Don't hesitate to reset your code if something isn't working. Accumulating layers of bad code from multiple prompts can lead to confusion, so it's better to start fresh with a clean codebase.

If you're unsure about a bug, try switching models. Different AI models may succeed where others fail, and if you find the source of a bug, reset all changes and give specific instructions to fix it on a clean codebase.

Use version control religiously, like Git, to manage your code. Always start with a clean slate before adding new features to ensure you can revert to a known working version if needed.

Frequent refactoring is essential once the code is functional and tests are in place, as it helps maintain a clean and manageable codebase.

It's important to quickly build and test ideas. If something doesn't work, it's better to kill it and start again rather than justifying its existence with additional features that may not address user needs.

When developing new technology, expect that it won't work on the first try or even the tenth; it's crucial to iterate quickly and avoid anything that slows down your progress.

Many developers overlook the importance of creating simple evaluations, which can help identify regressions and improve system performance incrementally.

When creating a prototype, prioritize speed and simplicity over perfection; aim to identify the core value of your product quickly, as you may discard most of your initial code.

I find that using asserts in programming helps catch potential issues early, preventing them from becoming real bugs later on.