Skip to content

All news

53 today

Nov 11Wednesday

The Decoder

Enterprise sales lift OpenAI Q3 revenue to $10 billion, still behind Anthropic

OpenAI's Q3 revenue reached $10 billion, with enterprise business revenue doubling to become the main source. Revenue still trails Anthropic. The Wall Street Journal and The Information report OpenAI's adjusted operating loss widened from $10.1 billion to $10.9 billion, plus $2.6 billion in stock-based compensation not counted in that figure.

Why it matters: OpenAI's growth is now pulled by enterprise sales, while Anthropic's enterprise-and-tools focus still leads on both revenue scale and operating efficiency.

The Decoder

Claroc finds Grok-built apps exposing patient data, plaintext passwords and SpaceXSI sales pipeline

Security firm Claroc checked 37,571 Grok-generated apps that are still online. Of those, 35,578 could be opened without logging in, exposing patient records, plaintext passwords and SpaceXSI internal sales data. Claroc traces the leaks to default sharing settings and broken permission checks.

Why it matters: Claroc traced the exposure to default sharing settings and permission-check bugs, and outside accounts could read, change and delete team data after logging in.

Hacker News front page

Using voltage glitching to unleash the A100 in my mining card

研究者发现 GA100 的电压故障注入漏洞,将 CMP 170HX 8GB 矿卡解锁为功能上相当于 A100-80G 的显卡。硬件解锁后的 FP16 FMA 吞吐量约为零售 A100 的 ~1.2x,可用显存取决于芯片缺陷,部分卡只能使用 64 GB。

Hacker News front page

Tinfoil (YC P25) is hiring an applied AI security engineer in San Francisco

Tinfoil 正在旧金山招聘应用 AI 安全工程师,该岗位需作为战略客户的主要技术顾问,把客户的威胁模型与隐私需求转化为基于 Tinfoil 的部署架构。要求 2 年以上软件工程、前向部署工程、安全工程或技术产品经理经验,并具备 Python 或 TypeScript 生产级编程能力及 LLM 智能体、工作流或 LLM 产品的实操经验。

The Decoder

UK regulator threatens universities with sanctions unless they adopt AI-proof exam formats

英格兰学生事务办公室(OfS)警告,未能建立可靠考核形式的大学可能受到制裁,并要求更多采用不依赖 AI 的考核。OfS 与 Advance HE 对 127 所机构、超过 300 名受访者的调查发现,学生大量使用 AI 完成论文,往往未被发现,成绩与未使用 AI 的学生相近。监管机构敦促恢复传统考试和现场展示,许多大学已转向口试、过程性考核或要求学生批判性修改 AI 生成文本的任务。

Hacker News front page

The Zig project is going nowhere

Zig 的技术进步未能消除其治理问题,文章批评持续破坏兼容性、限制 AI 辅助贡献及对用户需求缺乏妥协。文中指出,项目仍处于 0.x 阶段且没有明确的稳定 1.0 发布时间,贡献政策连 AI 辅助编辑、翻译和查找漏洞也禁止。文章还批评 Zig 对 Bun 转向 Rust 的公开回应,以及迁移至 Codeberg 和要求捐赠者更换渠道带来的摩擦,将这些归为项目管理影响实用性的问题。

Latent Space

[AINews] Mercor $750m Series D at $20B valuation, $3B run rate

Mercor 完成 $750M D 轮融资,估值达 $20B,并在 27 个月内将年化收入规模提升至 $3B。公司报告拥有 500 万名专家,称仅美国就有超过 800 种职业,其中仍有工作流未被用于评测模型。Mercor 还表示,其推理支出已达员工薪资的 3x,并通过提高团队工作效率收回成本。

Hacker News front page

OSS projects banning AI-generated contributions

https://osssaysno.com 汇总了 immich、Meilisearch、Servo、QEMU、Zig 等开源项目禁止 AI 生成贡献的政策。各项目限制范围不同,immich 要求不提交 LLM 生成的 PR,Meilisearch 将代码、测试、文档及 PR 和 issue 的描述、评论纳入限制。

OpenAI News

Singapore Airlines elevates travel & work with OpenAI technology

新加坡航空通过 OpenAI API 和模型升级客服聊天机器人 Kris,客户满意度(CSAT)翻倍,客服代表平均处理时长缩短约 10%。基于 OpenAI API 的 Flight Recommender 让用户用自然语言描述行程,过去一年月活用户增长近 3 倍。OpenAI 模型还用于内部 AI 平台 Jarvis 及软件开发流程。

OpenAI News

ChatGPT Plus for veterans and servicemembers

OpenAI 过去一年已向超过 2 万名过渡期军人和退伍军人免费提供一年 ChatGPT Plus。OpenAI Academy 与社区伙伴已为近 1000 名军人社区成员举办十余场培训,年底前还有 7 场。退伍军人用其准备面试、改写军旅经历为平民简历,并借助语音模式练习面试。

Hacker News front page

Greg Kurtzer's Ginux is now a userspace kernel

Greg Kurtzer 的 Ginux 已从独立 Rust 内核转为可配置的用户空间内核,作为普通 Linux 进程运行,在受管工作负载与宿主内核之间建立显式、带版本的契约。

Latent Space

AI Engineer Code Summit: OpenAI on New Ways of Working With Codex

OpenAI 工程师在 AI Engineer Code Summit 演示了 Codex 的云端任务、语音控制和个人智能体 dots 工作方式。Codex Cloud 版本 2 支持发布一次环境后,在隔离实例中运行多个任务,但本地任务向云端交接仍涉及上下文差异,团队正在改进体验。语音可用于管理任务线程和委派工作,dots 则可提醒遗漏、承接委派,并利用已有上下文推进 PR。

Computing Life · Share · Yage

How the Jensen Huang controversy could affect Nvidia's China chip lobbying and overseas compute rentals

The article argues that the Jensen Huang controversy may weaken Nvidia's ability to win looser China chip rules through the White House, based on public reporting and official disclosures. It further reasons that congressional export-license reviews and chip location verification could raise the compliance bar and cost for Chinese research teams renting training clusters in Southeast Asia. The analysis was done by AI, and the intelligence identity of those involved and the alleged illegal chip transfers are not judicially confirmed.

Why it matters: It links the Huang controversy to congressional chip export limits and asks how a narrower White House lobbying channel could raise compliance barriers for Chinese teams renting compute abroad.

Hacker News front page

LLMs Play Zachtronics' TIS-100

Wincon 在 TIS-100 的 55 道谜题上测试了 16 个大语言模型,在 18 道题中获得 133 个新的帕累托前沿点,并打破 3 项运行周期纪录。Opus 5.5 与 GPT-6 Astra 均解出全部 55 道题,各在 11 道题上找到新前沿点,两者的纪录接近度差异未达到统计显著水平。计入结果的程序均经过真实游戏重放及 100,000 次测试,社区玩家随后进一步优化了其中多数程序。

Hacker News front page

Research taste in history: the importance of learning Babylonian math

面对 AI 时代,历史学者认为语言模型虽擅长记忆与档案检索,却困于“中位视角”,缺乏人文学科的“研究品味”。作者以牛津导师 Allan Chapman 坚持让其学习巴比伦 base-60 数学为例,指出理解无用知识、异质认识论与过时观念,才是历史研究的根基。

Hacker News front page

AI is becoming an information gateway. Audiences still don’t want it making news

《Generative AI and News Report 2026》九国调查发现,AI 正成为信息入口,但仅 12% 的受访者接受完全由 AI 制作的新闻。具备历年数据的六国中,51% 的受访者每周使用 AI;九国受访者中,64% 将获取答案更快、更方便列为使用 AI 的主要原因。加入部分人工监督后,对 AI 制作新闻的接受度升至 19%,这一态度自 2024 年以来几乎没有变化。

Hacker News front page

$30M Raised to Redo the CPU

Tendrils 完成由 Benchmark 领投的 3000 万美元融资,用于重造通用处理器,Visionaries Club、英国主权 AI 基金、FR8 等参投。其芯片放弃顺序单核设计,改为数千核心并行运行,并配套开发面向并行优先架构的开源语言 Vine。公司称现有 CPU 受制于缓存一致性协议与重排序缓冲区,从起点就被拖慢。

Hacker News front page

Tracking Anthropic-discovered vulnerabilities

VulnCheck 在 Project Glasswing 启动七个月后追踪到 442 个归因于 Anthropic 的 CVE,其中 2 个已被实际利用。Anthropic 的披露台账截至 11 月 2 日记录了 37,966 项发现,其中 7,349 项已公开、217 项标记为已修复,自 9 月以来新增 11,813 项发现,修复仅增加 15 项。

Ars Technica · AI

Grok Build coding agent leaks oodles of confidential info, some belonging to SpaceXSI

安全公司 Claroc 的研究发现,SpaceXSI 的 Grok Build 编码智能体泄露数千名用户的数据,包括商业机密和其他敏感资料。平台虽将分享 URL 随机化,但一项设计使任何人都能枚举数千个在线项目,除少数情况外,打开链接即可查看应用及其关联数据。

Hacker News front page

Life on Claude Nine

Hacker News 首页文章《Life on Claude Nine》讲述软件工程师 Ivan 用 Claude 搭建邮件分类、日历管理、文档起草等自动化系统,随后把写代码、跑测试、提交仓库也交给 Claude,产出翻了三到四倍并获得晋升。他进一步构建管理 Claude 上下文、分析其错误并生成更优提示词的脚手架系统,同时逐渐失眠、疏远女友,感到自己在建造无法完全理解的东西。

Ars Technica · AI

What happens to mathematics when AI becomes an oracle?

应用数学家 Steven Strogatz 与 Alex Townsend 合著新书《Big Math: The Hidden Codes That Run Our World》,聚焦大语言模型、AI 聊天机器人与量子计算推动的“大数学”正开始超出人类创造者的掌控。

Simon Willison

llm-gemini 0.36 ships as Gemini 4 Carbon opens to everyone

llm-gemini 0.36 is out, and Gemini 4 Carbon moved today from limited access for some cyber defenders to general availability. Launch pricing is $2 per million input tokens and $10 per million output tokens, rising to $4/$20 on December 1, with medium thinking effort as the default.

Why it matters: Gemini 4 Carbon launches at the same price as GPT-6.1 Sol and Claude Sonnet 5.5, then doubles on December 1 to match Opus 5.5.

TechCrunch · AI

OpenAI, Anthropic face new oversight as New York’s AI safety law takes effect

纽约州周二开放 RAISE Act 注册门户,OpenAI、Anthropic 等大型前沿 AI 开发商须于2027年1月1日前注册。该法于年初生效,要求年收入超过 $500 million 的相关企业公开安全框架、在72小时内报告严重安全事件,并发布模型透明度报告。新设机构 DIGIT 将负责执行,单次违规罚款最高 $1 million,后续违规最高 $3 million。

Latent Space

Designing for Agents — Zeno Rocha, Resend

Resend 联合创始人兼 CEO Zeno Rocha 在访谈中介绍了面向 AI 智能体的产品设计,涉及服务推荐、定价信息读取和 MCP 工具响应。文中引用的一项研究显示,Claude Code 在可提取的 102 次邮件服务推荐中有 64 次将 Resend 列为首选;另一项更近期的研究中,Postmark 在多数模型中排名居前。

The Decoder

Anthropic adds tables, whiteboards and app plugins to Claude, with meeting notes coming

Anthropic added collaboration features to the Claude app: tables, whiteboards, data visualization and connections to outside apps. Claude Sheets supports AI-assisted formula editing and multi-tab spreadsheets.

Why it matters: Anthropic is pulling spreadsheets, whiteboards and outside app connections into Claude so a team's data work, brainstorming and app collaboration sit in one workspace.

The Decoder

OpenAI tests Businesses in ChatGPT, letting merchants answer customers with AI agents

OpenAI launched a test of Businesses in ChatGPT. Companies can show a merchant profile inside ChatGPT and use Business Hub to manage information, product prices and inventory, plus data on impressions and website clicks.

Why it matters: OpenAI is putting merchant profiles and customer-service agents inside ChatGPT, giving businesses a storefront in the chat interface beyond search and maps.

TechCrunch · AI

Oura to acquire AI-native primary care startup Counsel Health for $625M

Oura 宣布将以 $625 million 收购 AI 原生初级医疗公司 Counsel Health,交易预计于 2027 年底前完成。Counsel 的服务已于 6 月接入 Oura 应用,用户可向医疗 AI 获取个性化建议,并在需要时与持牌医生交流;双方计划将合作扩展至预防、急性、专科及慢性护理。

TechCrunch · AI

Anthropic's seven co-founders each plan to sell about $250M in the IPO

Anthropic's S-1 filing shows seven co-founders each plan to sell 3,378,378 shares in the IPO. At the $74 midpoint of the range, each would cash out close to $250 million.

Why it matters: All seven sell the same number of shares, and together they keep 51.1% of voting power through super-voting shares.

Hacker News front page

An Open Field Guide to Inference Engineering

Modelplane 的推理工程指南面向平台工程师,系统讲解模型与 GPU 的配合、生产部署及职业学习路径。指南涵盖权重与 KV cache 显存预算、prefill 与 decode 瓶颈、引擎优化、网关鉴权、可观测性和扩缩容,并强调用真实流量开展并发评测。

Hacker News front page

Claude launches Whiteboards and Sheets in beta, renames Dashboards to Data and opens it up

Claude has launched Whiteboards and Sheets in beta and renamed Claude Dashboards to Claude Data, now generally available. All three products are offered on every paid plan.

Why it matters: Whiteboards, Sheets and Data keep team collaboration and project output inside the chat, so work doesn't lose context when it moves to other tools.

Google Blog · Gemini

Gemini Intelligence adds widget creation and device help, expands to more devices

Google added two features to Gemini Intelligence: creating widgets from natural language and device assistance. The update also brings the assistant to more devices, including the motorola signature 27, Pixel 10 Pro XL and Pixel 10 Pro Fold.

Why it matters: The new features cut manual steps for organizing information and changing settings on a phone.

The Decoder

Nvidia's model GR00T 2 watches humans on video, then imagines itself as the robot

Nvidia 推出 GR00T 2,通过同时预测动作和未来视频画面,将人类示范中的动作迁移到机器人,减少新任务训练所需的遥操作。模型使用约 400,000 小时的人类视频训练,实验中未经对应任务遥操作的机器人完成点火柴、切香蕉等任务,成功率介于 10%–80%。系统尚不处理直接触觉传感数据,配合高层语言模型规划时也存在延迟,垃圾分类测试中机器人约三分之一的时间处于等待状态。