[smt88]

I sometimes meet devs who are "using it wrong" with under-baked prompts.

But mostly my experience is that people who regularly get good output from AI coding tools fall into these buckets:

A) Very limited scope (e.g. single, simple method with defined input/output in context)

B) Aren't experienced enough in the target domain to see the problems with the AI's output (let's call this "slop blindness")

C) Use AI to force multiple iterations of the same prompt to "shake out the bugs" automatically instead of using the dev's time

I don't see many cases outside of this.

[bsder]

> B) Aren't experienced enough in the target domain to see the problems with the AI's output (let's call this "slop blindness")

Oh, boy, this. For example, I often use whatever AI I have to adjust my Nix files because the documentation for Nix is so horrible. Sure, it's slop, but it gets me working again and back to what I'm supposed to be doing instead of farting with Nix.

I would also argue:

D) The fact that an AI can do the task indicates that something with the task is broken.

If an AI can do the task well, there is something fundamentally wrong. Either the abstractions are broken, the documentation is horrible, the task is pure boilerplate, etc.

[JamesSwift]

D) Understand and use context creatively. They know when to start new conversations, and how to use the filesystem as context storage.

[smt88]

Yeah except I do that with Claude Code, and my output is still shit most of the time. It might save me a little time or typing, but it definitely needs hand-editing. The people who say Claude is a junior dev (at best) are right.

That's why I think a lot of people who think it's a miracle probably aren't experienced enough to see the bugs.