Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The 100 hour gap between a vibecoded prototype and a working product

kanfa.macbudkowski.com

Post reply on HN
Print view View original

341–350 of 362 posts

Re: The 100 hour gap between a vibecoded prototype and a working product

#341

Post by nemo44x »

tossandthrow wrote: post #90
nemo44x wrote: post #3

The 80/20 rule doesn’t go away. I am an AI true believer and I appreciate how fast we can get from nothing to 80% but the last “20%” still takes 80%+ of the time. The old rules still apply mainly.

Yes, so 80% of 100 hours is considerably less than 80% of 600 hours

You get 80% done in 20% of the time. The LLM shrinks that 20%. So a 100 task maybe takes 5 hours instead of 20 which is great. But the remaining 80 hours are not as improved. So a 100hr job takes ~85 hours which is very good.

This is in-line with Googles study showing about a 10% productivity increase and other research I’ve read. I suspect this will increase with more integrations and workflow adaptations.

But even after power tools changed how quickly carpenters can frame and rough-in a house, the finishing work (which uses power tools too) still takes the majority of the time.

Re: The 100 hour gap between a vibecoded prototype and a working product

#342

Post by devld »

dehrmann wrote: post #311
devld wrote: post #221

My non-technical client has totally vibe coded a SaaS prototype with lots of features, way bigger product than OP and it sort of works. They spent like 200 hours on it. I wonder what would have been the time needed to clean it up and approve it is secure. I declined to work on it, as I was not sure if it's even possible or if it would be better to rewrite the entire thing from scratch with better prompts. I was not t…

> I declined to work on it, as I was not sure if it's even possible or if it would be better to rewrite the entire thing from scratch with better prompts. This is a bit of an unknown right now. If you get a working prototype, but need to productionize it, make sure it can scale, and get it looked over with a security mineset, how long it might take isn't clear, so finding someone who will do that is hard.

It sounds like a really crappy deal to a contractor. In the eyes of the client, you are basically a janitor and Claude is like a superstar that built the whole thing, while the truth is reviewing such code is harder than reviewing and quality approving one's own work. I hope it proves out to be uneconomical and we will be just rewriting the vibe code from scratch.

Re: The 100 hour gap between a vibecoded prototype and a working product

#343

Post by upflag »

The 100 hours are real and they're heavily back-loaded toward things that aren't features. In my experience the biggest gap isn't code quality, it's operational awareness. You can refactor the code, add types, write tests. But the part that kills you is when something breaks silently in production and you find out from a user three days later. I had a single-character typo in a deploy (traditional code, not AI) that cost $30K because we discovered it from revenue numbers, not an alert. The fix isn't reviewing every line, it's making sure you know when key flows stop working.

Re: The 100 hour gap between a vibecoded prototype and a working product

#344

Post by youknownothing »

mordechai9000 wrote: post #180
youknownothing wrote: post #138

Earlier quoted context omitted.

Yeah, I really wonder if someone would trust to do their taxes in a vibe-coded version of Turbotax...

Do you really need Turbotax? Just feed it the tax code, your financial data, and the relevant forms and it should be good to go. Now we have freed up the labor of accountants so they can go be productive in another segment of society. /s

well, you still need to drive some data: say you purchased some food, was that a personal purchase or was it business-related (e.g. a meal with a client). Software can help with that decision making.

But even if human interaction wasn't needed, the question still stands, would you trust an LLM to calculate your tax liability just by feeding it data?

Re: The 100 hour gap between a vibecoded prototype and a working product

#345

Post by jmt710 »

alexpotato wrote: post #10

I work as a DevOps/SRE and have been doing it FinTech (bank, hedge funds, startups) and Crypto (L1 chain) for almost 20 years. My thoughts on vibe coding vs production code: - vibe coding can 100% get you to a PoC/MVP probably 10x faster than pre LLMs - This is partly b/c it is good at things I'm not good at (e.g. front end design) - But then I need to go in and double check performance, correctness, information flow…

Do you ever think that maybe your biased? I ask because I get the feeling from a lot of professional programmers that they feel like they are better and smarter than everyone and everything else. No matter how good an LLM or AI in general gets at programming task, people who make a living programming will always have a problem with it. There is going to come a time when you're going to be obsolete. I hate to say it, but it's coming and the hostility twords the tech isn't going to save your job.

Re: The 100 hour gap between a vibecoded prototype and a working product

#346

Post by wierdbytes »

Jensson wrote: post #200
wierdbytes wrote: post #141

Earlier quoted context omitted.

> Where are all of the actual shipping vibecoded products that millions of users are using? Claude Code and OpenClaw - they are vibecoded. And I believe more coming.

Claude Code is not vibecoded, it is made using Claude Code but it is not vibecoded using Claude Code.

kind of - https://x.com/bcherny/status/2004897269674639461

Re: The 100 hour gap between a vibecoded prototype and a working product

#347

Post by tracker1 »

matt_heimer wrote: post #45
alexpotato wrote: post #10

I work as a DevOps/SRE and have been doing it FinTech (bank, hedge funds, startups) and Crypto (L1 chain) for almost 20 years. My thoughts on vibe coding vs production code: - vibe coding can 100% get you to a PoC/MVP probably 10x faster than pre LLMs - This is partly b/c it is good at things I'm not good at (e.g. front end design) - But then I need to go in and double check performance, correctness, information flow…

I'm building a Java HFT engine and the amount of things AI gets wrong is eye opening. If I didn't benchmark everything I'd end up with much less optimized solution. Examples: AI really wants to use Project Panama (FFM) and while that can be significantly faster than traditional OO approaches it is almost never the best. And I'm not taking about using deprecated Unsafe calls, I'm talking about using primative arrays b…

Maybe a silly question, but why Java? As a C# guy, my experience with AI is it hasn't been great with it, and I'd suspect similar for Java. I'd probably go with Rust, which my own efforts with AI has done really well with, even if I'm far from a Rust expert.

Re: The 100 hour gap between a vibecoded prototype and a working product

#348

Post by tracker1 »

ThrowawayR2 wrote: post #298
withinboredom wrote: post #199

Earlier quoted context omitted.

We were working on translations for Arabic and in the spec it said to use "Arabic numerals" for numbers. Our PM said that "according to ChatGPT that means we need to use Arabic script numbers, not Arabic numerals". It took a lot of back-and-forths with her to convince her that the numbers she uses every day are "Arabic numerals". Even the author of the spec could barely convince her -- it took a meeting with the Arab…

Kind of a tangent but that did make me curious about how numbers are written in Arabic: https://en.wikipedia.org/wiki/Eastern_Arabic_numerals

I guess "Western Arabic" would have been more precise.

Re: The 100 hour gap between a vibecoded prototype and a working product

#349

Post by spacecadet »

qsera wrote: post #227
spacecadet wrote: post #144

Earlier quoted context omitted.

I realize this sounds one sided. Ive also founded companies and worked across the range of startup to faang. Everything has changed... For the better if you ask me.

You are invested in some kind of AI start up, right?

No. I no longer believe in VC or PE. Evils that hurt society more then benefit...

Re: The 100 hour gap between a vibecoded prototype and a working product

#350

Post by spacecadet »

qsera wrote: post #220
spacecadet wrote: post #142

Earlier quoted context omitted.

Do you have a race car or race team? Happy to onboard you, otherwise, not here.

No, but I think I got my answer.

unlikely, people tend to assume wrong about me... as an outlier and all...
Post reply on HN