Hacker News Reader

Comment by rienbdj | Hacker News Reader

Effective AI use (in my experience) has human doing the load bearing parts by hand (schemas, api spec, overall architecture, domain types) then AI fills in the blanks.

View on HN

Could you briefly describe your workflow for doing that or give a pointer to a blog you wrote/like that aligns with the process? Thanks in any case, happy designing ;-)

I'm not who you asked, but my workflow is the same it always was, just that one tool in the process is an LLM. That's it.

If I find a task to be annoying and repetetive, I use an LLM to do it, then validate it. The quality of my work only improves, and so does the speed, and I don't lose my skill over time.

I am exploring ways to document the design for the agent to read and update. What makes it difficult is the lack of structure. Spec writing is not my core skill. Schemas and APIs are easier, there are declarative ways to document them. Runtime concepts and workflows have less structure and writing prose seems so unstructured for my taste. Formal languages are too rigid. But I could not find a better way.

For me writing manpages as a spec worked well. While the format is unintuitive and tedious (at least to me), I used AI to format the manpages from my descriptions and rendered to PDF to review. Though it's been mainly useful for CLI utilities.

Once I got the manpages ready and the way I wanted, I just stored those in an empty project folder and told the agent to implement as specified in a language of my choosing.

Maybe for other development scenarios like GUI there is some "native" way to spec the project in a common format like this.

Thanks, that sounds like a good direction to try.

Reply on news.ycombinator.com