跳到正文
8月31日周一
  1. The Decoder:AI News69

    OpenClaw 2.0 发布:简化设置、重构浏览器应用并支持多人会话

    OpenClaw 基金会发布开源 AI 平台 2.0 版,为迄今最大更新,包含超 16,000 个拉取请求。新版本简化首次设置,可自动检测 ChatGPT 或 Claude 订阅、API 密钥及本地模型;浏览器应用完全重构,新增显示会话进度的 Session Rail 功能。共享云会话支持多人协作,会话可在本地网关、配对设备或 Crabbox 租用的临时机器上运行。

  2. OpenClaw:Blog72

    OpenClaw 2.0 发布:迄今最大更新,由 933 位贡献者打造

    OpenClaw 2.0 正式发布,这是其历史上最大的一次更新,由 933 位贡献者(含 569 位首次贡献者)提交的超过 16,000 个 pull request 组成。新版简化了安装流程,重建了浏览器应用,并新增共享云会话功能,支持多人协作。OpenClaw 为开源项目,更新涵盖安装、消息、记忆、技能、模型、自动化、浏览器、插件及安全等全部模块。

  3. MarkTechPost64

    OpenClaw 2.0 发布:引导式模型设置、575 ms 控制 UI 启动与统一信任边界

    OpenClaw 2.0 正式发布,重写安装流程并重构浏览器控制 UI,将会话与转录迁移至 SQLite,新增共享云会话。引导式设置可复用本机已验证的 Codex、ChatGPT 或 Claude CLI 登录,并在保存前验证模型可用性;模拟测试中 UI 启动从约 1.6 秒降至 575 ms,JavaScript 请求从 140 降至 45。共享会话支持多人协作,但文档明确其并非安全边界。

  4. 公众号:腾讯混元51

    混元 Hy4 preview 调用激增,WorkBuddy 已紧急扩容

    混元 Hy4 preview 自 8 月 28 日在 WorkBuddy 首发接入后,因 Agent 能力提升引发调用激增,上线首日即出现排队,官方已对推理集群紧急扩容。Hy4 preview 限免至 9 月 10 日 23:59,每日免费额度有限;Hy3 限免延长至 9 月 30 日 23:59。Hy4 preview 暂不具备图像/视频等多模态生成能力,相关任务将自动切换至其他模型执行。

  5. IT之家·人工智能44

    北邮牵头:全球首个太空算力云实现常态化在轨试验服务

    北京邮电大学牵头的全球首个对外提供在轨试验服务的太空算力云,正式常态化提供在轨算力服务。该平台已累计完成数百次太空算力调用,服务用户超百家,基于国产原生软硬件生态,大模型推理能效比达10Tokens/J,太空算力服务覆盖率超10%。其依托天算星座运行,已累计在轨部署16颗低轨卫星,支撑60余家单位开展在轨试验。

  6. Rohan Paul48

    Rohan Paul 引述 Gavin Baker 观点称,数据中心正成为美国地方经济引擎:Quincy 镇贫困率从 29% 降至 6%,税收新建了高中、医院和图书馆;Loudoun 县每年从数据中心征收约 10 亿美元,Quincy 数据中心占房产税超一半。Baker 承认此前担忧过度,称水电、税收、就业等问题已基本解决,电价在多州正因数据中心下降。

    引用Gavin Baker@GavinSBaker

    Regret the tone of my post on data centers yesterday. What I should have said: There were reasonable concerns about data centers 18ish months ago: water, taxes, jobs, electricity prices, the environment and what they would do to small towns. Well-structured data center projects have largely addressed these concerns today and we should be celebrating this. On balance, data centers are awesome for America in every way. On water: U.S. data centers use a fraction of what golf courses use. A lot of the numbers from 18 months ago were off by over 1000x. Newer data centers use closed-loop systems or recycled water. Should be required by every town approving a data center project. On taxes: looking only at sales-tax exemptions, as Ronan Farrow did, is the wrong way to evaluate this. Data centers pay significant property taxes. Loudoun County, which is the wealthiest county in America, now collects on the order of $1 billion a year from data centers. In Quincy, WA, data centers are more than half the property-tax roll. Over time, property taxes can go to zero while government spending increases in these towns. On jobs: this has been unambiguously awesome for blue collar Americans. Demand for electricians, plumbers, welders, HVAC techs, and contractors has gone vertical, and it is not a one-time construction job. These buildings get upgraded and expanded over time. That is why the building trades are fighting for them, and why some unions are now treating opposition to data centers as a reason not to endorse politicians. On power: the original fear was that households would pay for the incremental electricity demand in the form of higher prices. That is why the ratepayer-protection deals and the new large-load tariffs exist. The right structure is: the data center brings or pays for new generation and signs a contract long enough that existing customers are protected. Where that is happening, utilities are cutting or freezing residential rates and saying so on the record. Where it is not, people are right to object. Electricity prices are going down *today* in a number of large states because of data centers. 
On the environment: data centers overwhelming use natural gas today, which is the cleanest power source outside of nuclear, solar and wind. And the companies that are building the data centers are committed to carbon neutrality such that an equivalent amount of solar will likely be built. Maybe more importantly, the data centers need batteries to function effectively and these batteries can also sell energy back into the grid (which recently prevented blackouts in Texas). Over time, data centers will run on solar plus batteries. On the towns: Poverty in Quincy, WA fell from 29% to 6%. Data center taxes paid for a new high school, a hospital, a library, police and fire stations. This is happening in many left for dead former mill and farm towns that had no other bidder for the land. Data centers are actually reindustrializing parts of America and creating the kind of working-class jobs both parties have spent decades claiming to support. That should not be a partisan issue. Data centers can and should be awesome for America and they increasingly, overwhelmingly are. Supporting the outsourcing of data centers to China will likely age just as well as support for the outsourcing of high quality, blue collar manufacturing jobs to China has aged. When the facts change, I change my mind. I hope that reasonable people who had good faith reasons to oppose data centers at least consider updating their beliefs given the change in the facts over the last 18 months. This really matters for America. I will say I also think the idea of making data centers beautiful is a good one that has yet to be implemented. Data centers should be just as beautiful as Grand Central Station. We can learn a lot from the railroad buildout. Neoclassical revival ftw. Might write up open-weight AI tomorrow as this is equally essential to America.

  7. MarkTechPost65

    语音与实时智能体最低延迟推理 API:以首 token 延迟(TTFT)为先的基准评测

    语音智能体的真实体验取决于首句延迟(TTFS)而非首 token 延迟(TTFT),后者仅标记生成开始,TTS 模型需等完整句子才能发声。评测覆盖 LLM、STT、TTS 及语音到语音全链路,并给出各层延迟预算:STT 约 100–200ms、LLM 300–500ms、TTS 100–200ms,端到端目标 700ms–1.2s。

  8. TechCrunch:AI42

    Caterpillar 将采矿自动化经验应用于 AI 部署

    工业巨头 Caterpillar 正将采矿自动化经验应用于更广泛的 AI 部署,推出 Cat AI Assistant,供现场技术人员通过语音指令获取维修流程和故障排查信息。该公司拥有约 160 万个联网设备和超 16 PB 结构化数据,并计划未来五年投入 1 亿美元培训 11.8 万名员工。其第二季度营收达 205 亿美元创历史新高,发电部门销售额因数据中心需求增长 72%。

  9. TechCrunch:AI67

    马斯克自建燃气轮机叶片铸造厂加速AI供电,但污染问题随之而来

    马斯克证实SpaceX在得州Bastrop秘密建设“叶片与导流叶铸造厂”,以自产燃气轮机核心部件,缓解AI数据中心电力瓶颈。他表示此举可将燃气轮机投产时间提前最多18个月。但燃气轮机排放已引发诉讼与健康担忧,弗吉尼亚州研究估计单座设施每年可致3.4至6.5例过早死亡。

8月30日周日
  1. IT之家·人工智能51

    网传微软 Win11 偷偷后台跑 AI?实为权限管理列表

    网传 Windows 11 在后台偷偷运行 AI 并拖累性能的截图实为误读。该“文本和图像生成”页面仅记录哪些应用拥有使用设备端生成式 AI 模型的权限,并非实际运行记录,且“最近活动”显示请求数为 0。相关功能在 Copilot+ PC 的 NPU 上运行,不会占用 CPU、GPU 或普通系统内存。

  2. IT之家·人工智能58

    马斯克确认 SpaceX 自建燃气轮机叶片铸造厂,称可提前 18 个月投运

    马斯克在 X 平台确认,SpaceX 正筹划在得克萨斯州巴斯特罗普建设燃气轮机动叶与静叶铸造厂,以突破 AI 数据中心电力扩张瓶颈。全球仅 3 家铸造厂能生产大型燃气轮机热端叶片,订单已排满至 2030 年。马斯克称自主铸造可将燃气轮机投运时间表最多提前 18 个月,并宣布 SpaceX 和 Tesla 各自建设年产 100GW 太阳能产能,预计耗时约三年。

  3. IT之家·人工智能47

    消息称百度智能云分拆平台产品事业部:MaaS 划入基础设施、Agent 独立成军

    百度智能云完成产研组织调整,原平台产品事业部更名为智能体事业部,MaaS 业务、千帆品牌及相关产品体系转入基础设施事业部,数据平台部更名为数据智能部并转入智能应用事业部。调整后,侯震宇、殷大伟、阮瑜分别负责基础设施、智能体和智能应用事业部,均向沈抖汇报。百度本月财报显示 AI 业务收入占比 50%,GPU 云同比增长 283%。

  4. Rohan Paul52

    马斯克计划在 SpaceX 内部自建涡轮叶片铸造,以绕过燃气轮机厂商售罄至 2030 年的瓶颈,使涡轮机提前最多 18 个月上线。SpaceX 与特斯拉正各自建设每年 100GW 的太阳能产能,但未来数年仍需天然气补充。

    引用Elon Musk@elonmusk

    SpaceX and Tesla are each building 100GW/year of solar production capacity as fast as possible, but natural gas will still be needed to supplement and bootstrap solar for several years. The limiting factor for nat gas turbine production is casting the blades & vanes. By doing in-house casting at SpaceX, we can accelerate nat gas turbines coming online by up to 18 months, which is a profound game-changer.

  5. MarkTechPost48

    使用 NVIDIA Earth2Studio 构建自定义批量集合天气预报工作流

    本教程演示如何用 NVIDIA Earth2Studio 构建集合天气预报工作流,安装组件并加载 FCN 预报模型,从 GFS 获取初始条件,实现将 10 米风分量转为涡轮容量因子的风电诊断及变尺度扰动系统。通过低层迭代器、坐标映射、批处理和 Zarr API 构建执行管道,用纬度加权 RMSE、fair CRPS 等指标验证预报,并可视化集合不确定性。

  6. Tencent Hy47

    腾讯混元 Hy4-preview 从发布首日起即支持在 vLLM 中运行,并已在 NVIDIA GPU 上完成验证。该模型总参数量 770B,激活参数 49B,采用 256 个路由专家加 1 个共享专家,支持 1M 上下文但每次查询仅关注 2048 个 token。

    引用vLLM@vllm_project

    @TencentHunyuan's Hy4-preview runs in vLLM from day 0, verified on NVIDIA GPUs. 🎉 - 770B total, 49B active, 256 routed experts plus one shared - 1M context, but each query attends to just 2048 tokens - Only 21 of the 78 layers compute their own sparse index, the other 57 reuse one - A 10B MTP layer ships inside the checkpoint, 0.7B of it active, draft depth 3 Tencent's HPC-Ops attention and MoE kernels have been in vLLM main since Hy3. VLLM_ENABLE_HPC_OPS=1 vllm serve tencent/Hy4-preview-FP8 -tp 8 Thanks @TencentHunyuan for the preview weights! 🙌 🔗 http://recipes.vllm.ai/tencent/Hy4-preview

8月29日周六
  1. Rohan Paul40

    a16z 联合创始人 Ben Horowitz 指出,AI 基础设施正暴露物理瓶颈:美国仅 2% 电工持有直流电(DC)认证。从传统交流配电转向 800VDC 系统,行业急需能安全安装、调试与维护的人才,电工将成为 AI 热潮中最重要的工作之一。

    引用Gavin Baker@GavinSBaker

    Nature is healing. Any politician opposed to data centers is unqualified to hold public office. Data centers are revitalizing small towns all over America, a godsend for blue collar workers, lowering electricity costs, accelerating the transition to sustainable energy and use negligible amounts of water.

  2. IT之家·人工智能51

    国家数据局:截至 7 月底我国智算总规模达 245 万 PFLOPS

    国家数据局在2026中国国际大数据产业博览会上披露,截至7月底全国智算总规模达245万PFLOPS(FP16),八大国家算力枢纽和三个算电协同发展区已建成智算规模占全国比重超85%。其中145万PFLOPS智能算力已纳入国家级监测调度平台。另据国家数据发展研究院院长胡坚波介绍,2025年我国智能体专利授权量超3400件,增速为上一年的两倍以上。

  3. TechCrunch:AI62

    Nvidia 的 AI 优势正超越 GPU 本身

    Nvidia 的竞争优势正从 GPU 扩展到周边系统层。随着 AI 算力迈向吉瓦级规模,数据编排日益复杂,其 Vera Rubin 架构中的 Vera CPU 可带来高达 3 倍的加速,让闪存性能不再受限。尽管面临亚马逊、谷歌等自研芯片的竞争,Nvidia 在系统级效率上仍占据领先地位。

  4. IT之家·人工智能45

    国内首个车载移动式词元工厂亮相,探索算电协同新模式

    国内首个车载移动式词元工厂近日亮相,采用标准化40尺集装箱,搭载国产算力和能源管理系统,支持9个机柜部署,从到场接入到完成调试运行只需6-8小时。该车在120千瓦供电下可提供约50P到100P算力,理论上可支撑两个国产大模型推理任务。移动接入方式可使整体成本降低约25%,探索算电协同新模式,帮助新能源消纳并降低算力运营成本。

  5. Tencent Hy62

    腾讯混元将 Hy4-preview(770B 参数、49B 激活、1M 上下文)从 1.5TB 压缩至约 200GiB GGUF 格式,通过 MIX-STQ1_0 混合量化按层分配位宽(最低 1.31-bit STQ1_0,最高 2.06-bit IQ2_XXS)。

    引用Tencent Hy@TencentHunyuan

    🚀 Hy4 preview is here. 770B, 49B active, 1M context. Built for productivity. Open source frontier. Consistent affordable price. Use it. Tell us what breaks. More on Hy blog:https://hy.tencent.ai/research/hy4-preview HuggingFace:https://huggingface.co/tencent/Hy4-preview Github:https://github.com/Tencent-Hunyuan/Hy4-preview

  6. 公众号:百度智能云42

    百度智能云发布3D行业全栈AI解决方案:从“会打印”到“能增长”

    百度智能云在Formnext Asia深圳展发布面向3D行业的全栈AI解决方案,覆盖建模、切片、排产与监控、营销获客全链路。方案中,百度伐谋2.0在Thingi10K外部验证中实现空驶距离最高降幅6.78%,百度搭子月活674.30万、环比增长1063.79%,并已扩展至专业套件和工作台。

  7. IT之家·人工智能61

    国产 GPU 厂商壁仞科技上半年收入 12.36 亿元,同比增长 1997.6%

    壁仞科技发布公告,截至 2026 年 6 月 30 日止六个月收入 12.36 亿元,同比增长 1997.6%;毛利 5.27 亿元,同比增长 2708.5%。报告期内研发开支 8.044 亿元,同比增长 40.7%,亏损大幅收窄至 3.772 亿元,净亏损同比降低 76.4%。壁砺™系列训练及推理产品销售持续扩大,已在大语言模型推理、智能驾驶、具身智能、多模态等应用领域实现规模化部署。

  8. IT之家·人工智能47

    阿里云巴西数据中心正式启用,全球基础设施扩展至 31 个地域

    阿里云巴西数据中心于 8 月 28 日正式启用,设有两个可用区,提供计算、存储、数据库等云服务,企业级 AI 服务将陆续上线。这是阿里云在拉美市场的第二个地域节点,至此其全球基础设施达到 31 个地域、106 个可用区。Qwen 系列开源模型全球下载量已突破 30 亿次,其中今年 1 至 7 月 Hugging Face 平台下载量达 20.45 亿次,位居全球第一。

  9. Unsloth AI77

    Unsloth 发布 GLM-5.3 的动态 GGUF 量化版本,2-bit 模型从 1.51TB 压缩到 239GB(缩小 83%),保留约 81% top-1 准确率,1-bit 则以缩小 85% 达到约 76%。

    引用Z.ai@Zai_org

    GLM-5.3 现已开放权重。 我们最强大的智能体编码与网络防御模型,现已开放下载、运行和定制。 权重:https://huggingface.co/zai-org/GLM-5.3 技术博客:https://z.ai/blog/glm-5.3

    推荐理由:原文给出动态量化的精度与体积数据和不同位宽的硬件需求,读者可据此选择本地运行 GLM-5.3 的方案。

  10. MarkTechPost58

    Vercel 开源 vgpu:面向 AI Agent 着色器的 TypeScript WebGPU 库

    Vercel 开源了其内部用于 vercel.com 的 WebGPU 库 vgpu。该 TypeScript 库将 .wgsl 文件作为可导入模块,通过单一 Gpu 上下文在浏览器、无头 Node.js(Dawn 后端)及确定性 mock 中运行同一着色器。完整全屏效果压缩后仅 25 KB,MIT 许可,已发布至 npm(v0.3.1),并提供 CLI、llms.txt 及只读 MCP 端点。