- cross-posted to:
- linux@lemmy.ml
- cross-posted to:
- linux@lemmy.ml
You must log in or register to comment.
Here is the link to the PR: https://github.com/void-linux/void-packages/pull/62482
Please stop citing from a third-party news site!
A maintainer. Apparently went into a hissy fit because people pointed out to him that he broke the rules. That said, the critics probably weren’t very nice to him either.
Ah, Linux drama.
BTW, the rules sound sensible to me:
"In this section, “AI” refers to generative artificial intelligence tools, including Large Language Models and similar technologies.
All contributions are expected to be made by humans. AI tools may be used for research and learning, but all content in contributions must originate from and be understood by the contributor. This includes code, documentation, issues, security reports, pull request descriptions, and comments in all Void Linux community spaces.
ALL usage of AI tools MUST be disclosed.
Do not bring AI review tools into pull requests.
Usage of AI-powered human-language translation tools is acceptable to facilitate contributions across language barriers."
Esp. this part: “contributions must (…) be understood by the contributor.”
I’m in a smaller community that sees a lot of AI-supported app development recently and while the outcome can occasionally be good, I’m not sure it’s worth taking those few good apples with the flood of vibe coded projects, and their “developers” that tend to get very defensive when simply asked about things.
Not saying this is the case here though.
but all content in contributions must originate from and be understood by the contributor.
why did you omit the important part of that quote
Probably because it isn’t actually the important part. And also because he did, in fact, quote it in full before repeating the part he found important.
yeah… but in my opinion it is the important part… imo it is okay to use an llm to get a reference implementation and then you refactor it into aworking piece that you understand and comment afterwards
Okay, the “must be understood” is the important part when discussing what’s good about the rule, and the “must originate from” is the important part when discussing what’s problematic about the rule.
I’m so tired of “Ai big number” headlines. 100+ packages - but how many good ones? (Genuinely, I don’t know, as I don’t the recognize those listed.)
We need projects that diligently keep the slop out. We don’t need every protect to do so, but it is critical that some do.
We have Meta and Microsoft to provide quick-to-market low quality shit.
The version of Linux that matters is the slow thoughtful good software.
I’m not mad at slop slingers, aa long as they don’t secretly inject slop risk where not welcome.
Edit: Thanks for responses.
For anyone who has read this far, I encourage seeing responses below for informative and nuanced thoughts on this.
- AI is here to stay. It’ll never go away, unless there is no more electricity on the planet. So, we should all come to terms with it and use it the best way we can.
- we shouldn’t just use its code blindly. I’ve vibecoded a couple of personal projects, and I can confidently tell you that the code isn’t the best and, at times, it can be massively messy and even shit, but it works. I didn’t care because those were personal projects that have zero risk for me. We all need to be in a mindset of agreement on certain rules. Because fighting over it won’t do anyone any good. This is reality and will still be reality even if we fought over it
I dunno. I feel like if this tech was so good the results would speak for themself. Instead I just get told that I must use it and that it is here to stay and other constant propaganda.
I’ve vibecoded a couple of personal projects, and I can confidently tell you that the code isn’t the best and, at times, it can be massively messy and even shit, but it works
as someone who doesn’t code, this terrifies me. We’re all kinda hoping that while we don’t exactly know what a program is doing, at least there was a human who wrote it to do something. Once we let AI write complex code that you guys don’t completely understand, how will we know what the hell is even going on??
I realise this might make me sound super dumb but honestly, as someone who is interested but simply can’t grasp the abstract nature of programming, this is how I feel when I read comments like these.
As someone who does code, this is a very valid concern you have. A lot of software is being built right now that not a single human actually understands. On some lever it could be possible to understand but that doesn’t change the fact no one does actually understand it.
To be clear, I’m not saying it’s so advanced that we don’t understand it. The problem is it’s such a giant pile of shit that we don’t understand it.
AI is here to stay. It’ll never go away, unless there is no more electricity on the planet.
‘AI’ doesn’t exist, at least when that term is meant to reference Artificial Intelligence in relation to LLMs, as they are intrinsically incapable of possessing or providing comprehension.
LLMs aren’t going away, but that doesn’t translate to an expectation that we should all have to be exposed to their explosively imprecise slop ingestion and excretions.
So, we should all come to terms with it and use it the best way we can.
This fascistic, rape culture rhetoric further diminishes arguments of its inevitability. If that’s your kink, go ahead and debase yourself, but just like religious dogma, it ceases to be acceptable or tolerable the moment it’s imposed on those who don’t, or can’t, provide explicit, informed consent.
Murder is here to stay; murder can solve many problems; murder is never going away, unless there are no more humans on the planet; we should all come to terms with it and use it the best way we can
We shouldnt just use it to solve problems blindly. It can have terrible consequences if not used carefully
We all need to be in a mindset of agreement on certain rules. Fighting over it wont do anyone any good; this is reality and will still be reality even if we fought over it
AI is here to stay. It’ll never go away, unless there is no more electricity on the planet. So, we should all come to terms with it and use it the best way we can.
Or choose not to, just because it’s there doesn’t make it mandatory or even a good idea.
Cool. Let the slop users go join the fascist DHH in Omacom. Then again, I’m giving a much closer look at Haiku & BSD again since a slop generator was used in the Linux kernel.
Then again, I’m giving a much closer look at Haiku & BSD again since a slop generator was used in the Linux kernel.
As far as I understand, the Linux kernel uses LLMs in practice for code review, not for code generation. And they do not replace human review with it, but use it in addition.
Seeking to understand: this type of code review LLM usage is disallowed in the Void project, correct?
You cannot make code reviews or security audits which originated from an LLM?
The rules are cited above. As I read it, you can make code reviews using LLMs.
But:
- You have to disclose this usage
- You have to describe the results in your own words
- You need to make a fix yourself.
It doesn’t matter. It was allowed at all. Linux is now as complicit with the slopiverse as X, Meta, Alphabet, Nvidia, and the rest of the bastards.
There are just two stances on this, and Linux has picked one. Either you’re in or you’re out. There are no more gray areas when the harm is proven and the damage is actively being done. There’s the moral choice where you denounce the evil, and there’s being evil.
