Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

I wonder if it would make sense to re-train an AI like ChatGPT on a large internal code base and then ask it to write new modules in the style of existing modules. If the code it produces is only 30% correct, it would still save a ridiculous amount of time.


Counting deployment. Testing. config generation . im expecting above 50% for ETL like projects.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: