How Codex happened: inside OpenAI's year-long turnaround to 40M weekly users
智东西 · wechat · 2026-10-20
Dan Shipper's long-form "How Codex Happened" (reconstructed by 智东西) traces Codex from a struggling cloud product to the foundation of the ChatGPT desktop app with 40M weekly users. Key beats: a Rust rewrite decoupling the agent core from the terminal, the cloud version's failure blamed on ignoring external users' setup costs, GPT-5 post-training using programmer supervision to fix real-world failures, and Altman's internal-multi-bet philosophy that dodged the innovator's dilemma. Next step: going back to the cloud for async agents.
Related event: Every's Deep Dive: How OpenAI Codex Turned Failure Into Victory(4 posts)→
More from coding & agent
- Factory adds Linear delegations, letting Droid agents run issues automatically — matanSF · 2026-10-21
- Agent memory, goals and emotion should be designed as one system, argues dev — mimi10v3 · 2026-10-21
- Meta launches Muse Wonder coding models with 1M context, $4/$20 per million tokens — AIatMeta · 2026-10-21
- CastWorkshop: open-source studio ties Ollama and ComfyUI into persistent AI characters — CastWorkshop · 2026-10-21
- Agents break in production for boring backend reasons, not model reasons — DayKnown1050 · 2026-10-21
- REA turns any site, app, or game into an RL environment via open-source AgentEnv — henloitsjoyce · 2026-10-21