OpenAI recently shelled out ~3.5 Bn to acquire a agentic code-editor startup Windsurf in an attempt to acquire the foothold in the most profitable GenAI-application layer -- Agentic Coding Tools.
Shortly after the acquisition, two products dropped:
- Codex -- an agentic coding assistant by OpenAI
- SWE-1 -- an agentic developer tool by Windsurf
and we at This Week in AI thought that was it.
But then Google IO happened. In the event's footnote was Google Jules (https://jules.google/) -- an agentic asynchronous coding agent -- Google’s first serious play in the space.
And Google delivered!
Jules comes with game-changing capabilities that step beyond both SWE-1 and Codex. For example, it integrates with your codebase via a secure cloud-based VM, validates code changes, and independently commits to GitHub.
Compare that with tools like Codex or Cursor, which typically read code line by line and make suggestions within the editor -- Jules takes a leap forward in autonomy and deployment readiness.
Now, combine Gemini 2.5 and Jules into an enterprise suite, and you can see how Google could do to OpenAI's ecosystem what Microsoft Teams did to Slack.
Needless to say, Google's supremacy is here to stay and I'm here for it.
Let me know what you guys are using at your company for your coding tools!