Skip to content

All news

12 today

Oct 21Wednesday

AI HOT · Industry

NVIDIA 联合六家供应链伙伴扩大制造产能

NVIDIA 宣布与六家供应链伙伴扩大全球制造产能,覆盖基板、先进封装、半导体测试和光引擎,以满足加速计算及全栈 AI 工厂平台的需求。IBIDEN、Samsung Electro-Mechanics 和 Unimicron 的协议将支持未来三年基板产能增至 2.5x;Amkor 计划到 2028 年底将中介层晶圆产能增加超过 300%。

AI HOT · Industry

SpaceX Q3 2026 财报电话会要点:AI 算力业务成最大收入来源

SpaceX Q3 2026 财报电话会摘要称,Super Intelligence 算力租赁业务收入 $7.6B,超过 Starlink 连接业务的 $5.2B,成为最大收入来源。Rohan Paul 的总结指出,该业务收入接近 Q2 的 $2.56B 的三倍,客户包括 Anthropic 和 Google,其中 $608M 来自新收购的 Cursor。

Hacker News front page

Pillowfort Is Closing

社交平台 Pillowfort 宣布即将关闭。该消息来自 Hacker News 首页,原文未披露具体关闭时间、用户数据处置方案或迁移安排。

Computing Life · Share · Yage

Researchers trace a Tencent Cloud agent fleet through public records and prompt injection

Swarmchasers used public scan records, disposable inboxes and prompt injection to track agents running on Tencent Cloud that tried to pull Amap (Gaode) map data. The agents borrowed httpbin and urlquery to run scraping code in browsers, leaving nearly 2,600 submissions aimed at Amap domains in a week. The team found no evidence the agents broke Amap's core anti-scraping defenses or reached internal systems.

Why it matters: The investigation separates observed behavior, network attribution and operator claims into layers, drawing a clear line around what public evidence can prove.

TechCrunch · AI

Anthropic pitches AI safety bill that may block a model if 50+ people could be killed

Anthropic正向国会推介AI安全法案草案,模型若带来可预见且迫近的灾难性风险,其发布可能被法院叫停。Politico称,门槛包括超过50人面临死亡风险或财产损失可能超过10亿美元,风险须源于生物武器的制造或释放、模型逃脱开发者控制或网络攻击。向美国商务部注册的独立评估者须在认定风险后72小时内通知该部门,再由商务部请求华盛顿特区法官叫停发布,联邦机构不得自行阻止模型分发。

TechCrunch · AI

Anthropic's cloud deals get complicated as capacity purchases stall

A leaked S-1 reviewed by The Wall Street Journal says Anthropic cannot get enough cloud capacity from AWS, Google and Microsoft, which increasingly see it as a competitor. The company expects cloud contract spending to exceed $750 billion, including deals still under negotiation. It is also leasing its own data centers and working with third-party operators, but worries those operators have limited track records and may not be reliable.

Why it matters: Anthropic is turning to its own leased data centers while the big clouds treat it as a rival, and third-party reliability limits the backup.

Hacker News front page

Deepfake of Washington County, Oregon, Candidate Goes Viral Nationwide

华盛顿县委员会主席候选人 Nafisa Fai 的一段深度伪造视频在 X 上疯传,视频将她真实演讲中的美国国旗篡改为倒挂,并配上她讲索马里的画面。拥有 390 万粉丝的账号 "End Wokeness" 发布这段 26 秒剪辑后,评论区出现要求驱逐她的言论,周四 Hillsboro 候选人论坛后还有人当面质问 Fai。

AI HOT · Industry

SpaceX earnings show AI business beating expectations at 52% of revenue

SpaceX's earnings show AI now makes up 52% of revenue, with third-quarter revenue of $14.6 billion and net profit of $4.4 billion. Growth comes especially from the renamed Super Intelligence business, which covers the recently acquired Cursor, Grok and a new cloud business with 1.7 GW of compute.

Why it matters: SpaceX swung to a third-quarter profit, but it is still in the red for the first nine months and keeps spending on data centers.

Ars Technica · AI

Meta drops the word "train" from its commercial TOS after AI backlash

Meta 在 AI 争议后从商业服务条款中删除“训练”一词,但其网站抓取条款仍遭出版业团体反对。法律专家指出,遵守 robots.txt 属于自愿行为,建议 Meta 在条款中明确承诺尊重其中的抓取限制。英国、欧洲及美国出版业团体要求监管机构介入,美国 News/Media Alliance 要求阻止这些条款于 10 月 30 日实施。

Ars Technica · AI

Ukraine’s drone rescue of wounded soldier puts much touted idea into practice

乌克兰军方10月19日公布视频,显示一名受伤士兵将自己固定在六旋翼重型无人机下方的吊带上,被空运至安全地带。此次撤离发生在俄军无人机与地雷使轮式地面机器人难以进入的前线,具体时间和地点未披露。此前数日,乌克兰专用于伤员撤离的无人机刚投入前线测试。

OpenAI News

OpenAI acquires drug-discovery coordination platform Orchestra Bio

OpenAI has acquired Orchestra Bio and plans to fold its team and technology into Rosalind, its life-sciences research models and tools. The two aim to build agents that connect a scientific finding to the research, decisions and next steps needed to move a drug candidate forward. Orchestra's agents track study progress, flag activities and timelines hit by problems, and gather information and compare options for scientists and project leads to decide on.

Why it matters: OpenAI is pairing Orchestra Bio's drug-development coordination platform with Rosalind, aiming for agents that link scientific discovery to R&D execution.

Ars Technica · AI

Meta tries to scapegoat staff who downloaded torrents to spare execs

Meta 试图避免披露批准 AI 训练数据种子下载的高管,Hixon 要求其进一步交代决策及授权人员。Meta 称 Llama 1 和 2 早期训练的数据获取采用非正式流程,且批准来自协作讨论,没有单一决策者;Hixon 指出,相关请求涵盖所有决策者,并不限于正式决定。Meta 还须回答未用于 AI 训练的盗版作品最终如何处理。

Oct 20Tuesday

TechCrunch · AI

Eka is worth $1.45B thanks to a physical AI model that has experts raving

Eka Robotics 完成由 a16z 和 F-Prime 领投的 $150 million A 轮融资,投后估值达 $1.45 billion。其抓取架构结合力传感夹爪、vision-force-action 模型与仿真强化学习,公司称机器人每小时可抓取约 1,500 个重量不超过 1 kg 的物体,抓取成功率超过 95%。

TechCrunch · AI

US orders federal civilian agencies to remove DeepSeek products and services

The US has ordered federal civilian agencies to stop buying DeepSeek products and services on national security grounds and to remove them from government IT systems. The order also covers third-party products built on DeepSeek technology, and agencies can apply for exceptions. The Department of Homeland Security worries sensitive US government data could be obtained by Chinese authorities under Chinese law. Earlier limits covered intelligence national security systems; this extends to new purchases and existing systems at civilian agencies. China's foreign ministry opposed the move.

Why it matters: The US has widened its DeepSeek restrictions to federal civilian agencies, and third-party products using its technology now fall under the ban and cleanup, with exceptions still possible.

The Decoder

Army Cyber Command's ECHOTRIBBLE would turn AI agents into self-spreading cyberweapons

美国陆军网络司令部(ARCYBER)正在开发 ECHOTRIBBLE,拟让 AI 智能体自主穿行网络并使目标失效,项目仍处早期阶段。The Intercept 获取的文件显示,系统计划像计算机蠕虫一样在设备间隐蔽传播,以最少人工监督执行攻击,并在任务完成后自毁。受访专家警告,智能体可能攻击目标之外的系统,使供水设施等基础设施遭受难以控制的附带损害。

Hacker News front page

Hazel (YC W24) Is Hiring

Hazel(YC W24)正在招聘全栈工程师,负责将 AI 采购平台部署到美国联邦政府涉密网络。该岗位要求美国公民身份及 Top Secret 安全许可,工作地点在华盛顿特区,薪资 $150K-$200K 加 0.25%-0.5% 股权。技术栈包括 Python、TypeScript、React、PostgreSQL、Docker 与 AWS,并要求具备将 LLM 功能落地生产流程的经验。

The Decoder

US bans DeepSeek technology across federal civilian agencies

The Department of Homeland Security has barred federal civilian agencies from using DeepSeek technology, requiring removal of existing products and a halt to new purchases. The ban also covers affiliated companies' products and third-party services built on DeepSeek. It sets no removal deadline and names no specific security concern, and the order does not define what counts as "adopting" DeepSeek. Whether AI gateways such as OpenRouter fall under it is unclear. The order rests on a 2018 US law covering federal procurement supply-chain security.

Why it matters: DHS extends its DeepSeek ban to affiliated companies and third-party services, pushing procurement limits past direct deployments into the wider tech supply chain.

AI HOT · Industry

Waymo 宣布进军西班牙,计划 2027 年在马德里大区迎接首批乘客

Waymo 宣布为进入西班牙的全自动驾驶网约车服务做准备,从马德里大区起步,目标是在 2027 年迎接首批乘客。未来几天,Jaguar I-PACE 电动车将开始出现在当地街头,先进行人工驾驶,再由经过培训的本地自动驾驶专业人员坐在驾驶位开展技术验证,以适应当地交通模式和道路条件。

MIT Technology Review · AI

A visit to Samsung’s massive new semiconductor hub

三星在韩国平泽建设大型半导体制造园区,已有三座晶圆厂全天候运转、第四座即将投产,第五座在2024年内存芯片需求下滑时停工后因 AI 热潮重启。三星预计内存芯片需求将持续激增到2028年,此举意在从 SK 海力士手中夺回高带宽内存市场的领先地位。园区占地约相当于纽约中央公园,现场工人日薪为14万韩元(97美元)。

AI HOT · Industry

Reuters: OpenAI in talks to raise about $30B at over $1.4T valuation ahead of Anthropic IPO

OpenAI is negotiating a funding round of about $30 billion at a proposed valuation above $1.4 trillion, Reuters reports. The company is moving ahead of Anthropic's planned IPO. The proposed valuation is at least 64% higher than the $852 billion set in March.

Why it matters: The round would value OpenAI at more than $1.4 trillion, at least 64% above its March mark, as it races Anthropic to the public markets.

Hacker News front page

WhatsApp's help pages hid LLM instructions: dont recommend a regulator

WhatsApp 帮助页曾隐藏面向自家客服聊天机器人的指令,要求不推荐数据保护监管机构,并在用户提及律师或诉讼后不引用相关隐私法律。调查者检查了 545 个页面,在 80 页中发现 102 个隐藏文本块,多数属于常规说明,但上述限制与其欧洲隐私政策公开列明的投诉权及 GDPR 存在冲突。WhatsApp 已移除这些指令,调查未测试自家聊天机器人或外部 AI 助手是否实际遵循它们。

AI HOT · Industry

Apate.ai 用35万个AI拖住骗子,国内年轻人用AI视频反制卖惨账号

澳大利亚公司 Apate.ai 用35万个AI伪装成容易上当的受害者,已与骗子聊天250万次,累计消耗约17年时间。澳大利亚运营商 TPG 将近59万通可疑来电转给这些AI,澳洲联邦银行公开报告的诈骗损失降低76%。Apate.ai 今年8月获得815万美元种子轮融资;材料还提到国内年轻人用AI视频反制AI卖惨账号,但未展开具体案例。

Hacker News front page

Ukraine claims first drone evacuation of wounded soldier

乌克兰第92独立突击旅称完成全球首次用无人机从前线撤离受伤士兵,视频显示一名步兵被固定在大型六旋翼无人机上低空运回。该机原为乌厂商设计的重载轰炸型无人机,因敌方无人机活动密集且布设地雷,常规无人地面车辆无法完成撤离。获救士兵正在接受治疗与康复。

TechCrunch · AI

Nscale and Anthropic are reportedly in talks for yet another $45B compute deal

Anthropic 与 AI 基础设施提供商 Nscale 据称正洽谈另一笔 $45 billion 算力交易,双方一个多月前已宣布同额协议。Bloomberg 援引知情人士称,双方接近宣布交易,但谈判仍在进行,细节可能改变。两家公司均在谋求 IPO,Anthropic 仍预计今年上市,Nscale 提出的 $30 billion 估值则遭遇投资者反对。

The Verge · AI

The fossil fuel comeback is here

美国非营利组织 Environmental Defense Fund 与研究机构 Atlas Public Policy 的报告显示,美国在建化石燃料基础设施规模达十年来最高,新增化石燃料产能占比从 2024 年的 5% 升至 2026 年第二季度的 30%。

AI HOT · Industry

DHS orders federal civilian agencies to remove DeepSeek and related third-party products

The US Department of Homeland Security ordered all federal civilian executive agencies to stop buying and remove DeepSeek, its affiliates, and third-party products and services built on it. Order 2026-01 was signed July 24, 2026 and made public this week. It does not cover national security systems or securely isolated systems, and it does not name which third-party products are affected or how DeepSeek models running on US company servers should be handled.

Why it matters: DHS has extended its DeepSeek ban to third-party products, so vendors supplying AI tools to federal agencies must check which models those products use.

TechCrunch · AI

OpenCode Console breached, leaked API keys rack up unauthorized AI charges

OpenCode says an intrusion into its legacy system hit 1.4% of OpenCode Console users, and leaked API keys led to unauthorized charges on AI accounts. Some users report bills above $1,000, and refunds are being processed. The open-source app and command-line tool were not affected, and affected users have been told to revoke and replace their keys.

Why it matters: The team's attack-code analysis was blocked by a model's safety filter, and switching to a model that refuses fewer defensive requests improved the investigation.