This was my exact first experience trying to build anything of significant size.
Several projects later, I can work with the AI now to produce several thousand lines of code, I understand it, it is what I asked for, has the architecture I wanted, is not bloated and is not buggy.
Spend time planning, figuring out the architecture you want, refining, breaking things down into smaller implementable features with implementation plans for each feature. Review these things aggressively. Make sure they align with your expectations.
Then get it to start writing code. Writing tests for everything. Include detailed runtime logging. Read what it produces, understand it and redo it if it wasn't what you expected. Your experience will be vastly superior to that first attempt.
Greenfields projects where this was possible only make up a tiny fraction of the stuff devs work on. Good devs find it easier, quicker, and more reliable writing the code themselves then explaining the context of the situation to someone.
At some point the idea of coupling your buisness to a set of tokens to an ai company that cant make profit on the tokens they sell to you - is going to crash. AI companies have an incentive to generate LoC because thats how they get paid. LoC = context. Your greenfields project turns into a unmaintainable bloated legacy codebase sooner then you expect. Like every other tech innovation in the last 10yrs, after the honeymoon period ends so do those perks that you enjoy today. Just imagine in 5years that time your AI will only recommend libraries they can yeild off, and will charge an even more premium price then you can imagine today to enjoy the benifits of ad free code generation. By that time your skills could be replicated by AI, because you keep training their AIs, and they will be worthless. Only those devs who maintain there understanding of code will be worth anything, and they will be brought in to cleanup the mess.
Several projects later, I can work with the AI now to produce several thousand lines of code, I understand it, it is what I asked for, has the architecture I wanted, is not bloated and is not buggy.
Spend time planning, figuring out the architecture you want, refining, breaking things down into smaller implementable features with implementation plans for each feature. Review these things aggressively. Make sure they align with your expectations.
Then get it to start writing code. Writing tests for everything. Include detailed runtime logging. Read what it produces, understand it and redo it if it wasn't what you expected. Your experience will be vastly superior to that first attempt.