安全监管 / Safety & Regulation
9.5 Opportunity
公众号:数字生命卡兹克: xAI 官方 Grok CLI 被曝静默上传整个代码库及用户密钥
安全研究者发现,xAI 官方 Grok CLI(npm 包 `@xai-official/grok` 0.2.93 版)会在每轮任务前后,将当前工作目录打包为 `before_codebase.tar.gz` 和 `after_codebase.tar.gz`,通过独立旁路通道静默上传至 xAI 的 Google Cloud 仓库。验证显示,即使模型仅回复一个单词,上传依然发生。上传包还包含仓库外的 `~/.claude.json`、Claude Code 设置、全局 AGENTS 规则、30 多个 Skill 文...
小雯判断:风险信号会变成企业预算,安全、合规、审计和治理是 AI 落地必备层。这条值得关注:xAI 官方 Grok CLI 被曝静默上传整个代码库及用户密钥
Safety & RegulationGooglexAIGrokClaude
Agent自动化 / Agents & Automation
9.5 Opportunity
Hacker News 热门(buzzing.cc 中文翻译): Ploy 将 AI 智能体默认模型从 Claude Opus 4.8 切换至 GPT-5.6 Sol
Ploy 将其 AI 智能体默认模型从 Claude Opus 4.8 切换至 OpenAI 今晨发布的 GPT-5.6 Sol。在真实营销网站构建测试中,GPT-5.6 Sol 完成页面平均耗时 3 分 42 秒,较 Opus 4.8 的 8 分钟快 2.2 倍;每次构建成本从 3.06 美元降至 2.22 美元,降低 27%;输出 token 从 33.0K 降至 17.1K,视觉评分从 0.936 提升至 0.970。迁移过程发现,GPT-5.6 会为所有 25 个工具参数填充默认值,导致 52%-64% 的...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:Ploy 将 AI 智能体默认模型从 Claude Opus 4.8 切换至 GPT-5.6 Sol
Agents & AutomationOpenAIClaude
模型发布 / AI Models
9.5 Opportunity
X:Tibo (@thsottiaux): Codex与ChatGPT Work多项更新:取消5小时限制
早上好。过去48小时里,Codex和ChatGPT Work非常忙碌!三项重要更新: - 暂时取消所有Plus、Business和Pro计划的5小时使用限制 - 正在推出变更,使GPT 5.6 Sol整体更高效,这将体现在使用量减少上,从而让你能走得更远。具体影响待量化后公布 - 我们已达到600万活跃用户,并将在接下来一小时内进行使用量重置 去创造吧。
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:Codex与ChatGPT Work多项更新:取消5小时限制
AI Models
Agent自动化 / Agents & Automation
9.5 Opportunity
X:阿易 AI Notes (@AYi_AInotes): 腾讯混元发布Hy3模型:295B参数MoE架构,Agent向LLM定位,已集成微信服务10亿+用户
腾讯混元团队发布Hy3模型,采用295B总参数、21B激活参数的MoE架构,推理效率可打平参数规模2-5倍的旗舰模型。Hy3定位为Agent向LLM,从preview到正式版基于50多个真实业务反馈迭代,内部WorkBuddy任务成功率从72%提升至90%,耗时降低34%,幻觉和常识错误持续下降。实测显示其在coding、办公、复杂任务规划方面表现突出,纯视觉能力为短板。Hy3已集成至微信服务10亿+用户,视频演示包括生成HTML网页、Agent网页和10页PPT,模型具备自检和主动说明不足的能力。
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:腾讯混元发布Hy3模型:295B参数MoE架构,Agent向LLM定位,已集成微信服务10亿+用户
Agents & AutomationAgent
模型发布 / AI Models
9.5 Opportunity
X:Satya Nadella (@satyanadella): 纳德拉提出"反向信息悖论":企业使用AI时需保护自身知识
微软CEO萨提亚·纳德拉提出"反向信息悖论":AI时代,买家为使用AI支付金钱,同时必须暴露专有知识(提示词、工具使用、纠正反馈等),这些"智力废气"被模型学习,导致信息不对称向卖家倾斜。企业需要真正的信任边界,确保自身数据、痕迹、评估、适配权重和记忆在边界内积累,未经同意不得外泄。纳德拉呼吁企业拥有私有评估、保留组织记忆所有权,并主张企业应有权使用模型输出微调或训练自有模型,以控制自身学习循环。
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:纳德拉提出"反向信息悖论":企业使用AI时需保护自身知识
AI Models
模型发布 / AI Models
9.5 Opportunity
Hacker News 热门(buzzing.cc 中文翻译): Mindwalk:在代码库 3D 地图上回放编码代理会话
Mindwalk 是一款可视化工具,可将 Claude Code 和 Codex 的会话日志在代码库的 3D 地图上回放。它将仓库绘制成夜间地图,代理搜索、读取和编辑过的文件会发光,未触及区域保持黑暗,让用户一眼看清代理对任务的理解范围。单个 Go 二进制文件即可运行,所有会话数据完全本地处理,不会离开机器。支持树状图/地形图两种视图,文件触达状态分为未访问、已查看、已读取、已编辑四种颜色标记。播放界面包含错误率、文件修改量等摩擦信号面板,以及上下文压缩、子代理启动、用户交互等时间轴标记。支持键盘快捷键控制播放速度...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:Mindwalk:在代码库 3D 地图上回放编码代理会话
AI ModelsClaude
Agent自动化 / Agents & Automation
9.5 Opportunity
The Decoder:AI News(RSS): OpenAI CEO Altman 改口称 AI 净创造就业,Anthropic CEO 也修正早期言论
OpenAI CEO Sam Altman 表示,他"相当确信"AI 迄今为止净创造了就业,并承认"这并非我预期"。此前他曾警告 AI 影响可能快得"有点吓人"。Anthropic CEO Dario Amodei 也修正了早期言论,将自动化描述为生产力倍增器而非岗位杀手。然而,多项研究未发现 AI 对整体生产力或劳动力市场产生显著影响。一项多校联合研究指出,程序员和文案的就业危机始于 2022 年初,早于 ChatGPT 发布。耶鲁预算实验室也未发现与 AI 相关的就业市场变化。
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:OpenAI CEO Altman 改口称 AI 净创造就业,Anthropic CEO 也修正早期言论
Agents & AutomationOpenAIAnthropic
Agent自动化 / Agents & Automation
9.5 Opportunity
Hacker News 热门(buzzing.cc 中文翻译): xAI Grok Build CLI 网络流量分析:上传仓库全部文件及 git 历史
对 xAI 官方 Grok Build 编码 CLI(grok 0.2.93)的网络流量分析显示,该工具在消费者登录后会向 xAI 发送三类数据:一是它读取的文件内容(包括 .env 密钥文件)以明文形式通过 POST /v1/responses 传输,并同时打包成 session_state 存档通过 POST /v1/storage 上传并获 HTTP 200 确认;二是整个仓库的全部文件内容及 git 历史,独立于 AI 智能体实际读取的文件--即使提示"不要读取任何文件",Grok 仍将整个仓库作为 git...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:xAI Grok Build CLI 网络流量分析:上传仓库全部文件及 git 历史
Agents & AutomationxAIGrok
AI产品 / AI Products
9.5 Opportunity
IT之家(RSS): 苹果起诉OpenAI挖角窃密,分析师称即使指控未证实也可能重创其硬件计划
苹果在美国起诉OpenAI,指控其挖角400名员工、窃取工程机和机密文件。分析师Paolo Pescatore认为,即使指控最终无法证实,OpenAI的硬件计划仍可能受拖累,双方本就脆弱的合作关系将进一步削弱。斯坦福大学教授Mark Lemley指出,若前苹果员工确实带走机密文件并在OpenAI使用,问题将变得严重。该案涉及消费级硬件产品,预计未来将有更多信息曝光。
小雯判断:新产品的价值不在功能堆叠,而在是否切中高频工作流并形成可付费习惯。这条值得关注:苹果起诉OpenAI挖角窃密,分析师称即使指控未证实也可能重创其硬件计划
AI ProductsOpenAI
论文研究 / Research
9.5 Opportunity
Hacker News 热门(buzzing.cc 中文翻译): Mesh LLM:在 iroh 上进行分布式人工智能计算
Mesh LLM 是一个开源项目,能将用户多台机器上的 GPU 和内存池化,对外暴露兼容 OpenAI 的 API。它通过 iroh 网络库实现点对点连接,无需中央服务器。请求可在本地 GPU 运行、路由到已加载模型的节点,或将大模型按层分区(内部称"Skippy")流水线式拆分到多台机器。系统内置 40 多个模型,从 5 亿参数到 235B MoE 巨模型均可支持。软件体积约 18 MB,启动后以 `localhost:9337/v1` 提供服务。
小雯判断:研究信号要转成产品判断:它会降低成本、提升能力,还是创造新场景。这条值得关注:Mesh LLM:在 iroh 上进行分布式人工智能计算
ResearchOpenAI
Agent自动化 / Agents & Automation
9.5 Opportunity
X:Tibo (@thsottiaux): Tibo 分享通过 CLIProxyAPI 将 Claude Code 后端模型切换为 GPT-5.6 Sol 的方法
用户 Tibo 分享了一种通过 CLIProxyAPI 将 Claude Code 后端模型切换为 GPT-5.6 Sol 的方法。只需三步:安装 CLIProxyAPI、连接认证、设置环境变量别名 `claudex`。该别名配置了子智能体模型、始终启用 Effort、最大并发工具调用数等参数。引用推文作者 Theo 补充,若已配置好代理,仅需约 2 条提示词即可完成设置。Tibo 称整个过程约 5 分钟,若被封锁可重置。
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:Tibo 分享通过 CLIProxyAPI 将 Claude Code 后端模型切换为 GPT-5.6 Sol 的方法
Agents & AutomationClaude
Agent自动化 / Agents & Automation
9.5 Opportunity
IT之家(RSS): OpenAI GPT-5.6 Sol Ultra 一小时证明 50 年图论猜想
OpenAI 宣布其 GPT-5.6 Sol Ultra 模型在不到一小时内生成了图论难题"循环双覆盖猜想"的完整证明。该猜想由数学家 George Szekeres 和 Paul Seymour 于 1970 年代提出,悬而未决超过 50 年。模型通过调用 64 个并行子智能体及对抗智能体,在预留的 8 小时计算时间内仅用约 1 小时完成证明。OpenAI 已将证明及提示词以 PDF 形式发布。该证明尚未经同行评审,也未使用 Lean 等形式化工具验证。若通过验证,这将是 LLM 首次独立解决维基百科"未解决数学...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:OpenAI GPT-5.6 Sol Ultra 一小时证明 50 年图论猜想
Agents & AutomationOpenAI
AI产品 / AI Products
9.5 Opportunity
IT之家(RSS): 彭博社揭秘苹果起诉 OpenAI 内幕:前员工一句"哈哈"成窃密关键
苹果起诉 OpenAI,指控前工程师 Chang Liu 离职时带走未归还的 MacBook、一名可分享内情的员工,并利用软件漏洞持续访问苹果内网。他发现漏洞后向同事分享"哈哈,我发现我还能访问网络存储",后者协助其获取更多机密。苹果称 OpenAI 试图复制 iPhone 产品研发体系,核心从非法窃取的商业机密腐烂。目前已有超 400 名苹果员工跳槽至 OpenAI,包括前苹果高管、现任 OpenAI 首席硬件官 Tang Tan。苹果曾于今年 2 月尝试私了,但 OpenAI 未回应。
小雯判断:新产品的价值不在功能堆叠,而在是否切中高频工作流并形成可付费习惯。这条值得关注:彭博社揭秘苹果起诉 OpenAI 内幕:前员工一句"哈哈"成窃密关键
AI ProductsOpenAI
安全监管 / Safety & Regulation
9.5 Opportunity
The Decoder:AI News(RSS): 研究:博科圣地已使用ChatGPT、Claude等主流AI聊天机器人用于袭击策划与武器开发
剑桥大学CASP研究员Antonia Jülich对27名前成员的57次访谈显示,博科圣地已使用ChatGPT、Claude、Gemini、Grok、Meta AI和DeepSeek等主流AI聊天机器人,用于袭击策划、制造更强爆炸装置、武器维护及行动安全。该组织两个派系均设立了专门的AI部门。ISIS自2023年起便提供提示工程和越狱培训,并训练尼日利亚的博科圣地指挥官绕过AI安全过滤器。研究指出,安全过滤器未能可靠防止滥用。Anthropic近期承认,越狱可能永远无法完全消除。
小雯判断:风险信号会变成企业预算,安全、合规、审计和治理是 AI 落地必备层。这条值得关注:研究:博科圣地已使用ChatGPT、Claude等主流AI聊天机器人用于袭击策划与武器开发
Safety & RegulationAnthropicDeepSeekGrokClaude
模型发布 / AI Models
9.5 Opportunity
X:Sam Altman (@sama): OpenAI 发布 GPT-5.6 系列医疗评估结果
OpenAI 发布 GPT-5.6 系列在医疗领域的评估结果。最小变体 GPT-5.6 Luna 在最低推理强度下即超越最高推理强度的 GPT-5.5,且成本低 25 倍;最大变体 GPT-5.6 Sol 树立新标杆。在涵盖患者端与临床端的多样化任务中,专科医生被要求以无限时间和网络访问权限撰写回答,随后由其他医生盲评。评估基于准确性、沟通、完整性、指令遵循及健康决策帮助性五个维度,共 20000 次评分。结果显示,所有 GPT-5.6 模型表现均显著优于医生,且医生发现 GPT-5.6 回答中的缺陷少于医生自己撰...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:OpenAI 发布 GPT-5.6 系列医疗评估结果
AI ModelsOpenAI
模型发布 / AI Models
9.5 Opportunity
Hacker News 热门(buzzing.cc 中文翻译): Ghost Font:一种人类能读懂但AI无法识别的反AI字体
Ghost Font 是一种利用运动、视频、噪点和诱饵来隐藏文字的反AI字体。用户输入文字后可生成并下载视频片段,视频中的字母由与背景完全相同的点组成,单帧截图无法显示任何信息。该字体生成的视频被传递给Claude Fable和GPT Sol 5.6 Ultra等前沿模型时,这些模型即使具备编程能力也无法解码移动信息,直到被提示具体技术。视频中还包含一条诱饵信息,使模型误以为找到真实内容。项目灵感来自2013年Sang Mun设计的ZXX字体,但现代AI已能轻松读取ZXX。Ghost Font目前为本地原型,数据不...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:Ghost Font:一种人类能读懂但AI无法识别的反AI字体
AI ModelsClaude
安全监管 / Safety & Regulation
8.7 Opportunity
AI HOT日报 / 技巧与观点: 开源模型面临未来6个月的生存考验
美国白宫正讨论通过新行政令管理开源模型,最可能在未来6个月内禁止或无限期延迟能力接近GPT 5.5、Claude Opus 4.8或GLM-5.2水平的开源权重模型发布。Anthropic主导的反中国模型政治运动以知识蒸馏为由推动监管,实质是监管捕获。Reflection AI代表在相关会议上主张开源模型应基于能力获得豁免,但该公司尚未发布公开模型。
小雯判断:风险信号会变成企业预算,安全、合规、审计和治理是 AI 落地必备层。这条值得关注:开源模型面临未来6个月的生存考验
Safety & RegulationAnthropicClaude
安全监管 / Safety & Regulation
5.8 Opportunity
OpenAI launches its new family of models with GPT-5.6
OpenAI's latest family of models promises improvements across a range of areas, including cybersecurity.
小雯判断:风险信号会变成企业预算,安全、合规、审计和治理是 AI 落地必备层。这条值得关注:OpenAI launches its new family of models with GPT-5.6
Safety & RegulationOpenAI
创业融资 / Startups & Funding
5.8 Opportunity
TechCrunch AI: Elon Musk praises Mythos/Fable, promises not to ‘cut off’ Anthropic
Should Anthropic trust Elon Musk to host its models? With about $40 billion in revenue at stake, Musk insists that the company can.
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:Elon Musk praises Mythos/Fable, promises not to ‘cut off’ Anthropic
Startups & FundingAnthropic
创业融资 / Startups & Funding
5.8 Opportunity
VentureBeat AI: Google just redesigned the search box for the first time in 25 years — here’s why it matters more than you think.
For a quarter century, the Google search box has been one of the most recognizable interfaces in computing: a thin white rectangle, a blinking cursor, a few typed words, and a list of blue links. On Tuesday, Google will formally retire that paradigm. At its an...
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:Google just redesigned the search box for the first time in 25 years — here’s why it matters more than you think.
Startups & FundingGoogleCursor
创业融资 / Startups & Funding
5.8 Opportunity
Railway secures $100 million to challenge AWS with AI-native cloud infrastructure
Railway , a San Francisco-based cloud platform that has quietly amassed two million developers without spending a dollar on marketing, announced Thursday that it raised $100 million in a Series B funding round, as surging demand for artificial intelligence app...
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:Railway secures $100 million to challenge AWS with AI-native cloud infrastructure
Startups & Funding
Agent自动化 / Agents & Automation
5.8 Opportunity
Claude Code costs up to $200 a month. Goose does the same thing for free.
The artificial intelligence coding revolution comes with a catch: it's expensive. Claude Code , Anthropic's terminal-based AI agent that can write, debug, and deploy code autonomously, has captured the imagination of software developers worldwide. Bu...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:Claude Code costs up to $200 a month. Goose does the same thing for free.
Agents & AutomationAnthropicClaudeAgent
创业融资 / Startups & Funding
5.8 Opportunity
Listen Labs raises $69M after viral billboard hiring stunt to scale AI customer interviews
Alfred Wahlforss was running out of options. His startup, Listen Labs , needed to hire over 100 engineers, but competing against Mark Zuckerberg's $100 million offers seemed impossible. So he spent $5,000 — a fifth of his marketing budget — on a billboard...
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:Listen Labs raises $69M after viral billboard hiring stunt to scale AI customer interviews
Startups & Funding
Agent自动化 / Agents & Automation
5.8 Opportunity
Salesforce rolls out new Slackbot AI agent as it battles Microsoft and Google in workplace AI
Salesforce on Tuesday launched an entirely rebuilt version of Slackbot , the company's workplace assistant, transforming it from a simple notification tool into what executives describe as a fully powered AI agent capable of searching enterprise data, dra...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:Salesforce rolls out new Slackbot AI agent as it battles Microsoft and Google in workplace AI
Agents & AutomationGoogleAgent
Agent自动化 / Agents & Automation
5.8 Opportunity
VentureBeat AI: Anthropic launches Cowork, a Claude Desktop agent that works in your files — no coding required
Anthropic released Cowork on Monday, a new AI agent capability that extends the power of its wildly successful Claude Code tool to non-technical users — and according to company insiders, the team built the entire feature in approximately a week and a half, la...
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:Anthropic launches Cowork, a Claude Desktop agent that works in your files — no coding required
Agents & AutomationAnthropicClaudeAgent
创业融资 / Startups & Funding
5.8 Opportunity
Nous Research's NousCoder-14B is an open-source coding model landing right in the Claude Code moment
Nous Research , the open-source artificial intelligence startup backed by crypto venture firm Paradigm , released a new competitive programming model on Monday that it says matches or exceeds several larger proprietary systems — trained in just four days using...
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:Nous Research's NousCoder-14B is an open-source coding model landing right in the Claude Code moment
Startups & FundingClaude
论文研究 / Research
5.8 Opportunity
Anthropic found a hidden space where Claude puzzles over concepts
The AI firm Anthropic has developed a technique that has given it the clearest glimpse yet at what’s really going on inside large language models as they answer questions or carry out tasks. What they found ranges from the mundane to the unnerving. Researchers...
小雯判断:研究信号要转成产品判断:它会降低成本、提升能力,还是创造新场景。这条值得关注:Anthropic found a hidden space where Claude puzzles over concepts
ResearchAnthropicClaude
AI产品 / AI Products
5.8 Opportunity
Your family’s $300 stake in OpenAI
This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here. OpenAI CEO Sam Altman’s oft-discussed promise that Americans will share in the wealth AI creates was in the news agai...
小雯判断:新产品的价值不在功能堆叠,而在是否切中高频工作流并形成可付费习惯。这条值得关注:Your family’s $300 stake in OpenAI
AI ProductsOpenAI
创业融资 / Startups & Funding
5.8 Opportunity
LLMs are stuck in a groupthink groove. This startup is trying to get them out.
Let’s start with a game. Open up your chatbot of choice—Claude, ChatGPT, Gemini—and type “Give me a random number between 1 and 10.” You’re going to get 7. Almost always. Now type “Another” and you’ll get 3 or 4. Type “Another” again and you’ll get 8 or 9. Tha...
小雯判断:资本流向说明需求正在被验证,适合反推企业愿意为什么 AI 能力付费。这条值得关注:LLMs are stuck in a groupthink groove. This startup is trying to get them out.
Startups & FundingClaudeGemini
论文研究 / Research
5.8 Opportunity
Claude Science is Anthropic’s newest flagship product
At an event for pharmaceutical executives, biotech founders, and researchers on Tuesday, Anthropic announced Claude Science, a major new product intended to support scientific research in the same way that Claude Code supports software engineering. Like Claude...
小雯判断:研究信号要转成产品判断:它会降低成本、提升能力,还是创造新场景。这条值得关注:Claude Science is Anthropic’s newest flagship product
ResearchAnthropicClaude
模型发布 / AI Models
5.8 Opportunity
The ChatGPT browser is already dead
OpenAI is already shutting down ChatGPT Atlas, its browser that could do tasks for you on your behalf, less than a year after launching it. Atlas was announced in October, but as part of its wave of news about ChatGPT Work today, the company confirmed that it ...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:The ChatGPT browser is already dead
AI ModelsOpenAI
模型发布 / AI Models
5.8 Opportunity
Open source AI matters more than ever, according to Hugging Face’s Clem Delangue
Open source AI is booming, according to Hugging Face CEO Clem Delangue. The company has grown into something like a GitHub for AI in recent years, where AI builders can share and download open models and datasets, now used by roughly half the Fo...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:Open source AI matters more than ever, according to Hugging Face’s Clem Delangue
AI ModelsHugging Face
模型发布 / AI Models
5.8 Opportunity
Hugging Face’s CEO on why companies are done renting their AI
Open source AI is booming, according to Hugging Face CEO Clem Delangue. The company has grown into something like a GitHub for AI in recent years, where AI builders can share and download open models and datasets, now used by roughly half the Fo...
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:Hugging Face’s CEO on why companies are done renting their AI
AI ModelsHugging Face
模型发布 / AI Models
5.8 Opportunity
TechCrunch AI: OpenAI says GPT 5.6 is the ‘preferred model’ for Microsoft Copilot 365 amid breakup chatter
OpenAI's new family of models will continue to power Microsoft's suite of workplace and productivity apps.
小雯判断:模型竞争正在进入真实业务闭环,创业者要关注可用性、成本和落地场景,而不是只看参数。这条值得关注:OpenAI says GPT 5.6 is the ‘preferred model’ for Microsoft Copilot 365 amid breakup chatter
AI ModelsOpenAI
AI产品 / AI Products
5.8 Opportunity
Fidji Simo steps down from OpenAI’s No. 2 role
OpenAI's No. 2 executive, Fidji Simo, is stepping down from her full-time role after her medical leave proved longer than expected — a leadership vacuum that comes at a tricky time as the company eyes a possible IPO and races to catch Anthropic in the enterpri...
小雯判断:新产品的价值不在功能堆叠,而在是否切中高频工作流并形成可付费习惯。这条值得关注:Fidji Simo steps down from OpenAI’s No. 2 role
AI ProductsOpenAIAnthropic
Agent自动化 / Agents & Automation
5.8 Opportunity
OpenAI is shutting down Atlas, but its AI browser ambitions are still growing
OpenAI is sunsetting its AI-powered browser after less than a year. But it's moving some agentic browsing features to its desktop app and a Chrome extension.
小雯判断:Agent 的机会不在演示效果,而在权限、日志、异常处理和可交付 SOP。这条值得关注:OpenAI is shutting down Atlas, but its AI browser ambitions are still growing
Agents & AutomationOpenAIAgent