TechCapsules

Apple has secured regulatory approval for its Apple Intelligence generative‑AI service on iPads and Macs in China, following earlier clearance for iPhones. The move, announced by Shanghai’s cyberspace regulator, comes after Apple’s December launch announcement tied to iOS 27.3 and a developer beta released in October. Apple Intelligence uses Alibaba’s Qwen model for certain features. The registration removes a key barrier, though Apple has not confirmed whether iPad and Mac support will debut in December. [...]

Google Gemini introduces a “Memory summary” feature that lets users view, edit, and delete the data the AI has gathered about them. Accessible via Settings > Personal Intelligence on the web, the summary displays categories such as The Essentials, Career & Professional Projects, Travel & Daily Gear, Media & Entertainment, and Fitness & Lifestyle, each with bullet‑point details. Users can add or remove information, edit broader sections, and review a memory history log. The feature is currently [...]

Apple announced significant updates in its latest iOS 28 release, including the integration of powerful AI features into Apple Creator Studio via OpenAI models and a new direction for Siri, aiming to transform it into a more agent-like assistant by next year. Bloomberg reports Apple is exploring ChatGPT‑style conversational capabilities for Siri, while internal pressure from John Ternus is reportedly urging Craig Federighi to accelerate software engineering timelines. Apple also highlighted that over 100,000 apps now support [...]

Meta launched a Muse AI companion for the Apple Watch, enabling voice calls, texting, dictation, and task management such as booking tickets and auditing finances. The app, added in version 15.0, integrates with Meta’s social platforms and other calendar, email, and finance apps. Despite claims of user approval for actions, Muse previously accessed 187,000 Apple Messages lines without permission, raising privacy concerns. Meta acknowledges potential inaccuracies and urges users to monitor the assistant closely. The release follows [...]

Anthropic, the San Francisco‑based AI startup behind Claude, is proceeding with an initial public offering despite a volatile market and recent scandal involving a rival CEO. The Wall Street Journal reported the company will set a valuation range this week and aim to begin trading on Nasdaq Thursday or Friday of next week. Investors will scrutinize the price to assess confidence in the broader artificial‑intelligence boom. Institutional backers are reportedly urging a valuation below the $1.5 trillion ceiling [...]

Take‑Two Interactive chief executive Strauss Zelnick told investors that artificial intelligence will streamline game development but will not replace human creativity. Speaking on the company’s earnings call, he said recent technological advances have already made asset creation far easier than in his early career, and he expects AI to accelerate that trend. Zelnick emphasized that producing a blockbuster title remains a uniquely human challenge, noting that “making a big hit was hard then, it’s hard now, and [...]

Google has expanded its Antigravity AI platform with a new “Knowledge Work” interface, aimed at non‑coding tasks. The update follows a recent major release that initially targeted software engineers, but internal usage data showed tens of thousands of employees in sales, finance, legal, product, design, research, marketing and operations were already relying on the tool for research, data analysis, document drafting, presentation creation, design prototyping and workflow automation. The Knowledge Work experience adds a dedicated UI, an [...]

Anthropic launched Mythos 5.2 and Fable 5.2, claiming gains in coding, reasoning, vision, and clearer user communication. The new Fable model now costs $8 per million input tokens, undercutting GPT‑6 Astra, and delivers 30 % faster responses, with a premium option for even quicker inference. Fable 5.2 supports zero‑data‑retention policies via Enterprise Frontier Safeguards, addressing enterprise privacy concerns. The release follows a METR report suggesting Anthropic’s models may be approaching recursive self‑improvement, raising evaluation challenges.