I think we got the hang of vibe coding banging our head against the laptop yesterday and looking to level up more with Bootoshi (
https://x.com/KingBootoshi)
Key takeaways:
- linting and typing instructions are important, or needs chatGPT intervention as AI just tries to delete them or doesn’t always remember to handle for it
- save save save when things work to your liking enough so you have an example to show AI to improve with and compare against
- simple things like naming conventions and promises are important to get right from the beginning if you want to save money
- telling the AI to step back and take stock of the bigger picture helps it offer better solutions as it’s better at analyzing so you can give inputs on what to do and not to do
- refactoring code, modules and pipelines can be really helpful to reduce redundancies but make sure to save to GitHub first
- when making requests for changes to an API, ask it to dig into all components of services and controller BE and FE pipes
- code snippets and examples help the AI find patterns quickly and efficiently
More to come tomorrow as we dig into database optimization and state handling. End of the day though you figure out what works for you with languages you ideally know and multiple AIs and toolings to cross reference things.