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

An AI Vibe Coding Horror Story

tobru.ch

Post reply on HN
Print view View original

201–210 of 224 posts

Re: An AI Vibe Coding Horror Story

#201

Post by zzyzxd »

Vibe coding is fun, but I can't trust it to make any serious decisions. Like, it knows what's the best way to do a thing, but when encounters challenges, it started to make all kinds of excuses to cut corners, just like humans. "but honestly, it's cluster internal traffic so unencrypted traffic is fine". "Given the urgency and tight timeline, your best option is bypassing the pipeline and deploying it manually". "Per my research, XXX also did this so you are fine".

If I don't have disciplines or principles, or if I am just technically incompetent, its suggestions would sound so reasonable.

Re: An AI Vibe Coding Horror Story

#202

Post by consumer451 »

ZephyrBlu wrote: post #167
consumer451 wrote: post #51

Earlier quoted context omitted.

Oh man, I have gone off on rants about software "engineering" here in the past. My first office job was as an AutoCAD/network admin at a large Civil and Structural engineering firm. I saw how seriously real engineering is taken. When I brought up your argument to my FAANG employed sibling, he said "well, what would it take to be a real software engineer in your mind!??" My response was, and always will be: "When ther…

People like to make this point, but traditional engineering has the opposite problem: insanely overwrought processes and box-checking that exists for no reason and slows everything down to a snail's pace. Yes there are safety-critical parts, but they surrounded by a ton of bullshit. It's also absurd to think that there is no company which does genuine software "engineering". If you break ads at Google/Meta, streaming…

That's quite the take. Throughout human history there were lots of instances of vibe-engineering and vibe-architecting, in the physical world.

Since the failings of not doing proper engineering is far more evident, the reasons for the "insanely overwrought processes and box-checking that exists for no reason and slows everything down to a snail's pace" go back to the earliest written law, AKA the Code of Hammurabi, circa 1754 BC! These rules are part of the core of our functional civilization.

Examples:

- Law 229 (Death of Owner): If a house collapses and kills the owner, the builder is put to death.

- Law 230 (Death of Owner’s Son): If the collapse kills the owner's son, the builder's son is put to death.

- Law 232 (Property Damage): The builder must replace any destroyed property and rebuild the collapsed house at their own expense.

- Law 233 (Structural Defects): If a wall "shifts" or is not built properly before completion, the builder must strengthen or repair it using their own silver/means.

Re: An AI Vibe Coding Horror Story

#203

Post by intended »

funkyfourier wrote: post #53
delis-thumbs-7e wrote: post #4

Meanwhile on Linkedin… Every sales bozo with zero technical understanding is screaming top of their virtual lungs that evrything must be done with AI and it is solution to every layoff, economic problem, everything. It is just a matter of time when something really really bad happens.

The Hindenburg of coding.

Nice. Another one I read on HN is “we are building hell on earth.”

Re: An AI Vibe Coding Horror Story

#204

Post by oxag3n »

What if that AI was trained to hide implementation as much as possible? E.g. by making the client as thin as possible, using OAuth for authentication, following some robust template, would it be better?

Avoiding such low-hanging-fruit disasters is doable and major players have incentives to fix it.

This however wouldn't change the goal - replace all software engineers with DIY nephews with LLM in their hands.

Even worse - people start thinking it's infeasible to learn hard things, because it prevents you from moving fast, from one domain to another.

My area of expertise is cryptography-adjacent. Non-trivial fraction of students think that cryptography and security is a dead-end skill as all companies in this area are being replaced with AI. I asked them to implement web-bot-auth as simple as possible, because I know how AI can read specification and follow it.

Re: An AI Vibe Coding Horror Story

#205

Post by dwaltrip »

ramon156 wrote: post #9
spaniard89277 wrote: post #7

I did something similar to a local company here in Spain. Not medical, but a small insurance company. Believe it or not, yes, they vibecoded their CRM. I sent them an email and they threatened to sue me. I was a bit in shock from such dumb response, but I guess some people only learn the hard way, so I filed a report to the AEPD (Data protection agency in Spain) for starters, known to be brutal. I've also sent them a…

You only burn your hand once, unless you're a company, then you never learn.

Ah yes, companies don't have hands. Nothing to burn!

Re: An AI Vibe Coding Horror Story

#206

Post by sajithdilshan »

t43562 wrote: post #196
sajithdilshan wrote: post #135

Earlier quoted context omitted.

> "this is possible and the AI will do it properly you don't have to worry" then it's not negligence really Then that's lack of due diligence and and any manager is forcing you to ignore that, you should report them to compliance team. You cannot blame everyone else and bear no responsibility for your actions. If you decide to vibe code blindly and ignore all the laws and standards, then that was your decision and yo…

Management know how to keep themselves safe! They watch your AI usage and pressure you to use more. They tell you that "of course you must check the code" but of course if you do then they can start telling you your performance isn't good enough. "Fred is much faster than you" and of course Fred says he checks everything but does he really? To do code reviews he just uses an AI anyhow so nobody catches what the AI do…

You do realize that bad management would still be bad management regardless of AI right? "Fred is much faster than you" has been used by management way before AI was there will aways be scenarios like that regardless of AI or any other tool or technology. Blaming bad management on advances on AI is quite naive and it actually doesn't achieve anything. If the management is the problem then you have to address that and fix that and AI has nothing to do it. Today the management would use AI as an excuse and tomorrow they would use the next hype word.

Re: An AI Vibe Coding Horror Story

#207

Post by krapp »

avazhi wrote: post #199
krapp wrote: post #151

Earlier quoted context omitted.

This is reality now, what do you want?

HN should just have a rule that all content should be human-generated. This post is literally an LLM writing about something an LLM did; it's a bot botting about a bot. Aside from how funny and dystopian that is - just ban it? Just make a rule that submissions require a human author. I don't think solving this is all that complicated, at least for now. It isn't like it's currently difficult to tell what is and isn't…

Almost everything of interest to HN is written by LLM now, at least partially, and if it isn't it will be soon. There already is a rule against comments being LLM generated (which gets ignored all the time) but if we extended that to articles there would be almost nothing to post.

Re: An AI Vibe Coding Horror Story

#208

Post by trick-or-treat »

voidUpdate wrote: post #180
trick-or-treat wrote: post #75

Earlier quoted context omitted.

Sure, but you'd hope that the LLM could understand that too.

And yet it seems it didn't

I'll bet you anything it did. If you ask any recent coding model to do this project is will strongly advise you to give up.

Re: An AI Vibe Coding Horror Story

#209

Post by trick-or-treat »

grey-area wrote: post #105
trick-or-treat wrote: post #68

Earlier quoted context omitted.

> Usually they would just use an off the shelf product and extend it AI does the same thing an agency or dev would do. Those vibe coding platforms have a template for these things which is usually Vite + React with Supabase for the backend, the same as a dev might use because surprise the LLM trained on the dev's work. OP's point is that you're not guaranteed a good outcome hiring an agency or solo dev either, in fac…

Apparently your assumptions about AI are completely wrong, if you read the article it produced terrible code.

Right, because the user provided terrible prompts. A dev / agency wouldn't have done much better.

Re: An AI Vibe Coding Horror Story

#210

Post by dd8601fn »

BitsAndObjects wrote: post #194
consumer451 wrote: post #181

Earlier quoted context omitted.

I used to manually curate a whole set of .md files for specs, implementation logs, docs, etc. I operated like this for a year. In the end, I realized that I was rolling my own crappy version of Jira. One of the key improvements for me when using Jira was that it has well defined patterns for all of these things, and Claude knows all about the various types of Jira tickets, and the patterns to use them. Also, the spec…

I self-host OpenProject - free and a little better than Jira.

This will probably make people laugh, but I just had claude make me one. It’s simpler than jira, but it’s good enough without the 10,000 things I don’t need.

I wouldn’t use it for work but it’s good enough to track my projects, note what’s in each release, has simple user and service account key issuance for api access, user roles and access control, project level configuration for kanban lanes and status mapping, claude can access everything via the api, simple project level document library with live preview markdown editing, etc.

Post reply on HN