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

Please Do Not Vibe Fuck Up This Software

github.com

Post reply on HN
Print view View original

101–110 of 534 posts

Re: Please Do Not Vibe Fuck Up This Software

#101

Post by rawoke083600 »

Been thinking of this mental model held by some "oh ai coding is always bad etc etc" (fair we all allowed opinions).

But why are we okey with colleagues making from time to time terrible blunders (hey we all human ). But when ai makes mistakes its a sweeping judgment of "oh ai coding is terrible".

We seen to not include all the amazing code they do right and security bugs they do find..

I feel if it was a human or colleague we be more fair with its failure and balance about his/her achievements also.

Just a thought.ymmv

Re: Please Do Not Vibe Fuck Up This Software

#102

Post by consp »

rawoke083600 wrote: post #101

Been thinking of this mental model held by some "oh ai coding is always bad etc etc" (fair we all allowed opinions). But why are we okey with colleagues making from time to time terrible blunders (hey we all human ). But when ai makes mistakes its a sweeping judgment of "oh ai coding is terrible". We seen to not include all the amazing code they do right and security bugs they do find.. I feel if it was a human or co…

When LLMs make mistakes, it is still the human making the mistake of trusting the LLM. And more often than not "AI" is hailed as costing no effort and being perfect in every way (yes exaggerated), which you can attack when it obviously is going to fail at some point.

Re: Please Do Not Vibe Fuck Up This Software

#103

Post by Quarrel »

ShinyLeftPad wrote: post #81
Quarrel wrote: post #59

Earlier quoted context omitted.

wtf is this comment section? The author of these commits were tridge & claude. What does tridge have to do to convince the open source community that he might be a legit programmer & have a clue? Samba? Whats that? Rsync? Never heard of it. Tivo? No clue (maybe more Australian context here than others, but still). Even the comments on the github issue, are totally devoid of the context that this is a very senior open…

> this is a very senior open source contributer who has maintained this project since he came up with the diff algorithm during his Phd People change. You can be Linus Torvalds for all I care, if one day you wake up and start pushing 9000 line commits created by LLM and with regressions, you're not that person anymore.

But DID anything change?

Of course I know that some people can just becoming psychotic out of nowhere. But why would I assume it?

Re: Please Do Not Vibe Fuck Up This Software

#105

Post by Lerc »

cwillu wrote: post #38
jochem9 wrote: post #28

This is the third HN post I read on this topic. Everytime the same tweet (or whatever it's called for mastodon/bluesky/etc). Did anyone actually debug the issue? Was it caused by poorly generated code, or was it caused a genuine (security) fix that accidentally caused it (potentially even in a way a human would to)?

https://github.com/RsyncProject/rsync/issues/929#issuecommen... has a partial list

That is a list of problems, not causes, which is what was being asked for.

It's possible it's some LLM randomness that caused bugs. That would suggest that some AI hygiene is in order.

If it is because of behaviour changes necessary to fix security issues, then the regressions might be from things that relied on unsafe features.

Do we know of actual specific causes yet?

Re: Please Do Not Vibe Fuck Up This Software

#107

Post by cyclopeanutopia »

bcrosby95 wrote: post #73
mcintyre1994 wrote: post #44

Earlier quoted context omitted.

I’ve never seen Claude do that. It makes the new tests pass by fixing previously unknown bugs in my experience.

I had it do it about a month ago. It changed test data which caused another test to fail and instead of isolating things it decided to flip an assert.

That's because Opus needed vacation and they routed your requests to its less sophisticated cousin, Claude Dynamite. ;)

Re: Please Do Not Vibe Fuck Up This Software

#109

Post by sureglymop »

rawoke083600 wrote: post #101

Been thinking of this mental model held by some "oh ai coding is always bad etc etc" (fair we all allowed opinions). But why are we okey with colleagues making from time to time terrible blunders (hey we all human ). But when ai makes mistakes its a sweeping judgment of "oh ai coding is terrible". We seen to not include all the amazing code they do right and security bugs they do find.. I feel if it was a human or co…

Because AI can't take responsibility. Humans can.

A human can not only learn from their mistakes and blunders but also, until very recently, the social pressure and fear of judgement would push (some) humans to try their best.

Now however, it is less socially acceptable to judge a human for mistakes made with AI coding because we are in a time of experimentation. So the blame has to go towards AI coding. Of course, coding with AI can be acceptable, if the human using the AI is rational and responsible.

But I think the bigger implicit point is actually that perhaps experimentation shouldn't be done on real projects and products as nonchalantly.

Re: Please Do Not Vibe Fuck Up This Software

#110

Post by SyneRyder »

dash2 wrote: post #63
Quarrel wrote: post #59

Earlier quoted context omitted.

wtf is this comment section? The author of these commits were tridge & claude. What does tridge have to do to convince the open source community that he might be a legit programmer & have a clue? Samba? Whats that? Rsync? Never heard of it. Tivo? No clue (maybe more Australian context here than others, but still). Even the comments on the github issue, are totally devoid of the context that this is a very senior open…

[flagged]

Those posts may not have been visible to everyone. The posts you're referencing are hidden for me behind a link "33 Remaining Items (load more)". Without the update, I didn't know to go look for them.

And honestly I noped out of scanning the entire comment thread by about #5 or #6... I could tell there was nothing productive in the remainder of the comments.

Post reply on HN