GitHub CEO: manual coding remains key despite AI boom

"He warned that depending solely on automated agents could lead to inefficiencies. For instance, spending too much time explaining simple changes in natural language instead of editing the code directly."

Lots of changes where describing them in English takes longer than just performing the change. I think the most effective workflow with AI agents will be a sort of active back-and-forth.

Yeah, I can’t count the number of times I’ve thought about a change, explained it in natural language, pressed enter, then realized I’ve already arrived at the exact change I need to apply just by thinking it through. Oftentimes I even beat the agent at editing it, if it’s a context-heavy change.

How active are you ok with/want? I've just joined an agent tooling startup (yesh...I wrote that huh...) - and it's something we talk a lot about internally, we're thinking it's fine to do back and forth, tell it frankly it's not doing it right, etc, but some stuff might be annoying? Do you have a sense of how this might work to your mind? Thanks! :)

Rubber duck. I’ve kept one on my desk for over a decade. It was also like a dollar, which is more than I’ve spent on LLMs. :)

https://en.m.wikipedia.org/wiki/Rubber_duck_debugging