OpenAI's product lineup is finally consolidating.
According to Wired, Greg Brockman — OpenAI co-founder and former CTO — announced in an internal memo a plan to merge ChatGPT and Codex into a single product experience. The appointment came during AGI deployment CEO Fidji Simo's medical leave, with Brockman taking an interim role overseeing product direction.
But "interim" at OpenAI often means longer than it sounds.
The Signal Behind the Merger
Brockman's words in the memo: "We're consolidating our product efforts to execute with maximum focus toward the agentic future, to win across both consumer and enterprise."
Two keywords: consolidating and agentic.
ChatGPT and Codex are increasingly overlapping in positioning. ChatGPT can already write code, and Codex can already chat. Maintaining two product lines with separate teams, separate UX, separate growth metrics — that's exploration in early days, waste after maturity.
For OpenAI, the deeper motivation is likely the urgency of the "agent narrative." Anthropic's Claude Code has good口碑 among developers, and Google's Gemini is catching up. OpenAI needs a unified entry point for agent capabilities, not users bouncing between two products.
What Brockman's Return to Product Means
The last time Brockman was deeply involved in product decisions was around the GPT-4 launch. Since then, he focused more on infrastructure and compute. Coming back to manage product signals two things:
First, OpenAI's product strategy has reached a crossroads requiring "founder-level judgment." Merging ChatGPT and Codex isn't just adding a tab — it means redefining "what ChatGPT is" from a chatbot to an agentic platform.
Second, Fidji Simo's medical leave may be longer than publicly described. If it were just a week or two, a co-founder wouldn't step in personally.
But Merging Is Easy, Integration Is Hard
Historically, product line mergers fail more often than they succeed. Microsoft tried merging Skype and Teams — both ended up worse. Google stuffed Google+ features everywhere — nothing worked well.
The core challenge for ChatGPT + Codex: two user groups with fundamentally different needs.
- Regular users want simplicity, intuitiveness, answers to questions.
- Developers want precision, control, workflow integration.
These definitions of "good" sometimes contradict each other. Shoving two products into one interface might satisfy neither.
Brockman's solution is likely layered: one entry point, different modes. But the specifics depend on the product rollout over coming months.
One Thing to Watch
A telling signal: what will the merged product be called? If it stays "ChatGPT," OpenAI believes brand equity matters more on the consumer side. If it gets a new name (or uses Codex as the primary brand), the developer market has been elevated to higher strategic priority.
OpenAI's IPO timeline is approaching (the Musk lawsuit just cleared one obstacle). Delivering a clear, unified, "agent story"-capable product before IPO is management's top priority.
Brockman's return to the product front is the response to that priority.
Main sources: