Live data from Hacker News

Vibe Coding Is Overrated

transitions.substack.com

21–30 of 93 posts

Re: Vibe Coding Is Overrated

#21
post #19

I have already vibe-coded countless number of scripts and two side projects that were on my "list" for years, these projects would have never seen the light of day if not for Cursor. That's the whole point.

The fact that they could be "vibe coded" means that they (or something similar) probably already existed, but you just didn't know they did.

Re: Vibe Coding Is Overrated

#22
post #12
post #7

Earlier quoted context omitted.

Wasn't this 12 months ago?

It was two months ago, and the prediction was 90% of code written by AI in 3-6 months, and ~100% in 12 months. Maybe he will be right but in a cursed monkey's paw sense where the amount of high quality code stays about the same, but that becomes a rounding error compared to the amount of barely functional spaghetti code being churned out by agents. Similarly ~100% of written word will probably be penned by LLMs soon…

It’s more about the qty of minutes/hours AI can crunch information for you, reducing the context switching or loss of context. A lot of the back and forth will go away, replaced with prompt and planning skills needing to be more specific, for long complex runs as needed, then clean up short runs based on UAT and QA inspection after.

Re: Vibe Coding Is Overrated

#23

It has really good use cases, I have been vibe coding some zsh scripts to automate some processes and it’s been amazing. No I don’t intend to actually learn zsh.

Agree that for scripts and simpler coding stuff it works well. For work I use just the bog standard MS Copilot (the one that is now integrated into Corporate M365/Teams), and it's been quite fantastic for just a general, well-tuned all-purpose LLM. I needed a simple API to update CSV files. A bit of googling led me to the deno-csv library, and I then pointed Copilot to the documentation, and it coded me up a working API after just a bit of back and forth.

Re: Vibe Coding Is Overrated

#24
It's interesting to me how I didn't know where the "vibe coding" term came from originally, and thought someone has decided to push and hype it up really hard as an alternative approach to writing software in general. And now that I see it, it sounds way more reasonable than the way it's been spreading out of context.

I'm sure it will go down in history as some sort of intentional movement, and then someone will occasionally point out "did you know that originally this idea was never meant to be used for serious work, Andrej literally said it was only okay for throwaway weekend projects." Big not-always-good things[1] started this way — by misinterpreting something.

Of course, it's possible that in the future code written this way will actually be maintainable, performant, and secure enough that you could start trusting it, and this will morph into a legit way of writing software, but right now the hype is way ahead of what Andrej meant when he coined the term.

[1]: Examples come to mind: Agile, Scrum, Javascript (2 week project originally), there're definitely more examples.

Re: Vibe Coding Is Overrated

#25
post #21
post #19

I have already vibe-coded countless number of scripts and two side projects that were on my "list" for years, these projects would have never seen the light of day if not for Cursor. That's the whole point.

The fact that they could be "vibe coded" means that they (or something similar) probably already existed, but you just didn't know they did.

But the amount of time saved because one could vibe code it rather than search through Google’s terrible results is immense and a benefit in and of itself.

Re: Vibe Coding Is Overrated

#26

Vibe Coding is the new offshoring - the difference is that basically anyone can do it.

This has been my conclusion as well. A lot of people complaining about the quality, bugginess and general fragility of vibe coded codebases are comparing it with their own personal, thoughtful, (and expensive!) dev process.

Having inherited a couple of hastily written but dirt cheap offshored codebases to fix, I would rather vibe code than try to untangle those garbage code bases.

Vibe coding is terrible compared to the thoughtful output of a seasoned dev, but is significantly cheaper, faster and easier to iterate on than the lowest bidder offshored. Undoubtedly most offshore coders will just vibe code anyway, so in the near term I suspect LLMs will impact these “programmers” the most.

Re: Vibe Coding Is Overrated

#27
post #17

I would guess that for every app that was vibe coded and made $80k in revenue, there are thousands of attempts that went off a cliff or into a wall, generating $0 revenue and wasting a human’s time. So it’s just survivorship bias. Sure, it’s possible to make something useful, but you will probably fail. I’m using AI to increase my productivity, but whenever I’ve vibe coded (not intentionally, just by getting caught u…

Unless I spend a considerable amount of time writing a wall of text to do anything beyond simple tab completion I have yet to see it save me a significant amount of time. The 1 time it helped me come up with a complex solution gets overshadowed by 10 rabbit holes of hallucinated API usage that I now have to sort out, clean up, etc.

Re: Vibe Coding Is Overrated

#28
Claude has helped me a lot to get through my small feature backlog faster.

But if you don't review the code you're asking for hurt. It has literally written stuff like "if pending changes > 20, delete all", etc.

Re: Vibe Coding Is Overrated

#30
post #19

I have already vibe-coded countless number of scripts and two side projects that were on my "list" for years, these projects would have never seen the light of day if not for Cursor. That's the whole point.

Is it. For small scripts like these you can get far with basic search and sources like docs/github/SO. Maybe cursor is more “fun”, “engaging”, and the like, but it’s not like it was a difficult job before.

It is, takes much less energy.
Post reply on HN