ChatGPT's New Coding Agent Is Huge, Even if You Aren't a Programmer
1 min read
Summary
OpenAI has launched a research preview of Codex, a new feature powered by a model trained to enhance coding and software engineering tasks.
The Codex agent can be found in the ChatGPT sidebar and can complete multiple tasks on an existing codebase, such as adding new features and fixing bugs.
It can also answer questions on request and performs tasks in a separate, preloaded repository.
OpenAI claims that tasks will take between one to 30 minutes to complete, depending on their complexity, and users can monitor the progress in real time.
Whereas ChatGPT can generate code, Codex can integrate with GitHub repositories, making it much more useful for programmers maintaining multiple projects.
Furthermore, Codex produces cleaner code and runs tests until code passes all given cases, making it more reliable than ChatGPT.