I recently (2 days ago) noticed i enjoy it a lot more if i use a super fast model with low reasoning (gpt 6 luna low effort + fast mode). That way, i can still have hands on the entire time and dont have to wait around for 20minutes for it to finish up and have made a bunch of decisions without me.
Also, LLMs are first and foremost excellent at reading ultra fast. Makes it excellent for summarizing and re-representing modules of your code
I’ve also found that fast cheap models are more sustainable because they make it easier to keep technical debt under control. Partially because I’m keeping sustained attention in one place. And partially because small fast models with low reasoning don’t seem to love technical debt either and will (unintentionally) start to give you negative feedback signals when you’re letting things get messy.
humlex · · focus · HN ↗
Also, LLMs are first and foremost excellent at reading ultra fast. Makes it excellent for summarizing and re-representing modules of your code
coffeefirst · · focus · HN ↗
Frenetic multitasking is for suckers.
bunderbunder · · focus · HN ↗