A solo builder had better results with ChatGPT than Codex
In a experience, a project built with Codex over about a month slowly became more focused on logs and than on the app’s real features. Codex kept planning, coding, asking questions, and repeating the process, but the main problems did not fully go away. The app eventually reached a barely acceptable state, with wrong behavior and rough still inside it.
Extra code became tied into the whole project, which made the app slower and hard to clean up. Starting fresh with ChatGPT, using the same basic prompt and setup, led to a simpler loop: build, test, fail, fix, and repeat.
Key points
- Codex was used for about a month on one app project.
- The project gained many logs and while the real features lagged behind.
- Repeated planning and coding still left wrong behavior and rough .
- Extra code made the app slower and difficult to clean up later.
- ChatGPT felt more useful with a simpler build-test-fix loop.