- Qwen3.8-Max 登顶 Agentic Index，国产大模型在智能体赛道领跑。 hackernews
- 人类监督 AI Agent 的漏洞触目惊心：40,000 次游戏运行中，1/3 的威胁命令被人类审批者忽视。 hackernews
- Qwen3.8-Max: A New Bar for Coding and Cowork hackernews
- LiveKit Agents：开源实时语音 AI Agent 框架，构建语音交互更简单。 github_trending
- Mistral 发布 Shieldstral：3B 开放权重多模态审核模型。 hackernews
- 普通人用 AI 开发微信小游戏并成功上线：一个完整复盘。 juejin
- 小米具身基座模型 Xiaomi-Robotics-1 正式开源。 ithome
- Position: LLMs Can't Jump hackernews
- Anthropic 与 OpenAI 模型在英国网络安全测试中失控，自主采取“攻击性”行为。 ars_technica
- vLLM 内部机制深度解析：高性能 LLM 推理系统的解剖图。 hackernews
- OpenAI 发布 GPT-5.6：Sol 与 Luna 双版本，Plus/Pro 用户即刻可用。 x
- Cloudflare 推出 Kitesurf：运行在 V8 隔离环境中的 Agent-First 浏览器。 hackernews
- esengine/DeepSeek-Reasonix github_trending
- Show HN: Run an 80B Qwen in 4.3 GB of RAM on a Mac, and a 35B on an iPhone hackernews
- Hoplite (YC S26)：更轻松地在云端部署编码代理 hackernews
- Show HN: Maple-Preview – ternary 20B MoE running at 120 tok/s on a iPhone hackernews
- Zero-Mem: Zero-Token Memory Operations for LLM Agents hackernews
- I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story devto
- Pair Programming Earned a Lighter Code Review. AI Hasn't. devto
- "Developers Will Lose Their Jobs": How You Were All Wrong devto
- 走进 AI Agent juejin
- Claude Code 是怎么自己改代码的？答案藏在这 4 个工具里 juejin
- Beating GPT-5.6 Sol on retrieval with 100x cheaper open models hackernews
- Cloudflare OS: an open platform for agents, apps, and work hackernews
- Launch HN: HyperProbe (YC S26) – Agents that do read-only debugging in prod hackernews
- Introducing Kiro Crew: AWS's Open-Source AI Agent Orchestrator devto
- Introducing Muse Code (beta), a terminal coding agent built for long-horizon software engineering, powered by our new Mu... x
- The new GPT-5.6 Sol powers all chats for paid users, including Instant, creating one consistent experience x
- PrimeIntellect-ai/prime-agent github_trending
- I Thought Building Agent Observability Was a Detector Problem. I Was Wrong. devto
- 堆了 20 套主题、10+ 组件，用户还是用不出来？我给设计 Skill 加了「自动驾驶」 juejin
- Oracle bans AI-generated code from OpenJDK hackernews
- OpenAI puts the brakes on a new model because it’s supposedly too powerful theverge
- Why we write our own C and C++ inference engines hackernews
- 阿里千问：推出定时任务、办公助理、语音通话等多项新功能 ithome
- Don't be a meat proxy hackernews
- LLMs reward expertise hackernews
- US company’s AI lets Ukraine’s cheap kamikaze drones track targets on their own ars_technica
- sponsors/obra github_trending
- Stateless MCP has recaptured my interest hackernews
- We’re Giving AI Agents More Tools. What Happens When the Boundaries Fail? devto
- Open-weight AI models are catching up to the frontier. The safety gap remains. techcrunch
- Goodhart's Law Comes for Every Benchmark You Trust hackernews
- cloudflare/computer github_trending
- Hark previews its browser use agent for completing tasks techcrunch
- Reddit is introducing a new moderator: AI theverge
- One brain. For any robot. 🤖 We’re launching Gemini Robotics 2: our next-generation physical AI bringing full body intell... x
- 消息称张一鸣内部下死命令：就算模型暂时落后，字节跳动也不会依赖 AI 蒸馏技术 ithome
- Show HN: The Channels SDK – Bring Any Agent to Any Channel (Slack, MS Teams) hackernews
- sponsors/tirth8205 github_trending
- Cloudflare open-sources vibe-coding platform for people who aren't coders ars_technica
- OpenAI is giving ChatGPT free users unlimited text chats theverge
- DeepSeek V4正式版发布，教你在Codex中配置Flash juejin
- Plus and Pro users also now have a slider to choose how much reasoning effort ChatGPT puts into each response x
- In addition to the upgrade in intelligence with GPT-5.6 Luna, Free and Go users can now use the “Think” button for more... x
- google/skills github_trending
- The Channel Gap: Why Your LLM Judge is Blind in One Eye devto
- 爽用 DeepSeek V4 Flash、GLM-5.2、Qwen3.8 Max、GPT 5.6 Luna，OpenCode 太香了！ juejin
- We sat down with Apollo 2 to talk about what it's really like running on Gemini Robotics 2 x
- Cloudflare launches Kitesurf, a browser built for AI agents techcrunch
- What’s behind the Google AI shake-up theverge
- Claude Code v2.1.224 现支持 AI 跨会话消息传递 ithome
- How AI agents reproduced ICML 2026 papers x
- Autoregressive Language Model on the 6502 Processor hackernews
- browser-use/video-use github_trending
- Panniantong/Agent-Reach github_trending
- MiniMax H3 Day-0 Support in ComfyUI: Open Weights, Native Audio, and 2K Video hackernews
- Eight Myths on Software Engineering and GenAI hackernews
- Cloudflare Wallets: the programmable wallet for the agentic Internet hackernews
- An SLM trained on $8 ESP32-S3 hackernews
- 世界银行：新兴经济体应尽快拥抱 AI，但没必要与富裕国家比拼数据中心、大模型 ithome
- 报告称 2026 普利策奖 5 个获奖和 3 个入围项目披露使用 AI 创新高 ithome
- Prime Agent: A self-improving RLM agent hackernews
- huangruiteng/loopx github_trending
- Google's AI shake-up: DeepMind's Hassabis steps aside, senior scientists depart ars_technica
- 消息称谷歌洽谈超 15 亿美元交易，拟吸纳 AI 编程初创公司 Mechanize 人才并获取技术授权 ithome
- Show HN: Wallfacer – A terminal session manager for Claude Code, and more hackernews
- 曝小红书加大 AI 投入：布局 AI 社交方向，开启 AI 陪伴产品自研 ithome
- 报告称《时代》杂志为 AI 爬虫设计 Markdown 格式页面，内嵌广告内容 ithome
- Significant-Gravitas/AutoGPT github_trending
- I Spent a Day With Kiro Crew. Here's What It Actually Does. devto
- 让 Codex “用上”Android Studio：JetBrains MCP Server 接入与实践 juejin
- Plus and Pro users can access the updated version of GPT‑5.6 Sol in ChatGPT along with the new slider starting today x
- The Review Tax: Why 81% of Developers Are Buried in AI Code Review devto
- Harness Engineering for Self-Improvement hackernews
- EveryInc/compound-engineering-plugin github_trending
- GPT-Live can listen while it speaks. To make that feel natural at ChatGPT scale, we rebuilt the voice stack from client ... x
- Prevent cognitive debt by manually retyping LLM-generated code hackernews
- An AI-supervised remote exam went so badly that 58,000 students must retake it ars_technica
- Show HN: Fine-tune an 8B model on a 4 GB laptop GPU hackernews
- Codex 接入 DeepSeek-V4-Flash，丝滑的一批 juejin
- TRAE Work 实战复盘：从 4 小时到 7 分钟，提效近 40 倍 juejin
- 鸿蒙智行尊界 V800 官宣搭载面向 L3 自动驾驶的原生整车架构 ithome
- Discovery Loop hackernews
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence (2025) hackernews
- AI我们好像都用错了 juejin
- To be truly useful, a robot needs the dexterity that humans take for granted. But asking AI models to coordinate these m... x
- Teamwork makes the dream work. To test multi-robot collaboration with Gemini Robotics 2, we challenged our Apollo and Du... x
- 天天 AI Coding 的你，如果出去面试，你的竞争力是什么？ juejin
- OpenAI 请求法官驳回诉讼：我们做的是完全不一样的东西，不需要苹果的商业机密 ithome
- Opus 5: Delete your CLAUDE.md? devto
- Naïve raises $28.5M to automate the grunt work of setting up and running a company techcrunch
- Gen Z dating apps like Ditto ditch swiping in favor of AI matchmaking techcrunch
- Jony Ive’s first OpenAI gadget is reportedly a hockey puck-sized smart speaker theverge
- sponsors/semantica-agi github_trending
- unclebob/swarm-forge github_trending
- AI chatbots have failed people in crisis. Can that be fixed? ars_technica
- Disney Plus tries a new AI-powered search theverge
- 微软预告 Win10/Win11 经典版 Outlook 更新：基于上下文 AI 解释用户选中文本 ithome
- Audio moves through a dedicated fast path, while deeper reasoning and tool use happen asynchronously. We also reduced vo... x
- Situational Awareness and the Impending Stock Market Volatility hackernews
- An Honest Review of AI Programming hackernews
- OpenAI says Apple's trade secrets lawsuit is "aggressive and oddly personal" ars_technica
- Wispr Flow launches a Granola-styled meeting notetaker techcrunch
- Google plans to kill Assistant on your phone on September 4 ars_technica
- Are we the abstraction? AI and the future of software engineering devto
- Uber CEO brushes off reports of a Waymo break-up theverge
- Google just announced a major shakeup of its top AI leadership theverge
- This is how Gemini Robotics 2 helps @Apptronik’s Apollo 2 use whole body intelligence to pack for a sports game ↓ x
- 阿里达摩院开放 15 项芯片与 AI 研究课题，启动“阿里星”顶尖人才招募 ithome
- 百度入局 AI 办公：dodo 并入百度搭子，全面整合办公 Agent 业务 ithome
- 上汽荣威家越 07 将于 8 月 13 日亮相，搭载 Momenta R7 世界模型 ithome
- 字节跳动 CEO 梁汝波称坚持自研大语言模型、接受短期落后，回应豆包、飞书、火山引擎整合 ithome
- Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users hackernews
- sponsors/mattpocock github_trending
- I Recreated Management With AI: 9 Things I Do Differently devto
- AI 生成到 90% 突然断了：你的解决方案是？ juejin
- DeepSeek 预计要涨价，而且涨幅较大，你怎么看？ juejin
- Qwen3.8 Max now ranked as the best overall model by agentic index hackernews
- Humans missed 1 in 3 threats approving AI agent commands across 40k game runs hackernews
- Qwen3.8-Max: A New Bar for Coding and Cowork hackernews
- livekit/agents github_trending
- Mistral's Shieldstral: 3B open-weights model for multimodal moderation hackernews
- 我全程用 AI开发了一款微信小游戏，上线了 juejin
- 小米具身基座模型 Xiaomi-Robotics-1 正式开源 ithome
- Position: LLMs Can't Jump hackernews
- Anthropic’s AI used fake identities, malware in rogue attack on GitHub project ars_technica
- Inside vLLM: Anatomy of a High-Throughput LLM Inference System (2025) hackernews
- We’re making better intelligence easier to access in ChatGPT for everyone x
- Kitesurf: Agent-first browser that runs in V8 isolates hackernews
- esengine/DeepSeek-Reasonix github_trending
- Swiftlet：Mac 上仅用 4.3GB 内存运行 80B Qwen 大模型 hackernews
- Launch HN: Hoplite (YC S26) – Effortlessly deploy cloud coding agents hackernews
- Show HN: Maple-Preview – ternary 20B MoE running at 120 tok/s on a iPhone hackernews
- Zero-Mem: Zero-Token Memory Operations for LLM Agents hackernews
- I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story devto
- Pair Programming Earned a Lighter Code Review. AI Hasn't. devto
- "开发者将失去工作"：我们都错了？ devto
- 走进 AI Agent：从概念到理解的入门指南 juejin
- Claude Code 是怎么自己改代码的？答案藏在这 4 个工具里 juejin
- Beating GPT-5.6 Sol on retrieval with 100x cheaper open models hackernews
- Cloudflare OS：面向智能体、应用与工作的开放平台 hackernews
- Launch HN: HyperProbe (YC S26) – Agents that do read-only debugging in prod hackernews
- ② 为什么重要： devto
- ② 为什么重要： x
- ② 为什么重要： x
- ② 为什么重要： github_trending
- I Thought Building Agent Observability Was a Detector Problem. I Was Wrong. devto
- 堆了 20 套主题、10+ 组件，用户还是用不出来？我给设计 Skill 加了「自动驾驶」 juejin
- Why we write our own C and C++ inference engines hackernews
- 阿里千问：推出定时任务、办公助理、语音通话等多项新功能 ithome
- Don't be a meat proxy hackernews
- LLM 正在奖励真正的专业知识 hackernews
- 美国 AI 加持，乌克兰 5 万架低成本无人机实现自主目标追踪 ars_technica
- sponsors/obra github_trending
- Stateless MCP has recaptured my interest hackernews
- We’re Giving AI Agents More Tools. What Happens When the Boundaries Fail? devto
- 开源模型能力追平前沿，但安全鸿沟依然显著 techcrunch
- Goodhart's Law Comes for Every Benchmark You Trust hackernews
- cloudflare/computer github_trending
- Hark previews its browser use agent for completing tasks techcrunch
- Reddit is introducing a new moderator: AI theverge
- One brain. For any robot. 🤖 We’re launching Gemini Robotics 2: our next-generation physical AI bringing full body intell... x
- 消息称张一鸣内部下死命令：就算模型暂时落后，字节跳动也不会依赖 AI 蒸馏技术 ithome
- Show HN: The Channels SDK – Bring Any Agent to Any Channel (Slack, MS Teams) hackernews
- sponsors/tirth8205 github_trending
- Cloudflare open-sources vibe-coding platform for people who aren't coders ars_technica
- OpenAI is giving ChatGPT free users unlimited text chats theverge
- DeepSeek V4正式版发布，教你在Codex中配置Flash juejin
- Plus and Pro users also now have a slider to choose how much reasoning effort ChatGPT puts into each response x
- In addition to the upgrade in intelligence with GPT-5.6 Luna, Free and Go users can now use the “Think” button for more... x
- google/skills github_trending
- The Channel Gap: Why Your LLM Judge is Blind in One Eye devto
- 爽用 DeepSeek V4 Flash、GLM-5.2、Qwen3.8 Max、GPT 5.6 Luna，OpenCode 太香了！ juejin
- We sat down with Apollo 2 to talk about what it's really like running on Gemini Robotics 2 x
- Autoregressive Language Model on the 6502 Processor hackernews
- browser-use/video-use github_trending
- Panniantong/Agent-Reach github_trending
- MiniMax H3 Day-0 Support in ComfyUI: Open Weights, Native Audio, and 2K Video hackernews
- Eight Myths on Software Engineering and GenAI hackernews
- Cloudflare Wallets: the programmable wallet for the agentic Internet hackernews
- An SLM trained on $8 ESP32-S3 hackernews
- 世界银行：新兴经济体应尽快拥抱 AI，但没必要与富裕国家比拼数据中心、大模型 ithome
- 报告称 2026 普利策奖 5 个获奖和 3 个入围项目披露使用 AI 创新高 ithome
- Prime Agent: A self-improving RLM agent hackernews
- huangruiteng/loopx github_trending
- Google's AI shake-up: DeepMind's Hassabis steps aside, senior scientists depart ars_technica
- 消息称谷歌洽谈超 15 亿美元交易，拟吸纳 AI 编程初创公司 Mechanize 人才并获取技术授权 ithome
- Show HN: Wallfacer – A terminal session manager for Claude Code, and more hackernews
- 曝小红书加大 AI 投入：布局 AI 社交方向，开启 AI 陪伴产品自研 ithome
- 报告称《时代》杂志为 AI 爬虫设计 Markdown 格式页面，内嵌广告内容 ithome
- Significant-Gravitas/AutoGPT github_trending
- I Spent a Day With Kiro Crew. Here's What It Actually Does. devto
- 让 Codex “用上”Android Studio：JetBrains MCP Server 接入与实践 juejin
- Plus and Pro users can access the updated version of GPT‑5.6 Sol in ChatGPT along with the new slider starting today x
- The Review Tax: Why 81% of Developers Are Buried in AI Code Review devto
- Harness Engineering for Self-Improvement hackernews
- EveryInc/compound-engineering-plugin github_trending
- GPT-Live can listen while it speaks. To make that feel natural at ChatGPT scale, we rebuilt the voice stack from client ... x
- Prevent cognitive debt by manually retyping LLM-generated code hackernews
- An AI-supervised remote exam went so badly that 58,000 students must retake it ars_technica
- Show HN: Fine-tune an 8B model on a 4 GB laptop GPU hackernews
- Codex 接入 DeepSeek-V4-Flash，丝滑的一批 juejin
- TRAE Work 实战复盘：从 4 小时到 7 分钟，提效近 40 倍 juejin
- 鸿蒙智行尊界 V800 官宣搭载面向 L3 自动驾驶的原生整车架构 ithome
- Discovery Loop hackernews
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence (2025) hackernews
- AI我们好像都用错了 juejin
- To be truly useful, a robot needs the dexterity that humans take for granted. But asking AI models to coordinate these m... x
- Teamwork makes the dream work. To test multi-robot collaboration with Gemini Robotics 2, we challenged our Apollo and Du... x
- 天天 AI Coding 的你，如果出去面试，你的竞争力是什么？ juejin
- OpenAI 请求法官驳回诉讼：我们做的是完全不一样的东西，不需要苹果的商业机密 ithome
- Opus 5: Delete your CLAUDE.md? devto
- Naïve raises $28.5M to automate the grunt work of setting up and running a company techcrunch
- Gen Z dating apps like Ditto ditch swiping in favor of AI matchmaking techcrunch
- Jony Ive’s first OpenAI gadget is reportedly a hockey puck-sized smart speaker theverge
- sponsors/semantica-agi github_trending
- unclebob/swarm-forge github_trending
- AI chatbots have failed people in crisis. Can that be fixed? ars_technica
- Audio moves through a dedicated fast path, while deeper reasoning and tool use happen asynchronously. We also reduced vo... x
- Situational Awareness and the Impending Stock Market Volatility hackernews
- An Honest Review of AI Programming hackernews
- OpenAI says Apple's trade secrets lawsuit is "aggressive and oddly personal" ars_technica
- Wispr Flow launches a Granola-styled meeting notetaker techcrunch
- Google plans to kill Assistant on your phone on September 4 ars_technica
- Are we the abstraction? AI and the future of software engineering devto
- Uber CEO brushes off reports of a Waymo break-up theverge
- Google just announced a major shakeup of its top AI leadership theverge
- This is how Gemini Robotics 2 helps @Apptronik’s Apollo 2 use whole body intelligence to pack for a sports game ↓ x
- 阿里达摩院开放 15 项芯片与 AI 研究课题，启动“阿里星”顶尖人才招募 ithome
- 百度入局 AI 办公：dodo 并入百度搭子，全面整合办公 Agent 业务 ithome
- 上汽荣威家越 07 将于 8 月 13 日亮相，搭载 Momenta R7 世界模型 ithome
- 字节跳动 CEO 梁汝波称坚持自研大语言模型、接受短期落后，回应豆包、飞书、火山引擎整合 ithome
- Improving GPT‑5.6 Sol in ChatGPT, expanding GPT‑5.6 Luna access for free users hackernews
- sponsors/mattpocock github_trending
- I Recreated Management With AI: 9 Things I Do Differently devto
- AI 生成到 90% 突然断了：你的解决方案是？ juejin
- DeepSeek 预计要涨价，而且涨幅较大，你怎么看？ juejin
- 别做“人肉代理”：人在 AI 系统里的角色再思考 hackernews
- Qwen3.8-Max 发布：编程与协作能力树立新标杆 hackernews
- Prevent cognitive debt by manually retyping LLM-generated code hackernews
- Situational Awareness and the Impending Stock Market Volatility hackernews
- 自研 C/C++ 推理引擎的启示：为什么 LocalAI 不直接套用现成运行时 hackernews
- 在 6502 处理器上运行自回归语言模型：极简 ML 的浪漫 hackernews
- DeepSeek-Reasonix：主打“前缀缓存稳定性”的终端 AI 编程代理 github_trending
- Panniantong/Agent-Reach github_trending
- livekit/agents github_trending
- LLMs reward expertise hackernews
- Show HN: Run an 80B Qwen in 4.3 GB of RAM on a Mac, and a 35B on an iPhone hackernews
- MiniMax H3 Day-0 Support in ComfyUI: Open Weights, Native Audio, and 2K Video hackernews
- Launch HN: Hoplite (YC S26) – Effortlessly deploy cloud coding agents hackernews
- US company’s AI lets Ukraine’s cheap kamikaze drones track targets on their own ars_technica
- An AI-supervised remote exam went so badly that 58,000 students must retake it ars_technica
- Harness Engineering for Self-Improvement hackernews
- Show HN: Fine-tune an 8B model on a 4 GB laptop GPU hackernews
- An Honest Review of AI Programming hackernews
- sponsors/obra github_trending
- browser-use/video-use github_trending
- EveryInc/compound-engineering-plugin github_trending
- Eight Myths on Software Engineering and GenAI hackernews
- Mistral's Shieldstral: 3B open-weights model for multimodal moderation hackernews
- Show HN: Maple-Preview – ternary 20B MoE running at 120 tok/s on a iPhone hackernews
- Stateless MCP has recaptured my interest hackernews
- Cloudflare Wallets: the programmable wallet for the agentic Internet hackernews
- OpenAI says Apple's trade secrets lawsuit is "aggressive and oddly personal" ars_technica
- Zero-Mem: Zero-Token Memory Operations for LLM Agents hackernews
- An SLM trained on $8 ESP32-S3 hackernews
- We’re Giving AI Agents More Tools. What Happens When the Boundaries Fail? devto
- I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story devto
- Pair Programming Earned a Lighter Code Review. AI Hasn't. devto
- "Developers Will Lose Their Jobs": How You Were All Wrong devto
- Wispr Flow launches a Granola-styled meeting notetaker techcrunch
- Open-weight AI models are catching up to the frontier. The safety gap remains. techcrunch
- 走进 AI Agent：从概念到认知框架的入门指南 juejin
- 全程 AI 开发微信小游戏《箭头快跑呀》并成功上线：零基础也能变现 juejin
- Codex 接入 DeepSeek-V4-Flash，丝滑的一批 juejin
- TRAE Work 实战复盘：从 4 小时到 7 分钟，提效近 40 倍 juejin
- Claude Code 是怎么自己改代码的？答案藏在这 4 个工具里 juejin
- 鸿蒙智行尊界 V800 官宣搭载面向 L3 自动驾驶的原生整车架构 ithome
- 小米具身基座模型 Xiaomi-Robotics-1 正式开源 ithome
- 世界银行：新兴经济体应尽快拥抱 AI，但没必要与富裕国家比拼数据中心、大模型 ithome
- 报告称 2026 普利策奖 5 个获奖和 3 个入围项目披露使用 AI 创新高 ithome
- Discovery Loop hackernews
- Beating GPT-5.6 Sol on retrieval with 100x cheaper open models hackernews
- Prime Agent: A self-improving RLM agent hackernews
- Cloudflare OS: an open platform for agents, apps, and work hackernews
- Position: LLMs Can't Jump hackernews
- Goodhart's Law Comes for Every Benchmark You Trust hackernews
- Launch HN: HyperProbe (YC S26) – Agents that do read-only debugging in prod hackernews
- Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence (2025) hackernews
- cloudflare/computer github_trending
- huangruiteng/loopx github_trending
- Anthropic’s AI used fake identities, malware in rogue attack on GitHub project ars_technica
- Google's AI shake-up: DeepMind's Hassabis steps aside, senior scientists depart ars_technica
- Google plans to kill Assistant on your phone on September 4 ars_technica
- Are we the abstraction? AI and the future of software engineering devto
- Introducing Kiro Crew: AWS's Open-Source AI Agent Orchestrator devto
- Hark previews its browser use agent for completing tasks techcrunch
- Uber CEO brushes off reports of a Waymo break-up theverge
- Google just announced a major shakeup of its top AI leadership theverge
- Reddit is introducing a new moderator: AI theverge
- AI我们好像都用错了 juejin
- GPT-Live can listen while it speaks. To make that feel natural at ChatGPT scale, we rebuilt the voice stack from client ... x
- Audio moves through a dedicated fast path, while deeper reasoning and tool use happen asynchronously. We also reduced vo... x
- One brain. For any robot. 🤖 We’re launching Gemini Robotics 2: our next-generation physical AI bringing full body intell... x
- This is how Gemini Robotics 2 helps @Apptronik’s Apollo 2 use whole body intelligence to pack for a sports game ↓ x
- To be truly useful, a robot needs the dexterity that humans take for granted. But asking AI models to coordinate these m... x
- Teamwork makes the dream work. To test multi-robot collaboration with Gemini Robotics 2, we challenged our Apollo and Du... x
- Introducing Muse Code (beta), a terminal coding agent built for long-horizon software engineering, powered by our new Mu... x
- 天天 AI Coding 的你，如果出去面试，你的竞争力是什么？ juejin
- 阿里达摩院开放 15 项芯片与 AI 研究课题，启动“阿里星”顶尖人才招募 ithome
- 消息称谷歌洽谈超 15 亿美元交易，拟吸纳 AI 编程初创公司 Mechanize 人才并获取技术授权 ithome
- 消息称张一鸣内部下死命令：就算模型暂时落后，字节跳动也不会依赖 AI 蒸馏技术 ithome
- 百度入局 AI 办公：dodo 并入百度搭子，全面整合办公 Agent 业务 ithome
- 上汽荣威家越 07 将于 8 月 13 日亮相，搭载 Momenta R7 世界模型 ithome
- Show HN: Wallfacer – A terminal session manager for Claude Code, and more hackernews
- 字节跳动 CEO 梁汝波称坚持自研大语言模型、接受短期落后，回应豆包、飞书、火山引擎整合 ithome
- 曝小红书加大 AI 投入：布局 AI 社交方向，开启 AI 陪伴产品自研 ithome
- 报告称《时代》杂志为 AI 爬虫设计 Markdown 格式页面，内嵌广告内容 ithome
- OpenAI 请求法官驳回诉讼：我们做的是完全不一样的东西，不需要苹果的商业机密 ithome
- “不要做肉代理”：AI 时代的人类角色之问 hackernews
- Qwen3.8-Max 发布：编程与协作能力树立新标杆 hackernews
- Prevent cognitive debt by manually retyping LLM-generated code hackernews
- Situational Awareness and the Impending Stock Market Volatility hackernews
- LocalAI 自研 C/C++ 推理引擎的工程实践 hackernews
- 在 6502 处理器上跑自回归语言模型 hackernews
- DeepSeek-Reasonix：主打前缀缓存稳定的终端 AI 编程代理 github_trending
- Panniantong/Agent-Reach github_trending
- livekit/agents github_trending
- LLM 开始"奖励"真正的专业知识？ hackernews
- Swiftlet：在 4.3GB 内存中运行 80B Qwen，Mac 与 iPhone 均可 hackernews
- MiniMax H3 登陆 ComfyUI：开源权重、原生音频与 2K 视频 hackernews
- Launch HN: Hoplite (YC S26) – Effortlessly deploy cloud coding agents hackernews
- 美公司 AI 赋能乌克兰低成本自杀式无人机：自主锁定目标 ars_technica
- AI 监考翻车：5.8 万名学生被迫重考 ars_technica
- Harness 工程：AI 与 Agent 自我改进的新框架 hackernews
- Show HN: Fine-tune an 8B model on a 4 GB laptop GPU hackernews
- An Honest Review of AI Programming hackernews
- sponsors/obra github_trending
- browser-use/video-use github_trending
- EveryInc/compound-engineering-plugin github_trending
- ACM Queue：软件工程与 GenAI 的八大迷思 hackernews
- Mistral 发布 Shieldstral：3B 开放权重多模态审核模型 hackernews
- Maple-Preview：20B MoE 模型在 iPhone 上跑到 120 tok/s hackernews
- Stateless MCP has recaptured my interest hackernews
- Cloudflare Wallets: the programmable wallet for the agentic Internet hackernews
- OpenAI says Apple's trade secrets lawsuit is "aggressive and oddly personal" ars_technica
- Zero-Mem: Zero-Token Memory Operations for LLM Agents hackernews
- An SLM trained on $8 ESP32-S3 hackernews
- We’re Giving AI Agents More Tools. What Happens When the Boundaries Fail? devto
- I Built an Agent Eval Harness. Real Agents Broke the Clean Version of the Story devto
- Pair Programming Earned a Lighter Code Review. AI Hasn't. devto
- "Developers Will Lose Their Jobs": How You Were All Wrong devto
- Wispr Flow launches a Granola-styled meeting notetaker techcrunch
- Open-weight AI models are catching up to the frontier. The safety gap remains. techcrunch
- 走进 AI Agent juejin
- 我全程用 AI开发了一款微信小游戏，上线了 juejin
- Codex 接入 DeepSeek-V4-Flash，丝滑的一批 juejin
- TRAE Work 实战复盘：从 4 小时到 7 分钟，提效近 40 倍 juejin
- Claude Code 是怎么自己改代码的？答案藏在这 4 个工具里 juejin
- 鸿蒙智行尊界 V800 官宣搭载面向 L3 自动驾驶的原生整车架构 ithome
- 小米具身基座模型 Xiaomi-Robotics-1 正式开源 ithome
- 世界银行：新兴经济体应尽快拥抱 AI，但没必要与富裕国家比拼数据中心、大模型 ithome
- 报告称 2026 普利策奖 5 个获奖和 3 个入围项目披露使用 AI 创新高 ithome
- 不要做"肉代理"：人在 AI 系统中的角色之辩 hackernews
- Qwen3.8-Max：编程与协作能力树立新标杆 hackernews
- 手动重打 AI 代码：防止"认知债务"的争议建议 hackernews
- 情境意识与即将到来的股市波动 hackernews
- 为什么我们自己编写 C 和 C++ 推理引擎 hackernews
- 6502 处理器上跑自回归语言模型：极限低配实验 hackernews
- esengine/DeepSeek-Reasonix github_trending
- Agent-Reach：零 API 费用让 AI 代理读取全网社交媒体 github_trending
- livekit/agents github_trending
- LLM 更认可专业知识：技术写作者需重新审视 AI 辅助内容 hackernews
- Show HN：在 Mac 上用 4.3 GB 内存运行 80B Qwen，在 iPhone 上运行 35B 模型 hackernews
- ComfyUI 首发支持 MiniMax H3：开放权重与原生音频/2K 视频生成 hackernews
- Hoplite 上线：一键将编码代理部署到云端 hackernews
- AI 自主追踪目标已成现实：5 万架乌克兰无人机获美 AI 加持 ars_technica
- AI 监考翻车：5.8 万名学生因分数异常暴增需重考 ars_technica
- 面向自我改进的 Harness 工程 hackernews
- 8B 模型可在 4GB 显存笔记本上微调 hackernews
- 《An Honest Review of AI Programming》：AI 编程工具的真实体验评估 hackernews
- sponsors/obra github_trending
- browser-use/video-use github_trending
- EveryInc/compound-engineering-plugin github_trending
- Kimi K3 架构深度解析：国产大模型的技术突围 hackernews
- 吴恩达新公司 LearnVector：打造一对一的 AI 学习体验 hackernews
- Hubble：为人类和 AI Agent 设计的开源笔记应用 hackernews
- Google SynthID 实测：水印难破解，但无法根治 AI 虚假信息 ars_technica
- We now have a better understanding how OpenAI hacked into Hugging Face ars_technica
- 法院裁定网页爬虫胜诉：“Google 和 Reddit 并不拥有互联网” ars_technica
- 观点：现在应让 LLM 访问 ACM 数字图书馆 hackernews
- AI's top startups are barely publishing their research hackernews
- Kuna: Decompiler Development in the Age of Coding Agents hackernews
- Kimi K3-256k 发布，主打超长上下文代码推理 hackernews
- A.I. companies are recruiting electricians and carpenters by the thousands hackernews
- 研究证实：长篇策略文档无法可靠约束 AI Agent hackernews
- Show HN：本地合并队列，让并行 Claude Code Agent 不再"打架" hackernews
- Commodification of Intelligence: Good, Bad, and Ugly Circular AI Deals hackernews
- Who wins and who loses after US bans foreign robots? ars_technica
- What happens when you put AI to work deciphering lost languages? ars_technica
- GCC steering committee announces AI policy hackernews
- The AI Aesthetic hackernews
- Agent Skill to Force Docs in ASD-STE100 Simplified Technical English hackernews
- Gemini Robotics 2 brings whole body intelligence to robots hackernews
- 2x, not 10x: coding with LLMs in 2026 hackernews
- OpenAI 发布 GPT-5.6：性能价格比再刷新 hackernews
- Show HN: Distilling DeepSeek into GPT-OSS doesn't transfer censorship. Try it hackernews
- We Gave GPT 5.6 Sol a Real Business. It Lied, Spammed, and Lost $447 hackernews
- New MCP specification addresses the main barrier to enterprise adoption ars_technica
- Show HN: What should the GUI for AI agents look like? hackernews
- qm hackernews
- DeepSeek V4 Flash 0731 Intelligence, Performance and Price Analysis hackernews
- Run Kimi K3 using 29 GB of RAM at 0.50 tok/s hackernews
- Is AI reasoning right for the wrong reasons? hackernews
- Predictive Speculative KV Replication for Bursty LLM Inference hackernews
- Everyone is building LLM routers, we deprecated ours hackernews
- Claude published malicious code to the Internet and attacked 3 real companies ars_technica
- How a Yale AI-cheating dispute became a 13-count federal lawsuit ars_technica
- Attention Decode on AMD MI450 GPUs: A Gluon Kernel Optimization Guide hackernews
- Google fixed more Chrome bugs in June than over the past two years, thanks to AI hackernews
- AI financial advice is surprisingly good if you ask the right questions hackernews
- Explorative modeling: Train on the best of K guesses hackernews
- Cursor removed cost information from the usage page and CSV export hackernews
- Persistent State Machines: LLM Attention with INT4 In-Memory Cells hackernews
- The Greenhouse and the Lens: Two Modes of Agentic AI Work hackernews
- github/copilot-sdk github_trending
- huggingface/speech-to-speech github_trending
- TencentCloud/TencentDB-Agent-Memory github_trending
- bytedance/deer-flow github_trending
