←back to thread

Coding with LLMs in the summer of 2025 – an update

(antirez.com)
600 points antirez | 5 comments | 20 Jul 25 11:04 UTC | HN request time: 0s | source
Show context
theodorewiles ◴[20 Jul 25 13:32 UTC] No.44625055[source]▶
My question on all of the “can’t work with big codebases” is how would a codebase that was designed for an LLM look like? Composed of many many small functions that can be composed together?
replies(5): >>44625070 #>>44625105 #>>44625128 #>>44625450 #>>44625922 #
1. antirez ◴[20 Jul 25 13:33 UTC] No.44625070[source]▶
I believe it’s the same as for humans: different files implementing different parts of the system with good interfaces and sensible boundaries.
replies(2): >>44625329 #>>44626082 #
2. dkdcio ◴[20 Jul 25 14:07 UTC] No.44625329[source]▶
this is a common pattern I see -- if your codebase is confusing for LLMs, it's probably confusing for people too
replies(1): >>44626026 #
3. physicles ◴[20 Jul 25 15:22 UTC] No.44626026[source]▶
This fact is one of the most pleasant surprises I’ve had during this AI wave. Finally, a concrete reason to care about your docs and your code quality.
replies(1): >>44634960 #
4. afro88 ◴[20 Jul 25 15:26 UTC] No.44626082[source]▶
Well documented helps a lot too.

You can use an LLM to help document a codebase, but it's still an arduous task because you do need to review and fix up the generated docs. It will make, sometimes glaring sometimes subtle, mistakes. And you want your documentation to provide accuracy rather than double down on or even introduce misunderstanding.

5. aitchnyu ◴[21 Jul 25 13:41 UTC] No.44634960{3}[source]▶
"What helps the human helps the AI" in https://blog.nilenso.com/blog/2025/05/29/ai-assisted-coding/

In future I'll go "In the name of our new darling bot, let us unit test and refactor this complicated thing".