3 ms·
> The question is whether someone who can't program at all can make something that works well with LLMs Is that where the goalposts are now?
by afro88 1y ago
> The question is whether someone who can't program at all can make something that works well with LLMs
Is that where the goalposts are now?
- stavros 1y agoNo, this is where this discussion is, since the top comment. Please go elsewhere for straw men.
- afro88 1y agoNo one mentioned anything about "people who can't program at all" until your comment. Up until then the discussion was about using LLMs for production ready code. It's a given that people working on production systems know how to program.
- stavros 1y agoThat is what "auto-edits" (in the first comment) and "vibe coding" (in the second comment) mean.
- afro88 1y agoThe bit you're missing is that people who can program also use "auto edits" and "vibe coding" when it fits the problem at hand
- stavros 1y agoAs someone who can program and used vibe coding, when they use vibe coding and don't look at the output, how do they ensure errors don't compound?
- afro88 1y agoI'll give you that - you can't vibe code, not look at the output, and push to main of a production codebase. Same as writing code normally. But you can vibe code and auto-edit your way to a good PR. But you're reviewing the code being output when the agent finishes each task. Maybe that's not pure vibe coding?
- stavros 1y agoThe definition of vibe coding is that you aren't looking at the code. That's what's never worked for me.