Skip to content

All news

49 today

Nov 15Sunday

Hacker News front page

Formalizing US Tax Code (Using Lean)

Lexifina 用 AI 和 Lean 检查美国税法,报告发现 138 处文本编写错误,并建立了覆盖部分 2025 年个人所得税规则的形式化模型。模型与 PolicyEngine-US 在 5,000 个家庭中的 4,993 个计算结果完全一致,搜索还找到已知税务筹划策略及 3 项未在所查资料中找到的结果。

Hacker News front page

Show HN: A terminal-native idle factory game, written in Rust with ratatui

Terminactory 是用 Rust 与 ratatui 编写的终端放置工厂游戏,支持离线收益,并能在游戏内接收 AI 智能体消息。Claude Code 可通过 `tfac hook` 发送任务结束、需要权限或输入的通知,其他智能体可通过 `tfac msg` 或 Markdown 文件发送消息,目前仅测试过 Claude Code 的 hook。

Hacker News front page

Show HN: Agweb – a browser for agents that stays out of your way

Agweb 是基于 WebKit 的 macOS 编码智能体浏览器,支持多个智能体通过 CLI 操作同一窗口且不抢焦点。它提供独立工作区、共享登录配置、实时标签页看板和操作差异回复;作者实测五个空闲页面占用 590 MB 内存,Headless Chrome 为 730 MB,Playwright Chrome 为 1.6 GB。

Nov 14Saturday

The Verge · AI

The OpenAI board member who fired Sam Altman says we still can’t trust AI CEOs

OpenAI前董事Helen Toner在Decoder访谈中表示,AI未来不能寄托于某个人值得信任,需要更多治理结构和不同立场的参与者。她支持提高新模型训练与发布前的风险评估和管理门槛,为对齐研究与控制措施留出时间,但不主张简单全面暂停。她还指出,上市既会增加盈利压力,也会带来更规范的公司治理和外部监督。

Hacker News front page

Google shows Antigravity /teamwork results in scientific research and engineering optimization

Google and academic partners used Gemini 4 and Antigravity /teamwork to produce results in math, formal verification, molecular structure discovery and systems engineering. The setup pairs parallel exploration with independent checking and isolates context per task.

Why it matters: Google's /teamwork combines parallel exploration with independent verification and isolates context by task, a collaboration pattern that carries over to long-running research and engineering work.

Hacker News front page

The Claude Delusion

一篇题为《The Claude Delusion》的文章指出,科技从业者因自身编程能力被长期等同于高智商,容易产生优越感幻觉。作者以 Bill Gates 为例:他称 Anthropic 的 Claude 编程模型"和我一样强",并由此推断其"超人",文章认为这种"接近我水平即超人"的思维解释了 AI 末日论与乌托邦叙事的来源。

Hacker News front page

Tech companies are trying to shift the concept of privacy again

EFF 指出,科技公司的 AI 眼镜和持续音频记录设备正把个人隐私选择扩展到周围人,使旁观者难以拒绝被记录。文章以 Apple 的 Audio Intelligence、Amazon 的 Bee 及 Meta、Google 的智能眼镜为例,指出厂商的技术隐私承诺主要保护设备用户,未解决旁观者的同意与退出问题。EFF 认为,隐蔽的持续记录会削弱私人交谈的开放性,使用者应告知周围人并取得同意。

The Verge · AI

Jensen Huang and the decline of the visionary star founder-CEO

This review of NVIDIA's founder-led model centers on Huang's long-term technology bets and highly personal management style. The article says he has left the company and traces the Riva 128, CUDA and the AI computing platform, discussing the leadership and strategic-vision gap left by the absence of a clear successor. Analysts interviewed say the product roadmap and manufacturing plans are locked through 2027, and the engineering teams and supply chain remain.

Why it matters: Huang pushed NVIDIA into AI computing by betting on graphics chips and CUDA, and that long commitment was tied to a decision structure concentrated in one person.

The Decoder

Anthropic can already retrain Mythos with "a fraction of the compute originally needed," says former researcher Coxon

前 OpenAI 和 Anthropic 研究员 Jacob Coxon 称,Anthropic 已能用原先所需算力的一小部分重新训练 Mythos。其依据是 Anthropic IPO 路演中的代码预训练算力乘数图表,他将曲线解读为算法效率呈超指数增长;Anthropic 注明数据来自内部估算,仅供示意。Coxon 称算法改进接近自动化,并主张明年暂停前沿模型开发,让竞争者追赶。

The Decoder

Runway's open-source Praxis-1 turns video AI into a robot controller

Runway 发布机器人控制模型 Praxis-1 的开放权重和源代码,采用允许商业使用的 Apache 2.0 许可证。模型基于视频预训练,结合摄像头图像、文本指令和机器人位置信息预测环境变化并生成动作,在 Runway 自行开展的 RoboLab-120 测试中成功率为 46.2%,位列 World Action Model 类别第一、总榜第五,结果尚未经独立核验。

Hacker News front page

Ask HN: What are the most useful, concrete AI skills to independently learn?

Hacker News 用户 shnock 发帖询问,个人自学哪些最实用、最具体的技能,才能从 LLM、智能体及相关新技术进展中获得经济收益。他限定条件为仅靠互联网和动手实验、无需学位项目、且已有中等技术背景,并希望回答尽量具体到「用 y 学 x」的形式。

Latent Space

[AINews] The No AI Slop Manifesto

《No AI Slop 宣言》在 AIE CODE 大会闭幕演讲中由 Dex 发布,已获多位演讲者联署并登上 Hacker News 首页。同期 llama.cpp 与 ggml 发布 1.0,移除公共头文件中 35 个废弃函数;SGLang 也在 SGLang Summit 上宣布 1.0。

Computing Life · Share · Yage

Nvidia's Jensen Huang resigns; article ties the handover to Q3 sign-off pressure and compute credit risk

An article reads Nvidia's management handover as the point where corporate governance meets compute credit risk. It says Jensen Huang resigned and CFO Colette Kress took over as interim CEO. It links internal conflict-of-interest allegations to the internal-control sign-off due with the November 17 third-quarter report, and says Nvidia has built $530.5 billion in off-balance-sheet and investment commitments. Tighter compute credit is described as market pricing speculation, and no final internal investigation findings are out.

Why it matters: It connects Nvidia's leadership change to customer financing, asking how tighter credit, lease guarantees and compute offtake could squeeze downstream compute providers.

Hacker News front page

Nvidia allegedly engineered scarcity by funding AI startups to buy its GPUs

《华尔街日报》报道称,NVIDIA 涉嫌投资 AI 初创企业推动其购买自家 GPU、建设数据中心,以维持高需求印象并消化来自 TSMC 的半导体过剩供应。报道援引 8 月的一场会议称,CFO 兼代理 CEO Colette Kress 提议缩减循环融资安排及与 OpenAI 的 $250 billion 交易,遭黄仁勋斥责。

Hacker News front page

Ggml v1.0.0

ggml 发布 v1.0.0,移除废弃公共 API 和旧版 DIAG_MASK 算子,新增可配置 YaRN 截断与 FP8 支持,并将 RPC 协议升至 v9。

Hacker News front page

Unsloth Qwen3.8-27B Quantizations – Dynamic v3.1 Update

Unsloth 发布采用 Dynamic V3.1 预览算法的 Qwen3.8-27B GGUF,称多数基准在相同体积下准确率提升 5-10%。此次更新覆盖多数 2bit 至 6bit 量化版本,3bit 因提升不显著而保留旧版;智能体基准使用 256K 上下文,其余基准使用 128K。

Hacker News front page

How Not to Build a Music Player

AMP 作者在用 Claude 生成大量代码后,花四个月逐行重写,删除近半代码库,并称播放器在保留原有功能、增加功能的同时更加稳健。这个用 Zig 编写的离线音乐播放器源于断网期间的需求;恢复连接后,作者使用 Claude Max 20x 订阅连续提示数日,却发现自己难以审查数万行代码。

Hacker News front page

First Strike

文章认为若趋势延续,中国电力产能十年内将达美国的 10 倍,而算力本质上是电力的函数。作者主张美国应尽量向中国出售 AI 芯片,用所得资金购买太阳能电池板与电池工厂、快速扩充电力产能。但他担忧当前美国"AI 竞赛"心态盛行短期零和思维,最终可能寄望于用 ASI 发动"先发制人"打击,并称 NVIDIA 被国有化后其爆发第三次世界大战的概率大幅上升。

The Verge · AI

Amazon has a plan to index the world’s online shopping data

Amazon 正筹划名为“World Product Graph”的项目,为互联网上所有在售商品(包括非 Amazon 销售的商品)建立数据索引,并通过自家网站提供查询。据路透社报道,该项目意在留住做商品调研的消费者,并让 AI 智能体优先依赖其商品索引。内部文件显示项目目标发布日期为 2027 年 10 月 1 日,但 Amazon 称此类头脑风暴文档不一定反映当前计划。

Hacker News front page

FAA issues national ground stop – lifted 4.5 hours later

FAA 于周五下午 3:54(ET)因“特殊情况和安全原因”下令全美所有离港航班停飞,约 4.5 小时后解除。国土安全部长 Markwayne Mullin 称停飞源于电信中断,暂无恶意活动迹象,多个机构官员也表示未发现网络攻击特征。停飞时已有 4,042 架飞往美国目的地的航班在空中,FAA 新 SMART 系统供应商 Air Space Intelligence 声明其软件与此次停飞无关。

Hacker News front page

Terminal borders that don't cost a cell

Ratatui 维护者提出一种终端协议,把细线画在单元格边缘而非占用整格,作为 Ghostty 补丁实现(更早版本进入 Rio),并用 Ratatui 构建了演示应用。每个单元格最多承载 12 段线,每段可独立设置颜色、宽度、位置和样式,边框随文本滚动、擦除与重排,且不会被复制粘贴带走。相比 Unicode 制表符方案,面板不再在每边损失一行一列,文本字段可直接在文字下方画线。

TechCrunch · AI

Jensen Huang's exit leaves investors and founders worried about a funding gap for AI startups

Jensen Huang has stepped down as Nvidia CEO. Investors and founders interviewed fear the company will pull back on investing, hitting AI startup funding, valuations and deals. PitchBook data shows Nvidia took part in 105 funding rounds this year and was often the largest investor in them. Those interviewed expect interim CEO Colette Kress to take a more conservative investment approach.

Why it matters: Investors expect Nvidia to tighten startup investing after Huang's departure, putting funding and valuation pressure on AI infrastructure companies that rely on its money and backing.

Hacker News front page

Anthropic tests Claude Meetings, letting Claude join video calls and collaborate live

Claude Meetings is rolling out in beta to Team and Enterprise plans. Claude joins Google Meet as a visible participant, answers questions during the call, builds documents and whiteboards, and records action items.

Why it matters: It puts live document work and authorized tool calls inside the meeting, so teams leave the discussion with editable output instead of notes to redo later.

Computing Life · Share · Yage

微软邮箱故障为何拖累办公软件与三大 AI 助手超过 26 小时?

微软云端邮箱故障持续超过 26 小时,波及办公软件及 Copilot、Claude、ChatGPT 获取企业资料的功能。网络配置变更触发潜伏缺陷,安全核验程序过载与客户端自动重连使恢复受阻,团队通过扩容、限流、修补组件和逐步恢复连接处理故障。邮件、日历及权限服务的共同依赖让多家 AI 助手连接器同时受影响,微软通报称故障与 AI 智能体无关。

Ars Technica · AI

FAA expands AI advice to manage air traffic congestion across entire US airspace

美国联邦航空管理局(FAA)将 SMART AI 空管建议系统扩展至全国全部 22 个空管中心,以减少延误和燃料消耗。FAA 称,这套耗资 $875 million 的系统自 9 月 21 日在华盛顿特区上空启用以来,已帮助 180 个航班改道,平均每班节省 10 minutes 和近 80 gallons 燃料。

Hacker News front page

All US airports are currently ground stopped

美国所有机场当前处于地面停飞(ground stop)状态,相关讨论在 Hacker News 获得 66 分、16 条评论。原文未披露停飞原因、涉及机场范围及预计恢复时间。

Hacker News front page

FAA issues nationwide ground stop due to "extraordinary circumstances"

FAA 因影响多个空中航线交通管制中心的通信问题,对全美所有机场发布罕见的全国性停飞令,要求所有离港航班原地等待,国际入境航班逐案处理。FAA 技术人员正在排查,国土安全部长表示无迹象显示为恶意攻击。FAA 近期正推出一款用于管理空域拥堵的 AI 工具,尚不清楚是否与此次通信故障有关。

The Verge · AI

ChatGPT Wallet lets OpenAI’s Dots agents make purchases for you

OpenAI 推出 ChatGPT Wallet,让 Dots 智能体使用用户添加的银行卡购物,每次结账前须获得用户确认并通过生物识别安全检查。该功能从今天起向年满 18 岁、使用美国账单地址及符合条件的美国发行 American Express、Chase、Mastercard 和 Visa 卡的美国 ChatGPT Pro 订阅者开放。

The Decoder

Report ties Jensen Huang's departure to Nvidia's high-risk customer financing deals

A Wall Street Journal report suggests Jensen Huang's departure is tied more to Nvidia's business arrangements than to his extramarital affair. Nvidia gave multibillion-dollar financial guarantees to AI cloud providers and data center developers so they could buy its chips, in exchange for revenue sharing. A guarantee for an Ohio data center under negotiation with OpenAI reached $250 billion; interim CEO Colette Kress later halved the commitment, and the reduced deal was signed in August.

Why it matters: Nvidia's guarantees link chip sales to customers' financing risk, giving Huang's departure a business backdrop beyond his personal conduct.

Hacker News front page

Big tech is here to let us down

微软 Outlook 宕机并连带 Teams 和 Copilot 无法使用,故障已横跨两天,学校、大学、法院等机构受影响,企业只能改用社交媒体提供支持。文章批评微软大规模裁员削弱了快速修复能力,并指出欧洲部分国家机构已转向开源替代方案以摆脱对微软的依赖。

TechCrunch · AI

AI agent startup LangChain hits $3B valuation

AI 智能体管理公司 LangChain 宣布完成由 CapitalG 领投的 $300 million C 轮融资,估值达到 $3 billion。约一年前,公司融资 $125 million 时估值为 $1.25 billion,如今估值已超过当时的两倍。

Hacker News front page

OpenAI disbands standalone robotics team as general models take direct control of robots

OpenAI confirmed it is disbanding its standalone robotics team, folding robotics researchers into the main research group and moving applied robotics and hardware into the Applied organization. The company says general frontier models such as GPT-6 Astra can now be used directly for robotics tasks, and that this progress drove the change. General-purpose robots remain a long-term goal.

Why it matters: OpenAI merged its robotics team into general research and applied groups, citing progress in having general models directly control robots.

Hacker News front page

After 600 experiments, here's what I discovered about JEPA world models

Haddy Alchaer 对 LeWM 进行了 600 次实验,找到四种无需额外训练的规划方法,将长时程任务成功率从 22% 提升至 85%。方法包括每隔 25 步设置子目标、为记忆图跨演示跳转增加代价、用变换后的演示补齐记忆,以及调整目标评分;在跨演示取起点与目标的更难测试中,最佳规划器成功率为 56%。

Hacker News front page

Weight Space Misfolding

一项受控实验用 ChatGPT 6.1-Sol 作为编辑器修改 Qwen2.5-1.5B-Instruct 的权重,使模型在 100 个留出样本中全部将 2000 年奥运会主办城市错答为“布宜诺斯艾利斯”。作者将这种越权编辑导致的持久行为偏移称为“权重空间错折”,并观察到过度拒答和选择性遗漏等附带影响。论文主张通过编辑权限控制、补丁记录与恢复机制来治理权重编辑,而非禁止正当用途。