OpenAI发布了Codex的Windows版Computer Use功能。根据演示视频,Codex已能处理像素级的Windows原生GUI操作,例如在收到“测试我正在做的WinUI应用”指令后,能直接打开Paint(画图)应用,流畅地选择工具、调色并一笔一笔绘制图案,这展示了其vision-action loop的成熟度。官方通过WinUI测试等场景,暗示该功能面向真实的生产环境。同时,通过ChatGPT移动应用,用户可以从手机端启动、审核和引导Codex在Windows机器上执行任务,形成“手机指挥+桌面执行”的跨设备协作流程。
Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
Claude Opus 4.8发布后,部分用户反馈其相比Opus 4.7升级感知不强,甚至认为模型变笨了。同时,有声音指出GPT-5.5更好用,并调侃该模型是否蒸馏了Qwen。此外,有用户在发布后尝试测试,发现自己的新注册账号意外变为“未注册”状态,此前该账号仅发送过一条消息,引发了对账号安全性的担忧。
Codex发布体验更新,支持Windows的Computer Use功能,但控制时用户无法操作。新增聊天记录管理(置顶、查找、归类)和个人资料页面(显示Token消耗、连续登录天数、最长任务)。ChatGPT上的Codex控制增加侧边对话(/side命令)、一键模型切换、iPad快捷方式,并在对话后显示Git Diff。
If you ever get tired of managing your Codex threads, just let Codex manage itself! Codex can now create threads, search...
10h 45m was codex picking the 5.5 party guests codex > settings > profile
OpenAI宣布其AI智能体Codex的Computer Use功能现已登陆Windows,允许Codex在Windows真实桌面执行任务。同时,ChatGPT手机App新增了对Windows端Codex的支持,用户可在移动端远程启动、查看和控制任务,实现跨设备协同。该功能目前处于早期体验阶段。
Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
关联讨论 3 条X:OpenAI Developers (@OpenAIDevs)IT之家(RSS)X:OpenAI (@OpenAI)tokens
If you ever get tired of managing your Codex threads, just let Codex manage itself! Codex can now create threads, search...
If you ever get tired of managing your Codex threads, just let Codex manage itself! Codex can now create threads, search...
For every ChatGPT conversation that started as "one quick thing" and became a full on saga: table of contents is here. A...
陶哲轩指出,研究过程中存在大量“认知摩擦”——例如验证想法、排除错误路径、将直觉转化为计算等试错环节,占据了主要时间。AI正在将这些摩擦成本降至零,使研究者能更自由地尝试“更疯狂的想法”。推文强调,许多非常规路径并非被证伪,而是被高昂的验证成本所阻碍。AI降低这一成本,让原本因“不便”而被放弃的弱信号得以被审视,这往往是发现的起点。
AI can give researchers the freedom to pursue "crazier" ideas. For Terence Tao, AI creates more room to experiment, test...
Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
OpenAI just dropped a completely new kind of model gpt-realtime-translate takes in speech audio from any language and ou...
codex 牛逼,我用新的 claude 4.8 找 bug,开了最高的 think, 找了半天搞出来一些不痛不痒的问题。还是 gpt 5.5 比较厉害,真让他找到了。
did you wish codex ask__user_question tool was available outside of plan mode
Greg Isenberg 认为,Claude Opus 4.8 的发布并未带来比 GPT-5.5 更有意义的提升,模型迭代(如4.6到4.7再到4.8)已进入类似 iPhone 常规升级的边际收益递减阶段。他指出,当前真正的创新发生在模型外围工具,例如 Claude Code 上线的 Dynamic Workflows 和 Codex 发布的带内置浏览器的桌面应用。他预测,模型正变得像“电”或 Uber 发动机一样可互换,未来用户将不再关心具体使用哪个模型。
I didn't cover Claude Opus 4.8 on my pod because I don't think it's MEANINGFULLY better than GPT 5.5 as of May 29th. We'...
OpenAI宣布其Codex的“计算机使用”功能正式支持Windows系统。该功能使AI能像人类一样操作桌面应用,此前于4月中旬上线时仅支持macOS。同时更新了手机远程控制能力,现可通过ChatGPT手机App在移动设备上启动、监控和审批运行于Windows电脑的Codex任务。这是继5月支持Mac主机后的扩展,目前仍属于早期体验。
Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
关联讨论 3 条X:OpenAI Developers (@OpenAIDevs)IT之家(RSS)X:OpenAI (@OpenAI)Windows users, this one's for you. Computer use now works on Windows, so Codex can take action on your Windows computer....
We're taking steps to accelerate defensive progress in biology: - Launching Rosalind Biodefense to help trusted builders...
Work continues on GPT-5.6! Earlier today a significantly better new checkpoint was made available internally
用户分享其AI驱动的PPT制作流程:先用GPT 5.5 Pro和Grok搜集资料并形成个人理解;再由Codex或CC加工成Markdown格式的经验帖;然后将内容上传至Youmind,生成大纲及20页高清PPT页面;最后导出图片包,在Keynote中完成自我介绍、FAQ等最终页面的制作。
OpenAI宣布通过新平台“Rosalind Biodefense”加速生物防御领域的防御性进展。该平台旨在帮助受信开发者构建新的生物防御和大流行病防备能力。同时,OpenAI正在扩大特定美国政府及盟友伙伴对GPT-Rosalind的访问权限,以支持公共卫生和生物防御任务。其目标是通过提供前沿AI工具,加强整个生态系统,提升社会在应对生物威胁方面的韧性。
OpenAI is retiring o3 from ChatGPT on August 26, 2026 and GPT-4.5 on June 27, 2026 (these changes apply only to ChatGPT,...
If you update to the latest ChatGPT iOS app version we have a few cool new things for Codex Mobile, including /side to a...
we shipped a new version of gpt-5.5 instant today. the previous model was too bullet pilled. the new one improves on som...
@giordanorandone @OpenAI Nah. The only thing that caught us off guard was a bug we discovered last minute. And we decide...
OpenAI 产品负责人介绍了 Auto Review 功能,其核心是一个 AI 智能体执行任务时,由另一个 AI 智能体实时验证其动作的安全性。该功能是 OpenAI 安全与对齐团队的研究成果,旨在让“AI 对齐”从研究概念变为实际产品,使用户能放心地让智能体处理敏感任务。
这周应该是Anthropic CEO 达里奥·阿莫迪人生中最高光难忘的一周吧,每晚应该都是这样甜甜入睡吧😄 - Opus 4.8 发布 - 以 9650 亿美元估值融资 650 亿美元 - 47B 美元 ARR 对比 OpenAI 25B 美元 - 再次重回全球AI铁王座称王,the king
We've raised $65 billion in Series H funding at a $965 billion post-money valuation, led by @AltimeterCap, Dragoneer, @G...
关联讨论 13 条Artificial Intelligence News(RSS)TechCrunch:AI(RSS)Anthropic:Newsroom(网页)X:小互 (@xiaohu)Hacker News 热门(buzzing.cc 中文翻译)X:Kim (@kimmonismus)X:Anthropic (@AnthropicAI)X:Rohan Paul (@rohanpaul_ai)IT之家(RSS)The Verge:AI(RSS)The Decoder:AI News(RSS)X:Testing Catalog (@testingcatalog)X:洪明 (@hongming731)