| I think the anxiety often comes down to how we choose to use such tools. I’ve started following a similar philosophy to keep my skills sharp: - I use AI primarily for a rough prototype, design elements, UI scaffolding, and quick proof-of-concepts. It’s great for seeing if an idea might work. But for the final product, especially the critical logic I write it by hand with some assistance AI autocomplete etc. - Even when AI generates a snippet, I force myself to audit every line. It often feels unnecessary or slow, but it’s the only way to ensure I actually understand the codebase. If you can’t explain why a line of code is there, you’ve lost control of your software. Might work for now and might actually be ok for the product you are building but generally not a fan of it. - This shift allows me to stop sweating the "dumb" recurring tasks and focus on the big-picture architecture. AI handles some boilerplate. I handle the systems design. At the end of the day, I still want to be the one who built the software, not just the one who prompted it into existence. But the same way i don't want to set every bit in the software code for a web-app and use abstract languages i think it has it's spot to get more done. Also of course instead of reading through hours of documentations i can read through a documentation summary that's most of the time good enough. |