Google Quietly Releases a Faster Model in Mythos’ Shadow
量子位 QbitAI3 days agoReleaseThe provided QbitAI title indicates that Google released a model quietly while attention was focused on Mythos. The only concrete performance claim available is that speed increased by 4x, but the model name, task scope, benchmark method, and availability are not provided. Based on the title alone, this appears to be a model-release item relevant to developers and AI practitioners tracking latency and throughput improvements.
DiffusionGemma: Google Launches High-Speed Open-Weight Gemma Diffusion Model★ 76
Simon Willison's Weblog3 days agoReleaseSimon Willison highlights Google’s new DiffusionGemma, an Apache 2 licensed open-weight Gemma model. He connects it to last year’s brief Gemini Diffusion preview, which he measured at 857 tokens per second. NVIDIA is currently hosting the model for free on its NIM cloud API, where Willison generated 2,409 tokens in 4.4 seconds, implying at least 500 tokens per second.
Nobody needs AI to search the Internet, court says in Google ruling★ 74
Ars Technica AI4 days agoRegulationArs Technica reports that Google lost a German court fight involving AI Overview, with the court rejecting the idea that AI is necessary for searching the Internet. The ruling matters because AI search products summarize web content in ways that may reduce visits to original sources. If courts treat AI summaries as optional rather than essential search infrastructure, Google and rivals may face tougher legal limits around content use, attribution, and publisher impact.
Google will save your Lens photos, Search Live recordings, and Translate audio for AI training
The Verge AI4 days agoEthicsGoogle has notified users via email that it will begin saving multimedia inputs—images from Google Lens, real-time recordings from Search Live, and audio from Translate—under a new 'Search Services History' setting. This data will be retained and potentially used to train and improve Google's AI models. Users concerned about privacy should review their account settings to manage or disable this data collection.
Apple Reveals New AI Architecture Built Around Google Gemini Models★ 78
Hacker News (AI keywords)6 days agoReleaseApple announced a major Apple Intelligence overhaul built around Apple Foundation Models co-developed with Google using technologies behind Gemini. The architecture supports on-device and Private Cloud Compute execution, with stronger reasoning, understanding, and multimodal capabilities. A new system orchestrator coordinates AI features across Apple platforms, though Apple has not yet specified which devices receive the higher-power model.
Google will pay SpaceX $920M per month for compute★ 78
TechCrunch AI9 days agoBusinessSpaceX announced a major compute rental deal with Google one week before its expected Nasdaq debut. From October 2026 through June 2029, Google will pay $920 million per month for access to about 110,000 NVIDIA GPUs, plus CPUs, memory, and related components. The agreement resembles SpaceX’s recent Anthropic deal and includes a 90-day cancellation option after December 31, 2026.
The LLM warnings Google fired Timnit Gebru over have all come true
Hacker News (AI keywords)10 days agoEthicsThe post frames Timnit Gebru’s dispute with Google as an early warning about large language model risks. Based on the available title, it appears to argue that concerns around bias, accountability, concentration of power, and deployment risks have since become visible in practice. This is best read as AI ethics commentary, not a model release or technical tutorial.
Gemini's new AI agent is about as good as Google's demo
The Verge AI13 days agoNew ToolGoogle's new 24/7 AI agent, Gemini Spark, can take on tasks for users and continue working on them. After receiving access last week, The Verge's reviewer found that Spark can perform surprisingly well, roughly matching Google's demo. The remaining question is whether that capability justifies the financial cost and potential privacy tradeoffs.
Google 推出全新開源大語言模型 Gemma 2：9B 與 27B 雙版本，效能超越同級模型★ 85
Hugging Face Blog717 days agoReleaseGoogle has officially launched the next generation of its open-source large language model, Gemma 2, with an initial release in two sizes — 9B (9 billion…
Google 官方推出 CodeGemma：專為程式碼生成與補全設計的輕量級開源模型★ 80
Hugging Face Blog796 days agoReleaseGoogle and Hugging Face have jointly announced the launch of CodeGemma, a family of lightweight open-source large language models (LLMs) designed specifically…