AI News AI资讯 4h ago Updated 1h ago 更新于 1小时前 50

Codex from 0 to 10M Users: Building ChatGPT Work — Akshay Nathan, OpenAI 从0到1000万用户构建ChatGPT Work——Akshay Nathan,OpenAI

Codex usage has exploded, with MAU increasing over 10x from January 2026, reaching 10 million users combined with ChatGPT Work shortly after launch. OpenAI reorganized to integrate Codex and ChatGPT Work under a unified "Superapp," shifting focus from coding agents to knowledge work agents. Knowledge workers now account for roughly 20% of Codex's user base, growing three times faster than developers, indicating a significant shift in AI adoption beyond traditional coding. ChatGPT Work enables us OpenAI Codex 用户量在2026年爆发式增长,MAU较年初提升10倍以上,ChatGPT Work与Codex合并后总用户数突破1000万。 Codex从专用编码工具演变为通用知识工作引擎,非开发者用户占比达20%,且增长速度是开发者的3倍。 OpenAI通过组织重组将Codex核心负责人纳入ChatGPT产品线,完成“超级应用”整合,推动代理能力向全场景扩展。 ChatGPT Work基于共享Agent Harness构建,支持跨文档、表格、幻灯片等原生工具的上下文感知与自动化编排。 产品形态正从“代码生成”转向“结果导向”,交互式Sites逐步替代传统演示文稿与电子表格,强调持久

75
Hot 热度
68
Quality 质量
72
Impact 影响力

Analysis 深度分析

TL;DR

  • Codex usage has exploded, with MAU increasing over 10x from January 2026, reaching 10 million users combined with ChatGPT Work shortly after launch.
  • OpenAI reorganized to integrate Codex and ChatGPT Work under a unified "Superapp," shifting focus from coding agents to knowledge work agents.
  • Knowledge workers now account for roughly 20% of Codex's user base, growing three times faster than developers, indicating a significant shift in AI adoption beyond traditional coding.
  • ChatGPT Work enables users to collaborate with agents across various tools and contexts, simplifying tasks that previously required manual operation of multiple applications.
  • The shared agent harness behind Codex and ChatGPT Work allows for persistent environments, artifacts, Sites, plugins, memory, and sub-agents, enhancing the delegation of tasks to AI.

Why It Matters

This development is crucial for AI practitioners and researchers as it highlights the rapid expansion of AI tools beyond their initial coding applications into broader knowledge work. The integration of Codex and ChatGPT Work demonstrates a strategic move towards creating more versatile and accessible AI tools that can cater to a wider audience, including non-developers. This trend suggests a future where AI becomes an integral part of everyday productivity, transforming how tasks are managed and completed across various industries.

Technical Details

  • Shared Agent Harness: Both Codex and ChatGPT Work utilize a common underlying agent framework, allowing for seamless integration and consistent user experience while maintaining distinct functionalities tailored to different user needs.
  • Persistent Environments: The introduction of persistent environments, such as Sites and artifacts, enables users to maintain continuity and context across sessions, enhancing the efficiency and effectiveness of AI-assisted tasks.
  • Memory and Context Gathering: Advanced memory capabilities allow ChatGPT Work to gather and retain context from various sources like code, Slack messages, documents, and local files, facilitating more informed and relevant assistance.
  • Sub-Agents and Task Delegation: The use of sub-agents allows for complex tasks to be broken down into manageable parts, with each sub-agent handling specific aspects, thereby improving the overall performance and reliability of the system.
  • Sandboxing and Security: Enhanced sandboxing mechanisms ensure that AI-generated actions are contained within secure environments, mitigating potential risks and ensuring user data integrity.

Industry Insight

  • Shift Towards Generalist Skills: As AI tools become more accessible, there will be a growing demand for individuals who possess a broad skill set with specialized expertise, enabling them to leverage these tools effectively across various domains.
  • Emphasis on Creativity and Judgment: With AI capable of generating content and performing routine tasks, the value of human creativity, critical thinking, and decision-making will increase, making these skills highly sought after in the workforce.
  • Integration of AI into Daily Workflows: The success of tools like ChatGPT Work indicates a trend towards deeper integration of AI into daily workflows, potentially leading to increased productivity and innovation across industries. Organizations should consider how they can adopt and adapt these tools to enhance their operations.

TL;DR

  • OpenAI Codex 用户量在2026年爆发式增长,MAU较年初提升10倍以上,ChatGPT Work与Codex合并后总用户数突破1000万。
  • Codex从专用编码工具演变为通用知识工作引擎,非开发者用户占比达20%,且增长速度是开发者的3倍。
  • OpenAI通过组织重组将Codex核心负责人纳入ChatGPT产品线,完成“超级应用”整合,推动代理能力向全场景扩展。
  • ChatGPT Work基于共享Agent Harness构建,支持跨文档、表格、幻灯片等原生工具的上下文感知与自动化编排。
  • 产品形态正从“代码生成”转向“结果导向”,交互式Sites逐步替代传统演示文稿与电子表格,强调持久化记忆与子代理协作。

为什么值得看

本文揭示了AI代理能力从工程领域向知识工作全面渗透的关键转折点,展示了Open如何通过统一底层架构实现多角色适配,为行业提供从工具型AI到自主型智能体的演进路径参考。对于关注生产力革命、人机协作范式转型及企业级AI落地的从业者而言,这是一份极具前瞻性的实战洞察。

技术解析

  • 共享Agent Harness架构:Codex与ChatGPT Work共用同一套智能体框架,支持持久化计算环境(Persistent Computers)、Artifact输出、Site生成、插件集成、内存管理(Chronicle)及子代理调度,形成可扩展的通用智能体操作系统。
  • 跨模态上下文聚合:Agent能自动收集来自代码库、Slack消息、本地文件、云端文档等多源信息,构建任务所需完整语境,无需用户手动切换应用或复制粘贴内容。
  • 差异化UX设计策略:尽管共享内核,但面向开发者的Codex保留Git可见性与命令行风格界面,而面向知识工作者的ChatGPT Work则采用自然语言交互+可视化成果展示模式,降低使用门槛。
  • 沙箱安全机制强化:依托Abhihek在AIE提出的沙箱方案,ChatGPT Work执行外部调用时具备独立隔离环境,防止恶意操作影响主系统,同时兼容未发布模型测试场景。
  • 渐进式智能释放:默认配置优化为适合大多数用户的平衡设置;高级用户可启用深层推理、Ultra模型或多代理协同模式以应对复杂任务,体现“简单入口 + 深度扩展”的设计哲学。

行业启示

  • 编程权限民主化加速:随着AI代理接管重复性编码与数据处理任务,“不会写代码但会用代码的人”将成为主流生产力群体,倒逼教育体系与企业培训重构技能树。
  • 产品边界模糊化趋势明显:工程、设计、运营职能因AI辅助而高度融合,未来团队结构将更扁平化,个体需具备“广域专长”而非单一技能深耕,创意判断力成为核心竞争力。
  • 从效率工具到行为伙伴跃迁:AI不再仅仅是加速器,而是主动参与决策流程的智能协作者(如自动生成绩效评估素材、规划家庭预算),要求企业在部署时必须建立信任机制与伦理审查框架。

Disclaimer: The above content is generated by AI and is for reference only. 免责声明:以上内容由 AI 生成,仅供参考。

Code Generation 代码生成 LLM 大模型 Product Launch 产品发布 Conversational AI 对话系统