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

Ask HN: What was your "oh shit" moment with GenAI?

news.ycombinator.com

Post reply on HN
Print view View original

981–990 of 1001 posts

Re: Ask HN: What was your "oh shit" moment with GenAI?

#981

Post by GistNoesis »

onraglanroad wrote: post #972
GistNoesis wrote: post #557

More like "oh shit, we are so screwed". It's already a better system administrator than I am. It can run plenty of obscure linux commands, trash the system and maybe restore system state to functional. I was vibe-setting my system permissions with some local qwen3.6 . It was all going well for 30 minutes. Then in between other commands, it made me run a variant of "sudo chmod 644 /usr/bin" Which it explained when the…

You don't need to actually reinstall to fix this. You can boot to single user mode and then you're root. Fix the permissions from there. Even if you've accidentally removed permission from chmod itself, you can use the linker trick to fix that (your LLM can explain that one but use a fresh context so it's not roleplaying where it doesn't know how to fix without a reinstall)

Before booting to the usb, the LLM did offer options to make me boot into a root prompt alternatively to the usb key, (but even there it was not booting properly, the failed services keeping displaying failures on the screen interfering visually with my inputting commands).

But as soon as I turned off the computer, and it no longer booted, I had to switch browser to an other machine to access the LLM and therefore could not access the context or conversation history which was stored in the browser and so the new LLM had no idea of all it had done before except from my prompt where I tried to explained what it had done.

Then from the live usb, the LLM made the situation even less recoverable when it started removing some system file in the hope to restore them cleanly with an apt install, probably because it didn't have a clue of the extent of the damages it needed to repair.

Thanks anyway, I'll try your solution if it happens again.

Re: Ask HN: What was your "oh shit" moment with GenAI?

#982

Post by LearnYouALisp »

ulfw wrote: post #567

My moment was when absolute everything I put into Gemini, ChatGPT et al comes back with a super convincing sounding lie followed by 'Oh you are absolutely right for calling me out on this'. It's a fucking joke and most people are blinded by it sounding very sophisticated and convincing

In what Stage of the Big Con are we?

http://www.amyreading.com/the-9-stages-of-the-big-con.html

Re: Ask HN: What was your "oh shit" moment with GenAI?

#983

Post by samsummer »

I was building a mini-app that could convert an image to a grid of SVG dots of different sizes—not a super complex project, but it involved some image sampling, SVG manipulation, etc. It probably would have taken me at least a day to write by hand, but I one-shotted it in about two minutes.

That was the first time I felt like GenAI would DRASTICALLY change the amount I could build in a day.

Re: Ask HN: What was your "oh shit" moment with GenAI?

#984

Post by vitorbaptistaa »

qnleigh wrote: post #949
vitorbaptistaa wrote: post #276

I am the CTO of a small NGO (10 people total, only 1 other junior Dev at the time). We supported two apps that were built by consultants. They were a mess. NextJS, React, about 4 micro services for a site that had 50 users per WEEK. I configured a devcontainer with the old codebase and an empty repository and asked Claude to rewrite it as an old school server side rendered Django app. Went to sleep. When I woke up it…

How much did the original apps cost you? That's gonna be direct job replacement for those people.

I wasn't involved at the time, so don't know. I just know that the cost was significant given the NGO budget, not only the initial development but also the ongoing support.

I feel as if we lifted an anchor that was slowing us down.

Re: Ask HN: What was your "oh shit" moment with GenAI?

#985

Post by frankieg33 »

shipping my first piece of software. i was playing the game Mewgenics. i knew why the developers decided to not show all of your cats at once (players will optimize all of the fun out of your game), but i thought that was particularly egregious given you can have upwards 100 cats, and keeping them all in working memory was too taxing.

so within a day, i had claude code make me a PyQt6 app that over the next ~2 months grew in complexity and functionality. I had 600 users. It was tranlated into Chinese, Russian, and Polish.

The speed of taking an idea to production was unbelievable. And I am not technical. I learned a lot about how software development worked. Super glad I did it.

https://github.com/frankieg33/MewgenicsBreedingManager

Re: Ask HN: What was your "oh shit" moment with GenAI?

#986

Post by Divyesh1125 »

N-Krause wrote: post #974

I did write a android app for my SO, rather over engineered - but I had fun programming it and SO uses the app daily. The thing is, the app was in ReactNative and I didn't regularly update the codebase with current ReactNative versions so, to get the app running on a newer Android version I had to either go through the pain of updating the app or rewrite it. I didn't really have the time and energy to rewrite it. So…

[flagged]

Re: Ask HN: What was your "oh shit" moment with GenAI?

#988

Post by darksim905 »

senko wrote: post #569
darksim905 wrote: post #434

Earlier quoted context omitted.

I know this isn't apples to apples, but given that I can't get Copilot or other tools to view a simple profile page on LinkedIn makes me curious/skeptical how this would work in this depth. I'm sure it's possible but I'm curious what the skills and toolchains involved were for you to get all that to work.

LinkedIn in particular is quite aggressively blocking any automated attempts to read or navigate through it. I post quite a lot there and wanted to have a copy of my posts on my blog[0] to preserve them. For a few months I was able to use a headless browser + claude code, then LI wised up and started logging it out, so I had to use a regular Chrome, log in manually and then tell the LLM to take over and slowly go thr…

useful, thanks!

Re: Ask HN: What was your "oh shit" moment with GenAI?

#989

Post by darksim905 »

AussieWog93 wrote: post #450
darksim905 wrote: post #434

Earlier quoted context omitted.

I know this isn't apples to apples, but given that I can't get Copilot or other tools to view a simple profile page on LinkedIn makes me curious/skeptical how this would work in this depth. I'm sure it's possible but I'm curious what the skills and toolchains involved were for you to get all that to work.

Full claude.md is here: https://github.com/EspoTek/.claude/blob/master/CLAUDE.md The skills I have installed are: ``` on frontend-design:frontend-design · plugin · ~90 tok · locked by plugin on agents-sdk · user · ~150 tok on cloudflare · user · ~130 tok on cloudflare-email-service · user · ~180 tok on durable-objects · user · ~130 tok on find-docs · user · ~300 tok on find-skills · user · ~110 tok on sandbox-sdk · u…

Thank you for the in-depth reply!

Re: Ask HN: What was your "oh shit" moment with GenAI?

#990

Post by jgerrish »

jgerrish wrote: post #980
shreddude wrote: post #44

I could go on and on, but Claude recently decompiled the firmware of my camper van, documented all the CAN interfaces, then programmed an ESP32 module to talk to the van’s integrated systems (power, HVAC, lighting, tanks). That sort of embedded systems integration is completely out of my wheelhouse. I honestly don’t understand AI naysayers. I use Claude every day both professionally as a Solution Architect and person…

> I honestly don’t understand AI naysayers. I use Claude every day both professionally as a Solution Architect and personally in a variety of projects I simply could not have ever approached alone. My brother works in wildlife trapping and management. I've been brainstorming and prototyping ESP32 sensors and mechanics for traps and educational devices with him. I probably won't end up doing the work with him, but I w…

I'm sorry, I said I probably won't work with them. That shouldn't be a probably, it's a won't work with them.

I hope they are inspired. I hope others besides me can advocate for UBI which would have lowered the pressure for irresponsible AI use. I hope I have easy healthcare options with a doctor without making it my adovcacy or story.

Post reply on HN