Post

Hendrik Mans
@hmans.dev

Hashing things out with an LLM works not because the LLM magically has all the answers, but because this process forces you to structure and reflect on your thoughts in a way that helps you find the answer yourself. Software developers worth their salt know this as "rubber ducking"

18 reposts161 likes

Andri Óskarsson
@andri.dk

Yes! Asking the right questions. Ask domain experts, peers, architect - but it's also very valuable to ask the LLMs questions. I think that separates the "vibers" from those who get seriously impressive results with LLMs.

codewright
@codewright.bsky.social

In my mind, I separate between people who mostly one-shot stuff and people who mostly work on iterative refinement of larger projects. The results of both can be very impressive, but the later offers more room for an experienced human to steer the ship towards more distanced destinations.

Steve
@stevelikesrobots.bsky.social

Yes. I “vibe code” but my projects are the result of weeks of effort and eventually it gets to a point where my stuff is good enough for me to use in my daily life.

AI Dive
@aidive.bsky.social

Claude Code's superpowers plugin makes it a gate: its brainstorming skill asks questions one at a time and refuses to write code before the design is agreed.

Will
@thealpacalypse.bsky.social

The duck also can be asked to be adversarial without being a dick. The duck will not get bored or annoyed. The duck has data-driven insight into my biases. The duck has the knowledge of 10,000 junior devs that all read the newest ycomb blog slop. The duck lived through the mainframe era