In short: of course you can use AI to code, but you should use it intelligently.
A few key points (with more here]:
First, AI doesn't do software end-to-end. It does it middle-to-middle. So you need to write a good prompt, and then you need to aggressively verify the output.
If you have AI copy on the page, that's the new lorem ipsum. It's a placeholder, but you usually don't want it to be the final version. You want to clean up the output of AI.
AI is helpful for data analysis, for frontend code, for images and for videos. In all those cases you can instantly evaluate the results with your eye. You essentially have built-in GPUs, so that's fast. But for backend code, for systems programming, and for smart contracts, you need to verify the logic step by step.
In general, the right amount of AI use is neither 0% nor 100%. If you're at 0%, you're too slow. And if you're at 100%, it's all slop. The right amount is situational, but 100% AI currently doesn't generate optimal results.
So: when writing your code, feel free to use as much or as little AI as you want to generate it. However, your output should not be raw AI without your thorough review and verification. After all, you're putting your name on it, and you're responsible for ensuring that the result is high quality. We want a fully-baked result, not a half-baked result.
With that said, here are some tools you may find helpful, both AI and not:
If you end up posting your code submission to your personal social media, please link back to this ns.com task page to tell your followers this was an entry for an Earn competition. This helps you, because your followers may upvote your submission. And it helps your followers, because they can submit entries of their own on other prizes.