今日必看

GitHub Copilot Agent 权限集中管理

If you administer GitHub Copilot Business or GitHub Copilot Enterprise, you can now centrally control which agent operations are blocked, require human approval, or can proceed without a prompt. Managed permissions cover shell commands, file reads and edits, and network domains.

CodeQL 支持 Linux ARM64 并扩展分析

CodeQL 2.27.0 is now available on Linux ARM64, adds a new Rust security query, expanded framework coverage for Java/Kotlin and C#, and analysis accuracy improvements across multiple languages. CodeQL is the static analysis engine behind GitHub code scanning, which helps you find

Mistral 用 AI Agent 迁移遗留代码

Mistral helped a European energy operator migrate 40,000 lines of Fortran 77 to C++. Learn how it was done, and the lessons to carry forward.

多模态服务的 EPD 推理拆分

Encode-prefill-decode (EPD) disaggregation is an inference optimization technique for multimodal models that separates the vision encoder stage from the prefill and decode stages. It is most effective for image-heavy prompts, short-to-medium outputs, and quantized mixture-of-expe

Paul Christiano 加入 OpenAI Foundation 董事会

Paul Christiano joins the OpenAI Foundation Board and its Safety and Security Committee, bringing experience in AI alignment, safety, and standards.

IBM 发布 Granite 时序模型

Anthropic 讨论经济未来情景

推动 AI 安全证据与政策行动

Chris Lehane argues that stronger AI capabilities require stronger safety evidence, shared standards, and durable policy action while the policy window remains open.

Google Search 新增足球功能

Track live game feeds, explore detailed stats, and get custom fantasy recommendations directly in Search this season.

LLM 系统工程设计的上下文建模框架

The development of generative artificial intelligence resources enables opportunities of speeding up systems and engineering design work. This contribution introduces a framework of formal operations for assembling context in LLM-based engineering design. This framework involves

OmniMed-FL: A Robust Multimodal Federated Learning Framework for Clinical Diagnosis

Simultaneous assessment of medical imaging and patient records is often required in clinical diagnosis. However, standard machine learning algorithms cannot analyze these data types together. Meanwhile, compliance with HIPAA and GDPR can constrain centralized aggregation of sensi

Qwen 3.8 follows GPT-5.5 Pro reasoning prefills

大喜：家居博主靠 "婆婆逼生二胎送别墅" 爆火，如何用编剧思维快速起号？

📌 一句话摘要 家居博主幕后操盘手大喜拆解如何用编剧思维批量打造爆款家居账号：从婆媳冲突选题、封面标题设计到评论区运营、矩阵孵化与商业变现。 📝 详细摘要 本期对话「一条就爆」MCN 创始人、全网 200 万粉家居博主幕后推手大喜。她只有高中学历，做过影楼修图师、销售培训和直播带货，凭借过年回村拍摄公公自建别墅意外起量，随后以「婆婆为了让我生二胎，送了我一栋别墅」等爆款选题切入家居赛道，四年孵化约 15 个账号。她强调做号前要做「资产审计」和定位，明确吸引哪类人群、赚哪部分品牌的钱；选题要抓住社会情绪（婆媳矛盾、女性觉醒、性价比消费），用反差与冲突制造情

I made an MCP app so claude code can record, edit and export demo screen recordings for you

Just say "film a 30-second demo of my app" and your agent opens it, moves the cursor, clicks through it, and hands you back a polished video. Works with Claude Code, Codex, OpenClaw or any compatible AI agent through MCP. Supports basically every app or website! Just ask and get

Why Emacs Consult async searches feel slow and how to speed them up

SODAX SDK

Digital asset flows you can build and deploy with AI. Discussion | Link

Suno launches v6 music models built with Warner, BMG, and Believe

Suno has unveiled a new AI music model generation, v6, in three versions, built together with Warner Music Group, BMG, and Believe. All older models are being shut down. Songs can now be partially changed through text commands or generated multimodally from text, audio, and image

Suno replaces its AI models with a new one trained on licensed music as copyright suits pile up

As it grapples with a bevy of lawsuits, Suno said its new model, Suno v6, is not trained using music it used to train previous versions of the AI model.

GPT-6 让 48 个网页验证码失效了，最聪明的 AI 和最笨的人类相遇了

机器人也把「我不是机器人」勾上了。 GPT-6 成功通过了「我不是机器人」游戏中的所有 48 个关卡。 「我不是机器人」是独立开发者 Neal Agarwal 在 neal.fun 上做的网页游戏，随着关卡推进，任务会变得越来越离谱、越来越不像「验证」，比如在涂鸦墙里找到威利、在配音板上玩「西蒙说」、画一个完美的圆，甚至在关卡里让你和「女友」分手、停车入位。 你可能会咬牙切齿做完，可 GPT-6 只是挥挥手表示：不吃压力。 人类的验证码已经拦不住 AI 验证码（CAPTCHA），全称是「全自动区分计算机和人类的公开图灵测试」。 由于它是用计算机来考人类，

“超脑+狼族”最新全阵容亮相！京东物流打造供应链应用场景最多、规模最大的AI具身机器人军团

四岁男童的肢体接触，不是制造流量漩涡的素材

📌 一句话摘要 人民日报评论指出，长沙四岁男童奔跑中不慎触碰女子引发的舆情争议，折射出「性骚扰」「维权」等词汇被流量绑架的社会现象，呼吁公共空间回归理性与宽容。 📝 详细摘要 文章以长沙一起四岁男童在餐厅奔跑时意外触碰女性乘客的事件为切入点，指出该事件在互联网上被放大为舆情风波，核心原因是自媒体借「性骚扰」「维权」等话语标签制造流量。文章批评将无心之失泛政治化、维权化的倾向，认为这会蚕食公共空间的善意与宽容。文中强调，纠纷解决应优先平等协商而非诉诸网络曝光，平台对涉事账号采取措施是必要的，舆论不应替代法律审判。最终呼吁社会少一些戾气，多一些理性包容。 💡

GPT-6 Astra, Looped Transformers, and Hidden Reasoning

A lot has happened in the last few weeks. I am sure that OpenAI’s GPT-6 Astra is top of mind for everyone right now. In particular, thoughts on its performance, the looped transformer/recurrent depth aspects, and rumors that Astra is “hiding” its reasoning trace (i.e., chain of t

Runway 推出 Adobe 插件，Premiere Pro 和 After Effects 内可直接生成和编辑

Runway 发布 Runway Plugins，面板可直接嵌入 Premiere Pro 和 After Effects，无需导出导入即可在时间线内生成图像和视频、重渲染片段并一键放置结果。

DeepSeek、千问、智谱轮番登场，PC 厂商终于等到了它们的弹药

没有新显卡，也没有跑分轰炸，这场 IFA 2026 的 AMD 开幕演讲只讲了一件事：Personal AI（个人 AI）。 以及支撑它的三样新品：Ryzen AI Max 400（代号 Gorgon Halo）平台、下一代 HP ZBook 笔记本的首度曝光，和一台 96 核、全液冷的「桌上超算」Threadripper Halo Station。 AMD 高级副总裁、计算与图形总经理 Jack Huynh 在 45 分钟的演讲里，从绕月飞行的 Artemis II、手术机器人讲到芬兰的 LUMI 超算和 PlayStation 5 Pro，回顾了一遍

[AINews] GPT-6 Astra: OpenAI’s biggest LLM launch of all time

The launch is barely 9 hours old, and with 36M views and 164K likes, already is OpenAI’s most successful launch since Sora and certainly GPT-4 or GPT-5 . You’ll recall we’ve previously observed that Anthropic tends to far outclass OpenAI in launch popularity. For the first time i