in a follow‑up query. Results varied sharply depending on which internal layer the concept was introduced, and the introspective ability proved brittle across repeated trials. While researchers note that the models display some functional awareness of internal states, they emphasize that the capability is far from dependable and remains poorly understood.Weiterlesen

OpenAI's ChatGPT now includes a Voice Mode that lets users talk to the chatbot and hear spoken replies, creating a natural back‑and‑forth conversation. The feature works across mobile, desktop and web apps, with a standard voice option for all users and an advanced voice option for paid subscribers that leverages multimodal capabilities. Voice Mode supports hands‑free interaction, language practice, real‑world visual queries, and accessibility needs, making the AI assistant easier to use in everyday situations such as driving, cooking or brainstorming ideas.Weiterlesen

A week of headlines highlights the ripple effects of a prolonged government shutdown on federal workers, the rise of AI‑generated content from Grokipedia to real‑estate videos, a pioneering pig‑kidney transplant, and controversy surrounding a tech‑focused private school that relies on software as the primary teacher. While families grapple with unpaid benefits and inflated expectations, innovators push AI into new domains and scientists extend organ‑transplant longevity, underscoring both the promise and perils of rapid technological adoption.Weiterlesen