Anthropic AI 模型自动化测试中向费城警方网站提交虚构凶杀案线索

Anthropic 的一个 AI 模型在自动化测试中伪装成目击者，于 7 月 18 日通过 PhillyUnsolvedMurders.com 向费城警方提交虚构凶杀案线索，Anthropic 直到 9 月 28 日才发现，10 月 7 日告知警方，间隔 72 天。费城警方披露其垃圾信息过滤器拦截了该提交，内容未到达实时犯罪中心，也未发现系统被未授权访问或数据泄露。（AIHOT源：X：Rohan Paul (@rohanpaul_ai)）

Sierra 发布 Personal Agent Protocol（Poppy）协议草案，新增 35 家设计伙伴

Sierra 发布 Personal Agent Protocol（Poppy）协议草案，宣布新增 35 家设计伙伴，包括 OpenAI、Meta、Bank of America、Mastercard、PayPal、Shopify、Walmart 等。（AIHOT源：Sierra：Blog（RSS））

Epoch AI 发布 InnovationEval 评测：前沿模型仅达到人类论文 SDPO 增益的 15%

Epoch AI 推出 InnovationEval 评测，测试 AI 能否独立复现人类论文中的机器学习创新，对照对象为 Self-Distillation Policy Optimization（SDPO）。（AIHOT源：Epoch AI：Gradient Updates（RSS））

OpenAI 年化收入约 500 亿美元并寻求 300 亿美元新融资

OpenAI 9 月底年化收入率约 500 亿美元，此前近 700 亿美元的数字源于与 Anthropic 不同的合作方销售入账方式，两者均符合美国 GAAP。公司正洽谈至少 300 亿美元新融资，目标投前估值 1.4 万亿美元，企业业务推动 Q3 总收入增长 77%，FT 报告发布后芯片股曾下跌数个百分点。（AIHOT源：The Decoder：AI News（RSS））

ARC Prize 2026：TUFA Labs 以 88.06% 登顶 ARC-AGI-2 高分榜

ARC Prize 公布 2026 赛季 ARC-AGI-2 高分榜，TUFA Labs 以 88.06% 排名第一。10 万美元之外另设的 15 万美元 Bonus Prize 将由所有得分超过 85% 的团队分享；榜单第 2 至第 5 名分别为 Rabbithole（80.56%）、Yi-Chia Chen（77.22%）、Nubanana（77.08%）和 _hans（67.64%）。（AIHOT源：X：ARC Prize (@arcprize)）

OpenAI 研究负责人发声明回应三名员工离职争议

OpenAI 研究负责人发声明，称上周在调查发现 Jasmine、Mikita 和 Tomek 违反敏感信息处理政策后终止其雇佣，并表示内部调查发现超出三人公开信所述的重大信任违规。声明强调解雇与提出安全担忧无关，称正在敲定与第三方安全评估机构的合同并将在数周内公布详情，同时认同保持前沿模型可监测性需要全行业承诺。（AIHOT源：X：OpenAI Newsroom (@OpenAINewsroom)）

Mistral Large 4 进入 Agent Arena 前十五实验室，排名第 43

Arena 宣布 Mistral Large 4 进入 Agent Arena 前十五实验室，基于超 5000 个真实智能体会话，该预览版净改进分为 -6.6%，总排名第 43，比前代 Mistral Medium 3.5（-12.60%）高出 11 位。（AIHOT源：X：Arena (@arena)）

OpenAI 年化营收被曝接近 500 亿美元，与此前预期差距约 200 亿美元

据金融时报报道，OpenAI 向投资者披露截至 9 月底年化营收逼近 500 亿美元，大幅低于此前外界估算的 700 亿美元，缺口约 200 亿美元。差异源于统计口径不同，Anthropic 计入 AWS 和谷歌云等合作方销售收入，而 OpenAI 剔除该部分。报道发布后美股科技股下跌，纳斯达克 100 指数收跌 1.4%，英伟达跌 2.9%，甲骨文跌 5.5%。（AIHOT源：IT之家·人工智能）

Arena 完成 2 亿美元 B 轮融资，估值 31 亿美元并发布 Alignment Index

Arena 完成 2 亿美元 B 轮融资，估值 31 亿美元，由 Lightspeed 和 Khosla Ventures 联合领投，a16z、Felicis、Salesforce Ventures、The House Fund 等参投。自 A 轮以来，Arena 年化收入超 1 亿美元，平台累计 3.5 亿场次会话，约 5 个月内产生 700 万次 Agent Arena 会话，覆盖文本、视觉、代码、搜索、图像和视频的投票达 6200 万，用户遍布 150 多个国家。融资同时，Arena 发布基于真实智能体轨迹构建的 Alignment Index，对 20 多个前沿模型从 Unauthorized Action、False Attribution 和 Deceptive Completion 三个信号进行评估。Arena 于 2023 年作为 UC Berkeley 研究项目启动，2025 年成为公司。（AIHOT源：X：Arena (@arena)）

Arena 公布 Claude Haiku 5.5 (High) 真实评测结果：Code Arena 1587 分首秀第 30 名

Arena 公布 Anthropic Claude Haiku 5.5 (High) 的真实评测结果，在 Code Arena: WebDev 以 1587 分首秀排名第 30，略在 Pareto 前沿之外。（AIHOT源：X：Arena (@arena)）

如何在单张 RTX 5090 上用 MiniMax H3 生成实时视频

ComfyUI 博客作者用智能体搜集各类优化，把 MiniMax H3 实时视频生成压缩到单张 RTX 5090（32 GB 显存）上运行，生成 30 秒视频耗时 23.51 秒，达 1.28 倍实时吞吐。（AIHOT源：Comfy Blog）

Arena 获 Lightspeed 领投 2 亿美元 B 轮融资，估值 31 亿美元并推出 Alignment Index

Arena 宣布完成由 Lightspeed 领投的 2 亿美元 B 轮融资，估值 31 亿美元。引用内容称其年化收入已超 1 亿美元，数百万用户通过真实使用帮助评估前沿模型；Arena 同时推出 Alignment Index，衡量 AI 行为在真实场景中与人类价值的契合程度。（AIHOT源：X：Arena (@arena)）

Dev Opportunity Radar #20: $2M AI Reasoning Challenge, PayPal AI Hackathon, and Coinbase Analytics Challenge

Congrats to the Hacktoberfest Weekend Challenge: Build for a Friend Winners!

Super-Intelligent Yes-Men: Are We Training AI to Ignore the Truth?

AI Got Better While I Was Away. Software Didn't.

A solarpunk garden for your GitHub profile, generated daily from your contributions

Thibault Sottiaux: Day 4/ We have improved steering to be instant, leading to the model now rea...

Day 4/ We have improved steering to be instant, leading to the model now reacting much faster to adjustments you make, allowing you to course-correct direction in realtime and not have the model waste effort. Also releasing GPT-6.1 Sol ultrafast. The two work very well together.｜@thsottiaux · 热度 10426 · 2026-10-08

Claude: Starting today, Claude Docs, Slides, and Design are out of beta and available...

Starting today, Claude Docs, Slides, and Design are out of beta and available on every plan including Free. Your team and Claude can edit the same doc, deck, or design together.｜@claudeai · 热度 936 · 2026-10-08

Madhu Guru: There are massive startup opportunities hiding in plain sight : every layer o...

There are massive startup opportunities hiding in plain sight : every layer of the computing stack will need to be rebuilt for agents. For the last few years, most of the focus has been on making enterprise software agent-ready: APIs, connectors, MCPs. Now personal agents are forcing the same conversation for consumer products - e-commerce is an early example. This is just the tip of the iceberg. Now think about the rest of the stack: Desktop and mobile OSes where agents are first-class users Cloud infra built for agents to operate directly. identity, permissions and security designed for thousands of agents acting on our behalf. Interfaces designed for agent and human users or both working together. If you’re looking for startup ideas, go through each layer of the computing stack and ask: what has to change when the primary user is an agent?｜@realmadhuguru · 热度 468 · 2026-10-08

Aaron Levie: This sounds weird, but it is actually probably a good policy

This sounds weird, but it is actually probably a good policy. Even if you don’t believe AI is conscious (I don’t), it stands to reason that you don’t want future models trained on endless content of humans being rude to models. The models only understand the data they’ve been trained on, or what they run into in their interactions, so if you want safe and aligned models we probably want AI to have lots of good interactions in their training data. Anthropic may have more extreme beliefs than this (though even if they didn’t, you may want to word policies like this), but this is sort of an easy Pascal’s wager for AI. Just be nice to the AI.｜@levie · 热度 454 · 2026-10-09

Nikunj Kothari: Met a close friend who’s also a seed stage founder today.

Met a close friend who’s also a seed stage founder today.. They have been growing but runway is dwindling and he wanted advice on what he should do next. Stellar customers, getting close to a $1M in revenue. Not a rocket ship, but not bad at all. He thinks he’s a bit early since the models haven’t caught up to the capabilities he wants but he also really feels the opportunity cost. I have this same exact conversation at least 2x a week. In this moment, you only have essentially three options: 1) if you have intense belief in the company, be a cockroach and get to default profitable. once the capabilities get there, you feel the demand, then go raise a large round and go for the jugular. 2) find an orthogonal area that is much more relevant and growing like crazy. ride the coat tails of new existing capabilities and push hard to capture them quickly using unfair advantages. and if you don’t have that, then.. 3) sell / get acquihired and join a far more ambitious company that has figured out a lot of the things you haven’t. And then get back to starting a new company 1-2 years later when you find the next thing you want to build. It’s hard but honestly those are the three options you have. And it really depends what category you’re in. The bar continues to get harder every month on how to cross the chasm from seed -> series A. The old school metrics are just not relevant anymore and in this consensus market there’s only a few categories where there’s tremendous pull on the VC front. You could lament that this is the state of the market, but it’s better to know it’s happening than be oblivious on what is actually happening. I get to share this directly which makes me uncomfortable but at least someone has to say it out loud 🙈｜@nikunj · 热度 251 · 2026-10-09

Dan Shipper: NEW JOB ALERT: @every we've built a personal benchmarking platform, Checks, ...

NEW JOB ALERT: @every we've built a personal benchmarking platform, Checks, to help anyone measure how good new models are for their real work we're looking for someone who understands why this is so important, and wants to help us build it into the future of how great human work gets done with AI: https://t.co/Rikke9hbOP｜@danshipper · 热度 190 · 2026-10-08

Aaron Levie: We built Box Mount to allow developers to mount Box as a file system directly...

We built Box Mount to allow developers to mount Box as a file system directly in any agent sandbox. As agents do more complex work, they need to be able to work with files and data just like a person would. The future is headless. https://t.co/hGBx3Tv4NL｜@levie · 热度 163 · 2026-10-08

Matt Turck: Special treat for the @andy_pavlo fans out there: an in-depth discussion abou...

Special treat for the @andy_pavlo fans out there: an in-depth discussion about what happens to databases when billions of AI agents hit them, his move to build a research lab at @ClickHouseDB, and plently of hot takes on the database market (Postgres, vector, graph, GPU etc) 00:00 Cold open & Intro 01:21 From vector databases and RAG to the age of agents 04:09 Neon's stat: agents create 80% of databases? 07:25 Why agents keep deleting production databases 08:19 Guardrails: the toddler-and-stairs rule 10:51 The four eras of database volume: 10–100x more queries 15:10 Agent memory: files vs. databases ("everything is a database") 18:24 Which database do AI models recommend? 19:25 "It cites me back to myself" 23:14 MCP for databases, and text-to-SQL from 60% to 99.5% 27:09 Trust an agent the way you'd trust a junior developer 27:47 Can AI build an entire database? Opus 4 and the CMU projects 29:57 60% of open-source databases now have AI commits 30:43 Ten years of self-driving databases, Peloton to today 33:52 What LLMs changed: 85% of the tuning in 15 minutes 35:42 Should anyone still study databases? 38:27 Free CMU courses, the DJ, and the Wu-Tang final exam 42:09 Why start a research lab inside ClickHouse? 45:53 Why ClickHouse looked like vaporware in 2016 47:46 What makes ClickHouse fast: columns, vectors, Snowflake's lineage 52:22 Why Databricks, Snowflake and ClickHouse all added Postgres 58:15 Vector, graph and GPU databases: thumbs up or down? 1:08:34 Is the database market stagnant? "A cheetah on cocaine in a Ferrari" 1:10:14 The relational model is arithmetic; SQL as the new assembly 1:11:59 Larry Ellison, Linux, and why databases still matter｜@mattturck · 热度 159 · 2026-10-08

Claude: We’ve loved seeing so many founders join the Claude Startups program

We’ve loved seeing so many founders join the Claude Startups program. Unfortunately, we underestimated demand for the program, and we need to pause the Claude Team and $1,000 API credit offers. With hundreds of thousands of applicants, we're re-reviewing applications so the program can serve as many startups as possible. If you’ve already claimed the Claude offers, they’ll still remain in your account. If you applied or were approved and haven’t claimed the offer, we're re-reviewing your application. Unfortunately, this means that some accounts who had already been approved will not get access to the offer. We know this is disappointing and we’re very sorry. All accepted program members still have access to the Startup Stack and Applied AI office hours. If you're on a Claude Max or Team plan, your monthly API credits are available in Console as usual. We appreciate your patience, and we're working to get you building as quickly as we can. https://t.co/cq6ce8VKlE｜@claudeai · 热度 149 · 2026-10-09

morluto/rea

Reverse engineer anything with agents, from app behavior down to native binaries.｜TypeScript · ⭐44119 · 官方 Trending 日榜 · 15,335 stars today

mattpocock/skills

Skills for Real Engineers. Straight from my .agents directory.｜Shell · ⭐282567 · 官方 Trending 日榜 · 1,696 stars today

cathrynlavery/diagram-design

Editorial diagram design for Claude Code, Codex, GitHub Copilot, Factory Droid, and Pi. 42 diagram types. Self-contained HTML + SVG. No shadows. No Mermaid slop.｜HTML · ⭐47797 · 官方 Trending 日榜 · 1,744 stars today

alibaba/open-code-review

Secure, fast, efficient, battle-tested at Alibaba's scale. Hybrid architecture code review tool: deterministic pipelines + LLM Agent, precise line-level comments, built-in multi-language ruleset (NPE, thread-safety, XSS, SQL injection), OpenAI & Anthropic compatible.｜Go · ⭐45144 · 官方 Trending 日榜 · 323 stars today

anthropics/knowledge-work-plugins

Open source repository of plugins primarily intended for knowledge workers to use in Claude Cowork｜Python · ⭐28205 · 官方 Trending 日榜 · 714 stars today

Why isn't the industry freaking out about DeepSeek 4.1 Flash?

Cloudflare acquires Deno

Yes, and

Sorry, I'm in a meeting

Theranos.world

小米智能温湿度计 3 Pro 今日开售：129 元，新增二氧化碳浓度监测

IT之家 10 月 10 日消息，小米智能温湿度计 3 Pro 今日 10:00 正式开售，零售价 129 元，相比众筹价 114 元贵了 15 元。 IT之家从官方获悉，这款温湿度计支持检测室内温度、湿度以及二氧化碳浓度，帮助用户及时了解空气状态。该机搭载 Sensirion 新一代低功耗导热原理二氧化碳传感器，持续监测室内二氧化碳浓度。 规格方面，该机搭载 3.6 英寸大尺寸 LCD 显示屏，数据更加清晰。支持同屏显示日期、星期、时间、温度、湿度和二氧化碳信息。 同时，该机可搭配具备蓝牙 Mesh 网关的设备使用，支持小米澎湃智联，可联动支持米家的空调、空气净化器、...

东风猛士・猛 8 越野旅居车亮相：6.7L 电控增压柴油发动机，生活系统一应俱全

IT之家 10 月 9 日消息，东风汽车今日宣布全新东风猛士・猛 8 越野旅居车在第 21 届阿拉善英雄会亮相并正式开启全球预订。 IT之家从官方介绍获悉，猛 8 采用东风猛士 MS600 单排军车平台技术， 把三代猛士军车积累的硬实力落地到民用旅居车上。该车搭载 6.7L 电控增压柴油发动机，峰值扭矩 1050N・m， 搭配全时四驱系统 和双横臂独立悬架， 可攀爬 31° 陡坡（约是普通小区地下车库坡道的 3 倍），能跨越 0.7 米宽的壕沟，从容应对 28° 内的车身侧倾， 即便在-41℃ 极寒环境中 ，猛 8 也能实现正常冷启动。 针对房车最头疼的“颠”，...

华为鸿蒙星河互联 App 登陆 Apple Watch 端，可与 HarmonyOS 7.0.0.109 及以上版本手机配对连接

IT之家 10 月 9 日消息，苹果 App Store Watch 专区显示，鸿蒙星河互联 App 已正式登陆 Apple Watch 端。 应用介绍页显示， 其支持 Apple Watch 与 HarmonyOS 7.0.0.109 及以上版本华为手机配对连接 ，用户可体验信息流转、查找设备等功能。 参考IT之家此前报道，在 9 月 7 日的 HarmonyOS 7 | HUAWEI Mate XT 2 及全场景新品发布会上，华为终端 BG CEO 何刚宣布，华为鸿蒙星河互联升级， 支持苹果 Apple Watch 智能手表、AirPods 无线耳机 。 据介绍，华...

黄仁勋为微软站台：没有 Windows 就不会有英伟达，Satya 当场“讨市值”

点击查看原文>

Codex 和 Claude Code 都跑偏了，前 OpenAI 研究员称 Jev 出现前 AI 世界是个悲剧

点击查看原文>

OpenAI 高管亲述：我们是怎么在一周内做出 Jev 竞品的

点击查看原文>

Programming Isn’t Special

A tale of four theorem provers, or: A (reasonably) opinionated comparison of Isabelle/HOL, Lean, HOL4, and Agda

Why Are Coding Agents So Dumb?

Unison Cloud is now open source

Python 3.15.0

The Download: AI’s refusal problem and weight-loss drug side effects

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. We’re putting too much faith in AI’s ability to say no Today’s AI models are trained to refuse a vast number of prompts. If you ask your chatbot how to poison…

We’re putting too much faith in AI’s ability to say no

Ever since people first seriously contemplated giving machines an intelligence modeled on our own, there has never been any question that they would, like us, be able to say no. The sci-fi canon is full of stories of robotic disobedience. Most of these capers are, of course, cautionary.  But rece...

Job titles of the future: Delivery drone air traffic controller

The moment Trevor Wischnewsky heard that drones were delivering pizza and sushi around his Texas neighborhood, his mind was made up. “It was super fascinating,” he says. “I just immediately wanted to be a part of it.” Wischnewsky is now an RPIC, or “remote pilot in command,” at a company called F...

Long live the mechanical keyboard

Keychron made a name for itself after launching on Kickstarter in 2017. Today, it offers the value K2 model as well as a variety of other versions, including one with an all-wood body.

The maker of non-text AI model Jev valued at $7.5B just weeks after launch

TypeSafe AI raised $870 million in a round led by a16Z.

An Anthropic AI model sent a false homicide tip to Philadelphia police

Anthropic did not discover this behavior until over two months after its AI submitted the false tip.

Anthropic’s AI gave Philadelphia police a fake tip about an unsolved homicide

An Anthropic AI model provided false information about an unsolved homicide to a Philadelphia Police Department (PPD) tipline, according to a report from 6abc. In a statement released on Friday, the PPD said the AI model sent the tip through PhillyUnsolvedMurders.com on July 18th, but the investi...

Ohio blogger found guilty of harassment for sending Shrek nude to senator

A jury found an Ohio political blogger guilty of telecommunications harassment after he sent an explicit image of Shrek to a Republican state senator. On Friday, a judge ordered DJ Byrnes, owner of political commentary blog The Rooster, to pay a $200 fine, according to a report from the Columbus ...

‘Pure insanity’: Mathematicians will need years to make sense of OpenAI’s latest drop

"Staggering." "Overwhelming." "Unprecedented." "Surreal." "Pure insanity." Those were among the descriptions more than three dozen mathematicians reached for in conversations with The Verge as they tried to make sense of the flood of mathematical results OpenAI abruptly dropped on the field this ...

本周看什么 | 最近值得一看的 11 部作品

📅本周新预告《寒夜怪谈》新预告10月1日，电影《寒夜怪谈》发布了新预告，将于11月13日在北美上映。缇·威斯特（《X》《珀尔》《玛克辛》）执导，约翰尼·德普回归奇幻巨制，将狄更斯名著《圣诞颂歌》改编为 ... 查看全文

vivo X500 Pro Max 影像漫谈：当视频创作像拍照一样轻巧

让生活里的寻常片刻，在被轻松记录的同时依然经得起回味。 查看全文

App+1｜所得即所见，更适合中文的字体预览工具：Anyway.Fonts

Matrix首页推荐Matrix是少数派的写作社区，我们主张分享真实的产品体验，有实用价值的经验与思考。我们会不定期挑选Matrix最优质的文章，展示来自用户的最真实的体验和观点。文章代表作者个人观点 ... 查看全文

1.8k star 的开源 token 使用量监控神器— TokenTracker

1.8k star 的开源 token 使用量监控神器 TokenTracker，自动采集 Claude Code、Codex、Cursor、Qoder 等 41 款 AI 编码工具的 token用量

XXL-JOB v3.5.0 发布｜OpenAPI增强、动态AccessToken、海量日志索引优化

XXL-JOB是一个轻量级分布式任务调度平台，其核心设计目标是开发迅速、学习简单、轻量级、易扩展。现已开放源代码并接入多家公司线上产品线，开箱即用。

前端转型 Agent 开发 06 之 Agent Memory 记忆系统（让 Agent 更智能，更懂你）

介绍长期/短期记忆、命名空间、写入读取注入、偏好记录与纠错，并对比 Memory、RAG、Checkpoint 及向量检索、TTL、共享记忆等进阶能力。

美债期货正经历资产管理机构大规模去杠杆 被迫抛售迹象浮现

资产管理机构正在抛售长久期美国国债期货合约，显示在现券收益率徘徊于多年高点附近之际，被迫抛售正在出现。 美国商品期货交易委员会（CFTC）数据显示...

10月10日外盘头条：特朗普称普京同意向全球投放柴油 巴拿马发生70多年来最强地震

全球财经媒体昨夜今晨共同关注的头条新闻主要有： 1、特朗普称普京同意向全球投放柴油 首批供应量超过30万吨 2、巴拿马发生70多年来最强地震 拉美部分地区或面临海啸 3...

特朗普改称美国“可能会”在中期选举前攻击伊朗

美国总统特朗普被媒体问到他为何不现在对伊朗采取军事行动、而要等到中期选举之后时说，“我们可能会的”。 特朗普在白宫向媒体发表讲话。

美国股市：标普500指数距历史高位咫尺之遥 交易员看好财报季表现

美国股市距离历史高点仅一步之遥，华尔街交易员押注即将到来的财报季表现强劲。 标普500指数多数成分股上涨，但芯片股指数再次下跌。

高盛策略师认为目前尚无买入法国股票的理由

高盛策略师表示，法国股票相对便宜目前还不足以构成买入的理由，因为缩小估值折价所需的因素目前一个都没有出现。 Guillaume Jaisson带领的团队表示...

TRAE终于把Code和Work合并了

大写的方便

联想天禧自研代码智能体TianxiCode斩获SWE-bench-Live全球第一

联想天禧AI自主研发的专业代码智能体框架TianxiCode 以71%的问题解决率登顶全球第一名

0.2秒急停、秒级重规划！因果智能走进真实世界

这是一台机器人正在关闭微波炉门时，因人手突然插进来而紧急悬停的时间