If you are already a good engineer, you get a massive advantage by utilizing an LLM that can hammer out code at the speed of light.
You still have to know what you're (or rather your agent) doing - if you don't, you're just shipping unmaintainable, insecure stuff that endangers you and your users.
We can build in minutes what once took hours.
But here's the real question: Are we becoming better engineers, or just better at prompting AI? 🤖
Let's hear your honest opinions.
commit 660430f
The problem is that in order to maintain your level of mastery and to become a better engineer, you need to practice wading into the details and thinking about the sharp edges. An AI likes to obscure those things.
And how do we own what is built!? In the sense of understanding what/where/how
Turns out, the good standard of code organization, styles are first class decisions for this?