#81
在生产环境中保护MCP：超越网关的纵深防御
推荐理由在生产环境中保护MCP：超越网关的纵深防御。
Pai 的 AI 雷达
含近几日仍热话题，按时间衰减排序。今天先看 Anil-matcha/Open-Generative-AI，其次 Anthropic asks users to stop being mean to Claude。并行还有 Anthropic bans users from abusive or cruel behavior towards Claude、B-Trees Are Back: Engineering Fast and Pageable Node Layouts [pdf]。
NVIDIA 与 Microsoft 在旧金山活动上宣布为 Windows PC 共同打造 AI Agent 软硬件。
Today, one in three pull requests on GitHub involves an AI agent . A year ago, that number was fewer than one in 10. If that pace holds, within the next two years, most of the code pushed to GitHub could be written by an agent. Much of it may never be fully read by a human. If de
Bradley Axen, Head of AI Capabilities at Block, on how his team evaluates new large language models, how Claude Fable 5 orchestrates company-wide code migrations, and why safeguards are a critical part of deploying frontier intelligence responsibly, at scale.
The missing map of the sky
Diffusion-based models decompose sampling into many small Gaussian denoising steps, an assumption that breaks down when generation is compressed to a few coarse transitions. Existing few-step methods address this through distillation, consistency training, or adversarial objectiv
📌 一句话摘要 本文揭示了华为流程体系背后的统一商业底层逻辑：商业的实质是确定性的生产与定价，企业需通过 IPD 供给发动机与 MTL/LTC 变现发动机的双轮驱动，将「从市场机会到产品发布」升级为「从市场机会到商业变现」。 📝 详细摘要 很多企业推行集成产品开发（IPD）时仅停留在「小 IPD」层面，交付了合格样机却未能带动利润增长。本文从商业底层逻辑切入，指出商业的本质是确定性的生产、转移、定价与分配。围绕这一核心，企业需应对四类不确定性：BLM 解决方向不确定性、IPD 解决供给不确定性、MTL/LTC 解决交易不确定性、ITR 解决使用结果不确定
College Planner is coming to ChatGPT for Teens to help students manage college applications, alongside new flashcards, quizzes, and a teen AI council.
If your Copilot usage metrics have shown agent activity or agent lines of code falling while Copilot usage kept growing, we’ve found the cause, and a fix is rolling out to each IDE. Several IDEs recently moved Copilot agent sessions to the Copilot SDK. Those sessions didn’t ident
Zach Yadegari, the teen co-founder of popular Cal AI calorie tracking app, has launched a new personal AI agent startup that competes with Instinct, Muse, and Bee.
📌 一句话摘要 作者指出，使用 Opus 进行视频生成的主要瓶颈是 CPU 计算能力，而非其他因素。 📝 详细摘要 这一条信息简明扼要地表达了作者的观察：在利用 Opus 进行视频创作时，当前主要的性能瓶颈是中央处理器的计算速度，而不是模型本身的能力或其他硬件组件。 💡 主要观点 CPU 成为性能瓶颈 作者明确指出，在当前的视频生成实验中，CPU 的处理速度已经成为限制工作效率的关键因素，而不是模型容量或内存的限制。 💬 文章金句 现在限制用 Opus 做视频的是 CPU 📊 文章信息 AI 初评： 66 来源： AI产品黄叔(@PMbackttfutu
The developer of Hermes Agent raised a $90 million Series B.
PLUS: Agilink's robot hands are getting crafty
Tickets for AIE NYC are selling out soon! See you next week! see past AINews issues for subscriber discounts. Pour one out for Mistral, who shipped a decent Large 4 “Le Chonk” model on the new 3800 GB300 cluster funded by their recent Series D . But they were overshadowed by more
PLATO V terminal, via Wikipedia . In 1960, a team at the University of Illinois released the first version of their PLATO “computerized instruction system,” a network of what would eventually be thousands of computer terminals connected to large mainframe computers. Users of PLAT
It’s been over a year since Reflection launched with us with big goals on coding (and hinted about their RL approach ): But they stayed “stealth” longer than Thinking Machines and it was not clear we would ever get a model launch out of them, as the broader open model ecosystem d