Ask HN: If AI writes the code, what matters?
AI is getting pretty good at writing code. Not all of it, obviously, but enough that I keep wondering what this does to the job of a software engineer.
If writing the code is no longer the hard part, what is?
Understanding the problem? System design? Knowing the product and users? Making good decisions about what to build?
What do you think will make someone a really good software engineer in the next few years?
4 comments
[ 0.26 ms ] story [ 3.1 ms ] threadAnd just say: LLMs only amplify the the knowledge you have, even the best models o use like Fable still suffer from promoting false narrative as a chart progresses. Often it’s stuff that can be ignored like the “not X but Y” crap it dumps because its reasoning had assumptions it invalidated. Sometimes it’s directly in your system architecture, because you never expressed preferences for solved foundational issues, you end up with generic http handler setups or whatever the hot web thing is today.
Takes knowledge and lots of it to really be on top of when these things go down a failure mode path.
I would say that the act of writing the code wasn't the hard part ever and love to cite this post: https://jaylittle.com/post/view/2023/4/low-code-software-dev...
> In both the AI Chatbot and the Low Code tool scenarios, the solutions each promise a shortcut around the complexity as perceived by a non-practitioner. That’s the essence of the trap. Practitioners know that the writing of the code is merely the last step in a long process that involves a lot of thinking, discussion and planning. The code is generally the end result and producing it is relatively easy once you truly understand the problem at hand.