Skip to content

Comment on Recursive LLM promptsparent

Comments

It's be nice if I could solve any problem by speccing it out in its entirety and then just implement. In reality, I have to iterate and course correct, as do agentic flows. You're right that the AI labs love it though, iterating like that is expensive.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.