AnySearch 面向全球推出 “学生与开发者成长计划”

7月24日，在 AnySearch 北京开发者沙龙活动现场，AnySearch 创始人兼CEO 薛冠群正式宣布推出“ AnySearch 学生与开发者成长计划”（AnySearch Student & Developer Program） 。该计划面向全球学生与开发者开放，旨在通过提供开放的 AI 搜索基础设施，支持更多 AI 创造者开展学术研究、开源创新和 AI 应用开发。 图片1 随着 AI 加速演进，AI 应用正在从简单信息交互迈向复杂任务执行，AI Agent 正在成为下一代应用的主流形态。在这一趋势下，实时、高质量的信息获取能力是 AI Agent ...

氪星晚报 ｜黄仁勋力挺中国AI开源模型；市场监管总局：深入整治“内卷式”竞争

大公司： 字节跳动在中卫成立朔方星河科技公司，注册资本24亿 36氪获悉，天眼查App显示，近日，中卫朔方星河科技有限公司成立，法定代表人为田凯，注册资本24亿人民币，经营范围包括信息技术咨询服务、软件开发、信息系统集成服务、网络设备销售等。股东信息显示，该公司由北京字跳网络技术有限公司全资持股。 周生生、老凤祥等品牌金饰价格集体下调 当地时间7月23日，黄金白银大跌，油价狂飙。现货黄金跌1.96%，报4049美元；现货白银跌3.60%，报57.5967美元/盎司。国际贵金属期货普遍收跌。COMEX黄金期货跌2.4%，报4052.30美元/盎司；C...

你的下一位面试官，可能不是人

谁能保证超级数字员工不是下一个AI落地的窗口   撰文  |  李嘉星 编辑  |  张薇 封面来源  |  元企AI   七月初，秋招开始预热。你投出去的下一份简历，可能不是先被HR看到，而是先被一个AI员工读完。 机器筛简历并不新鲜。真正有意思的是，如果这个AI不只是帮HR筛一下简历，而是像一个新入职的同事一样，被交代一个目标后，自己去找人、沟通、打分、汇报结果呢？ 最近，元企AI推出了第一款产品“超级HR”。 它不是一个只解决单点环节的工具，而是试图像一个员工一样接下任务：拆解岗位需求，去招聘平台找人，...

Midjourney V8.2 发布：专注美学提升与个性化理解

Midjourney 今日推出 V8.2 图像模型，重点提升美学质量、图像创意与个性化表现。低质量图像出现频率将显著降低，个性化功能能更精准理解用户审美偏好。V8.2 的个性化配置文件拥有更大、更优的图像选择池，建议用户尝试新旧配置文件体验最新版本。（AIHOT源：Midjourney：Updates（RSS））

英伟达、微软和Meta联合警告：应避免对开放权重模型过度监管

英伟达、微软和Meta联合签署公开信，警告对开放权重AI模型的过度监管将削弱美国在AI领域的竞争力。信中指出，开放权重模型能促进创新、降低准入门槛，并支持学术研究。OpenAI和Anthropic未签署该信函。（AIHOT源：Hacker News 热门（buzzing.cc 中文翻译））

Claude 5 代模型上下文工程新规则：Claude Code 系统提示词精简超 80%

Anthropic 为 Claude Opus 5 和 Claude Fable 5 等新一代模型删除了 Claude Code 超过 80% 的系统提示词，且编码评测无显著损失。（AIHOT源：Claude：Blog（网页））

Anthropic 发布 Claude Opus 5

Anthropic 发布 Claude Opus 5，其智能水平接近 Claude Fable 5，但价格减半。该模型在 Frontier-Bench v0.1 上性能超过 Opus 4.8 两倍以上，在 ARC-AGI 3 上得分是次优模型的三倍。Opus 5 即日起成为 Claude Max 的默认模型和 Claude Pro 的最强模型。（AIHOT源：Anthropic：Newsroom（网页））

Anthropic 联合 Andon Labs 发布 Drone-Bench，评估 AI 模型自主操控无人机执行定位追踪任务的能力

Anthropic 与 Andon Labs 合作推出 Drone-Bench，用于测试 AI 模型自主操控四旋翼无人机在室内环境中定位并追踪指定人员的能力。该基准将任务分解为 3D 地图重建、定位、导航、目标检测与跟随五个子任务，并通过软件复现实现快速评估。实验表明，该任务链的难度足以区分不同智能水平的模型，并揭示 AI 在物理世界操控能力上的进步轨迹。（AIHOT源：Anthropic：Research（发表成果 · 网页））

蚂蚁百灵发布Ling-3.0-flash原生混合推理模型

蚂蚁百灵发布新一代原生混合推理模型Ling-3.0-flash，总参数量124B，激活参数量仅5.1B，在传统推理、指令遵循与长文本等指标上对标甚至超越上一代旗舰Ring-2.6-1T。模型采用原生混合线性注意力架构与1/64稀疏MoE，并扩展至10，000+可交互训练环境，长输入下TTFT降低60%至80%以上。（AIHOT源：公众号：蚂蚁百灵（Ling））

Runway Agent 推出自然语言工作流功能

在 Runway Agent 中引入工作流。现在你可以通过自然语言构建、运行或编辑基于节点的工作流。工作流可大规模解锁高质量输出。 立即尝试，点击下方链接调用 / Workflow 技能。（AIHOT源：X：Runway (@runwayml)）

微软阐述开源模型助力美国竞争力路径

开放权重模型对健康的 AI 生态系统至关重要。我们与行业同仁一道，正在规划一条路径，让开放权重模型在保护国家安全的同时，增强美国竞争力并扩大经济机会。（AIHOT源：X：Satya Nadella (@satyanadella)）

百度搭子更新：电脑手机接力、桌面端内嵌浏览器上线，复杂任务可跨端连续执行

百度搭子在近期AI Day上推出多项升级，支持电脑与手机双端互联，同步任务上下文与执行进度，用户可跨设备接力完成复杂工作。桌面端内嵌浏览器正式上线，能自动打开多个网页执行调研、下载等操作，手机端支持云端远程操控。智能路由自动匹配任务模式，平均任务耗时降低20%，Token利用率提升25%；简单任务完成度达100%，复杂任务高交付率94%，积分消耗最高降低75%。（AIHOT源：公众号：百度智能云（文心））

FLUX 3 x mimic：新一代视频动作模型

Black Forest Labs 发布多模态基础模型 FLUX 3，联合训练图像、视频和音频，其中视频预测占训练算力的 95% 以上。该模型与机器人公司 mimic 合作推出 FLUX-mimic，已在奥迪生产线上测试部署。加入动作预测后，视频生成质量最初下降最多 10%，但经 3500 步训练后恢复原有水平。（AIHOT源：Hacker News 热门（buzzing.cc 中文翻译））

Kimi K3 在网络安全漏洞利用测试中大幅落后美国前沿模型，知识蒸馏或为原因

英国AI安全研究所与美国AI标准与创新中心联合评估显示，月之暗面的Kimi K3在ExploitBench基准上得分32.2%，远低于美国领先模型的76.2%，但优于智谱GLM-5.2的24.4%。（AIHOT源：The Decoder：AI News（RSS））

Claude-thermos：保持 Claude 会话缓存热度，避免重新编码费用

Claude-thermos 通过本地反向代理监控 Claude Code 会话，在主智能体因等待子智能体而空闲超过 5 分钟时，自动发送预热请求刷新提示缓存。实测约 185 次本地会话中，缓存过期导致的重新编码占账单约 22%。工具以 uvx 运行，支持自定义空闲阈值和预热间隔。（AIHOT源：Hacker News 热门（buzzing.cc 中文翻译））

🐍 Snake - rendered with 576 browser windows [warning - this WILL hurt your eyes...and PC!]

Dev Opportunity Radar #9: A Fully Funded AI Security Residency, SF Founder Residency, Figma Campus Leaders & More

6 Open Source Tools That Give You the Web Back

Teaching Antigravity to Direct: A Stateful Video-Editing Skill Built on Gemini's Interactions API and MCP

The World's Oldest Communication Protocol Is Music

Claude Blog: Claude Code now supports artifacts

Starting today, Claude Code can capture work progress as an artifact, which turn Claude Code's work into live, shareable visual pages— including PR walkthroughs, system explainers, dashboards, and release checklists—that update themselves as your session works. A Claude Code session can range from investigating an incident to refactoring a service to analyzing months of data. Artifacts translate the work into a web page anyone can open and explore, like a pull request walkthrough, a dashboard you can filter and sort, or even a release checklist that fills itself out as work gets done. Artifacts make it easier to collaborate on shared work, so teams can spend more time building and less time communicating status updates. Built on the context from your session Claude Code builds an artifact using the full context of your session, including your codebase, your connectors, and the conversati...

Claude: Voice mode now runs on Claude's more capable models and reaches the tools you...

Voice mode now runs on Claude's more capable models and reaches the tools you've connected mid-conversation. Talk through the hard problems out loud, in many more languages. https://t.co/k0CWAGjLdK｜@claudeai · 热度 10642 · 2026-07-23

Thibault Sottiaux: Should we rename ChatGPT Work to ChatGPT Vibe?

Should we rename ChatGPT Work to ChatGPT Vibe?｜@thsottiaux · 热度 1011 · 2026-07-24

Claude: Voice conversations now use more of the models you have in chat, including Cl...

Voice conversations now use more of the models you have in chat, including Claude Opus and Sonnet. Claude can also reach the tools you've connected mid-conversation, like your email and calendar. https://t.co/452G2ZZY1d｜@claudeai · 热度 890 · 2026-07-23

Garry Tan: Open weight models are very very important https://t.co/IwS4UYG3pD https://t....

Open weight models are very very important https://t.co/IwS4UYG3pD https://t.co/FhgtJzdqUl｜@garrytan · 热度 393 · 2026-07-23

Amjad Masad: Viktor was able to disrupt the agency model and make a lot of money by using ...

Viktor was able to disrupt the agency model and make a lot of money by using Replit. Then he figured: Why not automate the entire thing, not just the coding part? The Agency: It's merely an agent loop... So he asked our team for an MCP, and we built one and he's now built his autonomous agency.｜@amasad · 热度 333 · 2026-07-23

Peter Steinberger: We see that as well and added code paths that use the claude cli directly - h...

We see that as well and added code paths that use the claude cli directly - hard to fight the system. https://t.co/VM49zPSyk2｜@steipete · 热度 270 · 2026-07-23

Aaron Levie: The best way to think about AI is as a force multiplier for the fields you al...

The best way to think about AI is as a force multiplier for the fields you already know about (or for the rate at which you want to learn about a new one). The 3rd category -no existing judgment, no interest in developing it- will basically produce slop and won’t drive that much economically productive activity. What will happen is that the experts will get better and better at their craft and be able to do so much more. Because they can wield these tools in ways that actually drive high quality output, veer agents in the right direction when they go off, and actually incorporate the work into something useful. The expert engineer with agents will do far more productive work, precisely because they know how to steer the agent properly. The designer will product far better outcomes using AI than someone without an eye for design. And so on. As a result of this dynamic, specialization will continue to be important, if not even more so as the tools get more powerful. Because the expectation of the market gets to be much higher. Getting good at any craft will continue to be necessary in the future.｜@levie · 热度 214 · 2026-07-24

Guillermo Rauch: 🔴🔊 AI Gateway keeps getting better

🔴🔊 AI Gateway keeps getting better. Unreal product velocity from the team. https://t.co/OWkksdt10W｜@rauchg · 热度 204 · 2026-07-23

Amjad Masad: My chess autoresearch agent got a PhD in modern LLM finetuning

My chess autoresearch agent got a PhD in modern LLM finetuning. https://t.co/xw9mvVRhgY｜@amasad · 热度 188 · 2026-07-24

Swyx: one thing i think people dont appreciate enough about @poolsideai is their un...

one thing i think people dont appreciate enough about @poolsideai is their unusual degree of openness — not only have they shipped an excellent Small model that somehow beat @thinkymachines at coding, but most people (like @eliebakouch) have been shouting out their excellent papers, but also they're among a rare few to actually expose their full eval dataset as well - beautifully published, with across 6 public benchmarks with 4 runs each and hundreds of turns per run. you can satisfy for yourself if they rewardhack. brilliant.｜@swyx · 热度 140 · 2026-07-23

mattpocock/skills

Skills for Real Engineers. Straight from my .agents directory.｜Shell · ⭐186612 · Search 增补 · 近快照 +2189星 · 创建 2026-02-03

Graphify-Labs/graphify

Turn any codebase, with its docs, SQL schemas, configs, and PDFs, into a queryable knowledge graph. A /graphify skill for Claude Code, Cursor, Codex, and Gemini CLI: local deterministic AST parsing, every edge explained, no vector store.｜Python · ⭐95221 · Search 增补 · 近快照 +645星 · 创建 2026-04-03

DietrichGebert/ponytail

Makes your AI agent think like the laziest senior dev in the room. The best code is the code you never wrote.｜JavaScript · ⭐88961 · Search 增补 · 近快照 +566星 · 创建 2026-06-12

Panniantong/Agent-Reach

Give your AI agent eyes to see the entire internet. Read & search Twitter, Reddit, YouTube, GitHub, Bilibili, XiaoHongShu — one CLI, zero API fees.｜Python · ⭐60602 · Search 增补 · 近快照 +453星 · 创建 2026-02-24

Leonxlnx/taste-skill

Taste-Skill - gives your AI good taste. stops the AI from generating boring, generic slop｜JavaScript · ⭐67252 · Search 增补 · 近快照 +865星 · 创建 2026-02-19

Claude Opus 5

https://www.anthropic.com/claude-opus-5-system-card

My security camera shipped a GitHub admin token in its login page

India's first privately-developed rocket reaches orbit on debut launch

Nvidia, Microsoft, Meta warn against overregulating open-weight models

Letter: https://images.nvidia.com/pdf/Open-Weights-and-American-AI-L... [pdf] https://x.com/JensenHuang/status/2080643682408321103 , https://xcan

If coding has been solved, why does software keep getting worse?

香港 TVB 官宣更名无线集团，由传统电视台升级为跨媒体娱乐集团

IT之家 7 月 24 日消息，今日，@TVB官方微博 发布消息，TVB 宣布正式更名为“无线集团”， 由传统电视台升级为跨媒体娱乐集团 。业务更多元化，拥抱创新 AI 技术，市场拓宽到大湾区。 IT之家从官方公告获悉，TVB 中文名称由 “电视广播有限公司”更改为“无线集团有限公司” ，英文名称由“Television Broadcasts Limited”更改为“TVB Limited”。 无线集团行政主席许涛在致辞中表示：“由 1967 年启播以来，‘无线’已扎根香港超过半世纪，记载了香港的历史，亦承载了几代香港人的集体回忆。我们制作的内容，除了透过电视机、也透...

宁德时代管理层回应 200 亿-400 亿元回购计划：公司股价近期被低估

IT之家 7 月 24 日消息，据财联社报道，在今日举行的宁德时代半年度业绩解读会上，对于投资者所问回购股权原因，公司管理层表示，本次回购规模相对更大，核心逻辑在于公司内部判断自身发展前景、财务状况、盈利水平及行业市场地位均具备突出优势，但受资本市场阶段性波动影响， 当前公司股价存在明显低估，因此启动回购 。 今日收盘后，宁德时代 A 股股价为 383.01 元，A 股总市值为 1.77 万亿元。 据IT之家此前报道，宁德时代今晚发布公告，拟使用自有或自筹资金， 以集中竞价方式回购 A 股股份 ， 回购资金总额不低于 200 亿元且不超过 400 亿元 ，回购价格上限 573 ...

谷歌签署欧盟 AI 生成内容透明度行为准则，并呼吁避免监管复杂化

IT之家 7 月 24 日消息，谷歌宣布签署欧盟《人工智能法案》框架下的《AI 生成内容透明度行为准则》 ，以支持在欧洲市场负责任地使用人工智能。 据悉，该准则核心要求主要集中在以下四个方面： 机器可读的标识与水印：确保第三方工具或平台能够通过技术手段自动检测并识别该内容是由 AI 生成或修改的； 人机交互的明确告知：当 AI 系统直接与人类互动时，必须在设计上保证用户能意识到他们正在与机器对话； 深度伪造内容的强制声明：内容发布者必须以显眼的方式声明“深度伪造”内容是 AI 生成或修改的； 公共利益相关文本的标注：如果 AI 生成的文本旨在向公众提供涉及“公共...

“因为AI不再招研究生了！”菲尔兹奖得主转身加盟OpenAI：数学家这份工作很难维持

点击查看原文>

GKE 安全蓝图加入云厂商 AI 安全框架阵营

点击查看原文>

深度对话：AI Agent 真落地，为什么需要 GPU 原生认知数据库？

点击查看原文>

Don't take the black pill

Making GIFs from 35mm film photography

Delightful integration tests in Rust

FreeBSD ports frozen after someone commits the entire 150MB Linux Copilot binary

What are you doing this weekend?

The quest to keep organs alive outside the body

This week, I covered a fascinating effort to preserve organs outside the body. There’s a huge shortage of donor organs, and one of the main reasons is time—they survive only a matter of hours outside the body, even when they’re kept on ice. Doctors dream of organ banks—stores of human organs that...

The Download: an organ transplant breakthrough, and homegrown Chinese chips

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. Supercooled kidneys have been transplanted into pigs in a “landmark achievement”  When it comes to organ donation, time is everything. As soon as an organ has ...

Supercooled kidneys have been transplanted into pigs in a “landmark achievement”

When it comes to organ donation, time is everything. As soon as an organ has been carefully removed from a donor’s body, it starts to deteriorate. Surgeons have a matter of hours to get it into a recipient. Leave it too long and the organ will become unusable. In most cases, organs will be kept…

Prentis, new AI lab co-founded by Reid Hoffman, Marc Pincus in talks to raise $100M

The neolab is betting that automating routine computer tasks will soon outpace coding as AI's biggest use case.

TechCrunch Disrupt 2026’s new Smart Money Stage explores fintech, payments, AI, and everything between

Money has evolved into far more than the cash in your wallet or your bank account. And at TechCrunch Disrupt 2026, we’re devoting an entire stage to that progression.

Vietnam is looking to restrict social media for kids; here are the growing number of other countries doing the same

Australia was the first country to issue a ban in late 2025, aiming to reduce the pressures and risks that young users may face on social media, including cyberbullying, social media addiction, and exposure to predators.

After backlash, Meta pauses plan to ‘rate limit’ its smart glasses

Remember when Meta was planning to charge a $20 monthly subscription fee for the smart glasses feature that lets people hear each other more clearly - even though that feature runs locally on your glasses and doesn't require the cloud? Meta has paused those plans, spokesperson Tyler Yee confirms ...

Meta just created a moderation nightmare for its smart glasses

Meta's smart glasses have been a PR headache for the company. Public backlash has been swift, and fierce; people are concerned about the erosion of privacy and expansion of surveillance. Some especially bad actors are using the glasses to film themselves "pranking" random strangers. Women have be...

Qualcomm is about to raise prices and that’s bad news for everyone

Qualcomm sent a letter to customers on Friday warning of plans to increase its prices by "a percentage in the double digits," Bloomberg reports. The price hikes will go into effect starting with products shipped after September 1st. Qualcomm claims it has "exhausted its ability to absorb higher c...

本周看什么 | 最近值得一看的 13 部作品

📅本周新预告《复仇者联盟5：毁灭之日》正式预告7月20日，电影《复仇者联盟5：毁灭之日》发布了正式预告，将于12月全球献映。罗素兄弟执导，克里斯·埃文斯、克里斯·海姆斯沃斯、佩德罗·帕斯卡、小罗伯特· ... 查看全文

角落新声｜数字与实体、有线与无线：我的多场景有声角落构建

编者按：本文是「角落新声」征文活动的入围文章。本次主题征文由声擎冠名，邀你前来分享那个让你放松娱乐、拥抱汩汩心流、逃离尘世纷扰、找回本真自我的理想一隅。8月7日前投稿参与，即有机会获得声擎Cyber5 ... 查看全文

赛博对话｜对话潘玮哲x老麦煮机：电子产品在悄悄变贵！

不是说好AI来临人类就解放了吗，怎么我连手机都买不起了？ 查看全文

分不清AI Workflow与Agent？3个实战案例彻底讲透，做AI应用不再踩选型坑

😵写AI应用半年，一直踩Workflow/Agent选型大坑 做AI开发这段时间，踩过无数哭笑不得的坑： 复杂规划场景硬套固定工作流，流程写满几十层判断，需求一变全重构； 简单简历筛选流水线盲目上Ag

被AI绑架了！

AI已经不是火不火的问题了，是已经实实在在进入我这样人的生活里. 这就带来一个问题，比如以前，日常中会想到的一些技术点，想来juejin这里装个X。

LC 51 N 皇后：我以为难的是回溯，结果栽在了对角线下标

题目说的啥 其实规则说穿了很简单。 给你一个 n×n 的棋盘，要放 n 个皇后。 皇后能攻击同行、同列、同斜线的棋子，所以摆放的时候要保证两两之间都打不到对方。 最后返回所有合法的摆法，用 Q 代表皇

中国模型加速AI明星“内卷”，Anthropic上新Opus 5，性能逼近Fable 5价格打对折，刷新ARC-AGI-3纪录

来源：华尔街见闻 AI模型的竞争，正在同时向两个方向加速：一边是价格不断下探，另一边则是模型能力继续冲击通用人工智能（AGI）的边界。

大宗商品综述：布伦特回落 基本金属涨跌不一 金价微升

布伦特原油已从每桶100美元回落。尽管近期美伊敌对行动激化且白宫不断发出武力威胁，但石油仍能通过中东贸易航线运输。基本金属价格周五涨跌不一...

7月25日外盘头条：特朗普称美伊仍在保持沟通 高通宣布产品价格上调 马斯克身家一周缩水1300亿美元

全球财经媒体昨夜今晨共同关注的头条新闻主要有： 1、美军导弹袭击伊朗，特朗普仍为外交协议留有余地 2、特朗普威胁对欧盟产品征收新关税 以报复欧盟对谷歌罚款 3...

美军称在阿曼湾对一艘商船开火

央视记者当地时间24日获悉，美军中央司令部称，当天在阿曼湾向一艘商船开火，使其失去航行能力，该船至少4次试图突破美方针对伊朗港口的封锁。

纽约汇市：美元指数持稳 势创一个月来最佳单周表现

彭博美元即期指数周五基本持平，本周有望创下美联储6月释放鹰派信号以来的最佳单周表现。 彭博美元即期指数日内变化不大，本周上周0.4%，因为在地缘政治紧张局势加剧后...

合肥又押中AI独角兽：多模态赛道，3个月融了21亿

跑出原生全模态新路径

国产世界模型登顶李飞飞团队榜单！适配国产昇腾算力、代码权重全开源

给它一张图，还你整个世界

菲尔兹奖得主王虹，也发过NeurIPS

王虹主页唯一没挂链接的论文