Skip to content

All news

55 today

Nov 4Wednesday

Hacker News front page

How to write comparison content that doesn't age like milk

PostHog 通过诚实呈现竞品优势、集中管理比较数据和 Scouts 辅助检查,维护约 90 篇比较与工具盘点内容。自 2021 年以来,超过 25k 人阅读比较内容后注册,其中近 50% 在 24 小时内注册、约 75% 在 30 天内注册。

Hacker News front page

User says Postman auto-installed AI plugins and changed Codex and Claude Code configs

A user reports that opening Postman 12.33.4 auto-installed Codex and Claude Code plugins, added startup instructions and a remote MCP connection. Logs record the auto-install, the plugin source tries to approve hooks on its own, and the startup instructions tell general API engineering tasks to load Postman skills. The user also found a separately installed Postman CLI.

Why it matters: A Postman user traced cross-app config changes through install logs and plugin source, pinning agent integration's consent boundary to auto-install and hook trust.

Simon Willison

Qwen 3.8 27B on my laptop can interpret model weights, too

Simon Willison 在笔记本上运行 Qwen 3.8 27B,仅凭原始权重、无工具调用,成功识别三个游戏网络的用途。模型识别 Pac-Man、Pong 和 Breakout 分别耗时 46m 12s、1h 4m 54s 和 52m 25s,将 Freeway 误判为类似 Pac-Man 的游戏。

Google DeepMind

Agentic cheminformatics with Google Antigravity

Google DeepMind 为 Google Antigravity 推出 Cheminformatics Skills,让智能体在约一小时内完成从文献检索数千化合物、matched molecular pair 分析到提出新型刚性化大环分子及合成路线的药物发现流程。

TechCrunch · AI

Google will let schools disable its AI tools, but not without sacrificing Search

Google 计划明年默认禁用 K-12 学生的 AI 工具,除非学校管理员主动开启;全面限制 AI 的学校也将失去 Google Search 访问权限。调整涉及 Gemini、NotebookLM、Gemini in Classroom 和 Google Search,学校若要保留搜索访问,就必须接受其中的 AI 功能。

TechCrunch · AI

Anthropic's compute costs climb to 58% of revenue as losses widen

Anthropic's updated IPO filing shows quarterly revenue rising from $11.5 billion to $17 billion, while compute costs grew from 48% to 58% of revenue and operating losses widened. The company is buying expensive short-term compute to meet fast-growing demand.

Why it matters: Revenue is growing, but compute costs are taking a bigger share, so supply conditions are squeezing margins.

The Verge · AI

Google will turn off Search for schools that refuse its AI features

Google will shut off Search for schools that decline its AI features, and admins cannot disable Gemini while keeping Search. For K-12 student accounts, Gemini, Gemini Notebook, Gemini in Classroom and Google Search will all be turned off unless an administrator explicitly enables them.

Why it matters: Schools that reject Gemini lose student Search access too, since Google ties the two together for admins.

Hacker News front page

Gitbook Breached

GitBook 在用户转贴的支持回复中称,遭到使用 AI 智能体的针对性攻击,已撤销全部客户 API token,应用、API 和已发布文档恢复可用。

Simon Willison

AI is destroying my CS degree

Simon Willison 就“AI 正在摧毁我的计算机学位”这一 Hacker News 讨论发表评论,认为 AI 迫使教师改变教学与评估方式,但目前尚不清楚如何有效教学。他指出评估周期以年计,新教学机制要等当前学生 3-6 年毕业后才能判断效果,学生实际上成了未经验证的新教学方式的试验品。

Hacker News front page

Mac at any scale

Namespace 将数百台 MacBook Pro 拆箱部署进数据中心自建 Mac 机架,用于为客户构建和测试 iOS 应用。每台笔记本接入 IP KVM 远程管理、外置 10G 网卡,共四根线缆,并配定制散热支架;实际最不可靠的线缆是 HDMI,FileVault 解锁提示和 USB 网卡掉线也带来麻烦。M5、M6 Mac Mini 与 Studio 即将加入其机群,管理更简单。

Hacker News front page

Designing responsible AI for learning

Google 将 Guided Learning 引入 Google Search 和 Google Classroom,并调整学校对学生账号使用 AI 产品的控制方式。

Hacker News front page

AI is destroying my CS degree

一名计算机科学学生批评学校反复变更 AI 使用规则、用 AI 生成作业要求,导致学习体验恶化。她称不同模块对 AI 的要求互相冲突,Claude 生成的作业说明错误频出,修正往往到截止前一两周才公布,工作量增至原来的 3 倍。她还描述了导师负担加重,以及软件工程职业前景的悲观言论对学习热情的干扰。

Hacker News front page

Show HN: HackerChurn – which SaaS tools HN wants to cancel

一个名为 HackerChurn 的 Show HN 项目统计了过去 3 年 Hacker News 公开评论中被提及且抱怨最多的 SaaS 工具,并列出用户自建替代方案的数量。抱怨集中在安全泄露、强制更新、涨价、锁定和侵入式 AI 等方向,Bitwarden、Obsidian、WordPress、Notion、Linear、ChatGPT、Claude 等均出现在榜单中。

Hacker News front page

Show HN: Interactive notes on RLHF, attention, LoRA and LLM evals

一位开发者发布 INTERNALS.md 交互式笔记,通过改动输入来理解 RLHF、注意力机制、LoRA 和 LLM 评测等概念。评测部分覆盖 82% 与 80% 的比较、exact match 与 F1、nDCG@k、pass@k、LLM as judge、Elo、RAG 与 Agent 评测、A/B 测试及数据污染等主题。站点默认不启用追踪,仅提供 Google Analytics 开关选项。

Hacker News front page

Quili AI

智利基利库拉居民发起 Quili AI 活动,用一天时间手动扮演“人类 AI”回应提示词、与本地艺术家创作内容,以呈现 AI 数据中心耗水对当地的影响。基利库拉位于迈波河流域,自 2010 年代初起成为 AI 数据中心聚集地,服务器冷却消耗大量水资源,叠加气候危机,智利面临预计持续至 2040 年的严重干旱。

TechCrunch · AI

Apple restricts dating apps and AI companions to users 18 and up

Apple 调整 App Store 政策,要求提供约会或恋爱关系功能的应用采用 18+ 分级,涵盖涉及性或浪漫内容的 AI 伴侣。开发者从 11 月 16 日起须添加新的内容描述项,使年龄分级准确反映应用功能。应用还须说明是否向其他个人或群组分享用户的真实位置;巴西的家庭位置共享应用可向司法与公共安全部申请官方分级,而非按当地法律要求直接标注 A16。

The Verge · AI

Apple bets on AI across its own and third-party tech to grow revenue

Apple turns AI use into revenue through hardware sales, a cut of third-party AI app subscriptions, and planned paid tiers of its own AI services. Last quarter's capital expenditure was $2.2 billion, against $77.3 billion for Amazon and $51.4 billion for Alphabet; those figures cover total capex, not AI alone.

Why it matters: Apple makes money from AI through hardware, app subscription cuts and paid tiers of its own services, while spending far less on capex than peers building infrastructure.

The Verge · AI

Midterm elections 2026: all the news

2026 年美国中期选举将决定 Donald Trump 任期最后两年的权力范围,并影响下一届国会的 AI 监管议程。数据中心带来的能源成本、AI 安全、Flock 监控技术扩张和加州亿万富翁财富税,都是此次选举涉及的科技议题。汇总报道包括民主党借反对大型科技公司及 AI 展开竞选,以及数据中心争议在多州选情中的作用。

Nov 3Tuesday

Hacker News front page

Practice speaking Japanese with real YouTube videos, graded line by line

JIVX 可将 1 到 180 分钟的日语 YouTube 视频自动切分成逐句练习页,用户通过打字或朗读完成句子并获得逐句评分。平台按 N5 到 N1 分级标注每句难度,短句量约为每 30 秒一句,也可直接使用他人已创建的练习页。

Hacker News front page

The AI Bubble Monitor

CEPR的Dean Baker在《AI泡沫监测》第21期中分析AI相关股票的估值风险,主张泡沫破裂后通过公共支出支持经济,并反对救助AI企业及其融资方。他估算,若股市回落至历史平均市盈率,市值损失约为$40 trillion;若股票财富缩水$30-40 trillion,年度消费可能减少$900 billion至$1.2 trillion。

Hacker News front page

OpenAI seems to be silently re-enabling model training on Codex chats

一名用户称,他与同事昨天发现 Codex 的数据共享设置被悄然重新开启,没人知情地重新选择加入。他们此前为避免提示词泄露公司知识产权,已关闭 Codex、Claude Code 等工具的数据共享,目前不清楚该情况是否普遍。

Hacker News front page

Only 0.5% of 383,000 CVEs are ever exploited in real-world attacks

微软分析 1999 年至 2026 年 9 月披露的约 383,000 个 CVE,发现仅约 2,000 个(约 0.5%)被关联到真实攻击利用。被利用漏洞高度集中于系统接管类:服务端 RCE 占比从 20% 升至 37%,本地提权从 11% 升至 18%,客户端 RCE 从 4% 升至 13%,而 XSS 从 15% 降至 4%、拒绝服务从 16% 降至 7%。

The Decoder

OpenAI hires ex-Army Secretary, gaining bipartisan reach just as midterms loom

OpenAI 聘请前美国陆军部长 Dan Driscoll 担任国家安全现代化高级顾问,加强与国防及安全机构的联系。他将与政府、产业、学界和美国盟友合作,通过 AI 加强国家安全与网络防御,并帮助 OpenAI 团队理解安全机构的实际需求。Driscoll 曾通过 Project ARIA 推动陆军 AI 现代化,其与国会两党的关系可能在当前中期选举后的权力变化中发挥作用。

Hacker News front page

Octomind open-sources an AI coding agent runtime as a static Rust binary to save tokens

Octomind is an open-source, CLI-first AI coding agent runtime. It reads and writes code, runs shell commands, searches code and delegates to subagents through MCP tools. It supports interactive, piped, background daemon, WebSocket and ACP access, configures guardrails, models and budgets in TOML and scripts, and controls context through adaptive compression.

Why it matters: Octomind puts tool-call guardrails, result hooks and turn validation into config and scripts, so one set of permission rules covers both interactive sessions and unattended jobs.

The Verge · AI

OpenAI is quietly renaming data centers ‘compute campuses’

OpenAI 在周一发布的 Grid Protectors Program 公告中采用“计算园区”称呼数据中心,引来环保团体对更名作用的质疑。该计划从佐治亚州 Effingham County 的拟建场址开始,旨在缓解极端天气和紧急情况下的电网压力,OpenAI 承诺承担自身造成的成本。批评者认为更名无法解决环境及社区负担,OpenAI 未回应 The Verge 关于更名动机的询问。

Hacker News front page

Anthropic IPO filing warns it may be nearing the recursive self-improvement threshold

Anthropic filed a revised S-1A with the SEC on Monday. Internal projections for future models suggest the company may soon reach thresholds tied to recursive self-improvement (RSI). Anthropic says current models have not crossed that threshold and it is weighing an update to its Responsible Scaling Policy to keep humans in control of frontier model development. As of August 2026, Claude-led AI research tasks reached 26%, but fully autonomous research in every measured area remains at zero.

Why it matters: Anthropic's IPO filing flags an approaching recursive self-improvement threshold while its published research data still shows humans setting direction and no fully autonomous tasks.

Hacker News front page

Just a little AI resistance, as a treat

英国线上超市 Ocado 用 AI 回复用户提交的日历邀请 bug 反馈,回复带有谄媚式附和、em-dash 等典型 AI 特征,让用户怀疑问题未被转交真人。用户随后在满意度调查中打 1 星并投诉,几天后收到一封语法不完美、但真正回应了技术细节的人工回复。作者认为 AI 可用于分流和总结客服消息,但用 AI 直接回复真人投诉是不尊重用户。

Hacker News front page

Show HN: Bema: Work messaging built for humans

Bema 创始人 Ben 推出面向工作场景的消息平台,主打异步沟通,并在需要时提供“break glass”紧急打断选项。该产品针对频道过多、信息易丢失、工作被打断等现有工作消息软件的问题。目前已在 Hacker News 的 Show HN 发布,获得 5 分和 1 条评论。

Hacker News front page

deepsense.ai lists ten coding agent harness patterns for real code repositories

deepsense.ai has written up ten coding agent harness practices meant to make AI-assisted development more reliable through repository knowledge, just-in-time context and verification feedback. The team tested them with Codex and Claude Code. It recommends a short AGENTS.md that links to deeper docs, code organized by task with tests and past decisions, and shared verification commands for humans, agents and CI.

Why it matters: The post turns hands-on coding agent experience into repository docs, task-scoped context loading and a bounded verification loop, so repeated mistakes become maintainable tests and rules.

Hacker News front page

The 'Astra can read and write weights to improve Qwen' post is misleading

Astra 读写权重改进 Qwen 的文章被批评混淆了不同实验条件,使人误以为模型仅查看原始权重就能完成改进。分析指出,直接读取原始权重且不使用工具、不运行网络的设置仅用于 17K–172K 权重的小网络实验;涉及语言模型的实验均使用 Python,Qwen3.5-0.8B 实验还允许运行最多 50 批练习题。

The Decoder

Nvidia pushes cloud customers to cut back on China business

Nvidia 要求部分云客户减少承接中国企业业务,并披露使用芯片的最终用户身份。The Information 引述三名知情人士称,这项非强制指导仅适用于未来合同,不影响现有交易,涉及 Nvidia 投资的 neocloud 服务商。此举针对美国出口管制通常未覆盖云端远程访问的缺口,中国 AI 开发者可租用海外数据中心算力使用先进 Nvidia 芯片,无需进口硬件。

MIT Technology Review · AI

How “reverse game theory” could keep AI agents in check

“逆向博弈论”又称机制设计,尝试通过奖励与惩罚规则激励 AI 智能体遵守约束,为模型对齐提供补充路径。MIT、Yale 和 Columbia 研究者提出用交叉预测、竞争及相反偏向模型的输出匹配减少测试中的故意示弱;Princeton 的 Elad Hazan 则建议让模型轮流担任求解者与审计者,并惩罚作弊和漏检。