I had a report for AI slop today. 1m LOC in 3 months. How do you even modify and distribute that? I do not think we had a rule against it.

IMHO if we allow, the community gets drowned out by AI vomit and it is harder to see handcrafted loved projects that are easy to modify and distribute.

My mind is to ban it to preserve open source.

Are folk in agreememt with this line? Would love to hear your thoughts.

Yeah I don't think we need slop projects being posted here. I know a lot of people are using these tools even for personal projects nowadays but tbh I'm not really interested in exploring any project heavily written by AI.

Yes. They are useless. Get rid of them.

I think one of the appealing traits of open source is that you can identify what piece of software and hardware does what (and how, if you're interested).
You can post your exact method, and another person can follow along and replicate the result.

But if AI tells you how to build everything, you don't get the opportunity to learn, and your stack depends on what that particular model thought was best on that particular day. Maintenance also becomes harder if you just did a sudo apt install whatever Claude told me to without knowing what everything does and how it interacts

I had a report for AI slop today. 1m LOC in 3 months. How do you even modify and distribute that?

That makes some sense but does "AI slop" here refer to code with this type of problem or all code made with AI assistance?

Let people choose by their owns.
In my client, I can filter by texts; I already used to hide some topics.
Maybe it's more convenient to include a tag in the title. For example: if your contribution/project is made by or include generated content from a LLM, include the tag [AI]. That way, anyone can hide these contributions by filtering by the text "[AI]".

Yeh I’d prefer no slop. You’d probably also need to clarify what the definition or threshold for this might be. Just the code? Slop docs? LLM used for any aspect whatsoever? Other communities must have a definition that might help. But your goals with it as stated are sound. Let’s not drown in vomit.

This was a problem in !selfhosted@lemmy.world and they've done a good job of dealing with it. You could take a look at that past discussion to see what worked / what didn't.

IMO there is an organized effort behind some of these new accounts pushing AI projects, and you need to deal with it in some form to stop them from coming

Most important step is to decide exactly where you want to draw the line.

Yes

Agreed ! I think a complete AI ban on every element of the project is too strict. But I also have no interest in a massive AI repository with no supervision (and who can supervise a huge AI commit ?)

U also like that for small private projects you can literally contact the developer for questions/suggestions/comments. There is no point if all they did was ask Claude

Let people choose what they want to see.

People here tend to call any hint of LLM use "slop", and arguing on a threshold is a waste of everyone's time that will end up disappointing most. Tag it and let people choose.

I like this idea. Tag it. Mandatory disclosure. People should just be honest about the amount of AI they used and how.

"I vibecoded this awesome thing with 100k LOC in a weekend, check it out" vs "I used AI to help with planning and architecture, but wrote the code myself" are two very different things.

I for one would really appreciate just one single FLOSS community that flat out banned any and all "AI" contributed code instead of following the approach of others. Somebody have the spine to stand out, plz.

I'm not a developer, but I wanna be able to keep up with open source news without having to navigate a slippery slop(e) into clanker dominated territory.

Yea I think so. If the idea is too controversial we could make an open source AI community so everyone is happy.

I'm conflicted. I'm anti-AI due to distrust in capitalism and our pseudo-democratic plutocracies, the profit motive to replace the human workforce, the reckless criminal-incompetence of its implementation, and apparent stupidity of the average hairless primate in understanding that you can never trust a correlation engine... but it's just a tool, and I've also found it incredibly useful in my limited and controlled usage; getting it to write small functions or blocks of code, then amending and composing them myself.

How do we determine what is "heavily written by ai", and all the obvious slop factory bullshit, without banning strong projects who are just leveraging it for legitimate dev work; as a glorified autocomplete, test autofill, etc but still read and validate every line of code?

Like most people, I'm finding it increasingly difficult to find trustworthy projects amongst the see of amateur slop, but I think banning it out-right would likely result in many false positives, and ultimately be counter-productive.

As much as I don't want to depend on metrics, I think a better approach would be to define a bunch of heuristics based on historically-competent projects. Things like defining an upper limit based on LoC vs commit time and number of devs, project age, etc. Ultimately I feel the industry should do what cybersecurity should be doing; moving to some sort of cryptograhically-verified social trust model where engineers/professionals have to build trust in the real world, and take turns in "vouching" for projects and code releases.

I would support such a ban so long as there is some process to prevent it becoming a witch hunt.

I have personally already received one false accusation of using AI in this community and I wouldn't want to continue posting projects here if it becomes an atmosphere of suspicion.

The worst is when you see something you're curious about and ask for clarification, and they just paste your question into the LLM and paste its response back to you.

Agreed. TLDR, you must tag posts there with the following:

Post Tagging

There are three types of tags for project promotional / AI posts.

• [CBH] - Code By Human - A project promotional post with a project that did not use AI in any capacity. No disclosure necessary.
• [AIP] - AI Project - A project promotional post with a project that used AI in development in any capacity. Disclosure is required for how it was used.
• [AIT] - AI Topic - A discussion topic that includes AI. This may be for a discussion on hardware to self-host on, llama.cpp vs ollama, your experience using vulkan, etc.

I'm team "no slop allowed", but you have to define what slop is somehow.

I think there's a bunch of metrics to consider, but it might be difficult to assess as a mod.

I had a report for AI slop today. 1m LOC in 3 months.

I don't understand this, is it human slop?

Would love to hear your thoughts.

AI is here and not going away. Almost every programmer I know, professionally and hobby, uses it to a greater or lesser extent.

We have plenty of crap software in foss that was authored by people, always have, even when it was called public domain. I've certainly shared crap code going back over three decades and I don't need a clanker to make it bad.

Bad software gets rejected or unused because it's not very good. Why can't those same standards apply whoever or whatever created it?

The only thing I think is important, is to be honest.

Thanks for sharing this. Maybe correct tagging could be useful. Makes it easier to ignore but not overly heavy and possibly toxic.

Maybe we use AI to detect the bad AI, and then blame that AI when it doesn't work?

Both the Linux Foundation and the EFF have permissive policies towards AI generated code, even Linus Torvalds accepts it as a tool.

Some FOSS projects that have AI generated code in it:

• The Linux kernel
• Fedora
• Mesa 3D
• Jellyfin
• Firefox
• Chromium
• Homebrew
• OpenSSL
• FFmpeg
• Curl

Should we ban them all? If so, this community would be either dead soon OR it would be hypocritical.

Or because they started pre-AI it's ok now? Shouldn't we care more about disclosure, policies, licensing, code quality and maintainability?

Users should downvote crappy software and crappy devs, regardless of the tool it was written with.

I'm fine with AI. I'm not fine with slop. Ban slop, not LLMs.

You are the same.

LLMs are slop.

You have it backwards. It's better to keep a tight grip on planning and architecture than implementation of individual functions.

You're right, AI is here to stay.

Thank god people like you exist, or we would have also missed the train of the blockchain and NFTs.

And how could this amazing technology that literally doesn't have a single positive point, and has a lot of points ranging from slightly negative to completely immoral and illegal, ever go away? I don't see how, at least Grok said so in the middle of a paragraph telling me that nazis were right and that I should go kill minorities and vote for Trump.

define slop. it means different things to different people.

Experienced developers using AI as a tool and properly understanding their code base are not slop.

AI slop is when someone with no software development experience prompts a piece of software into existence without any understanding of how that piece of software actually works. The project's are usually heavily pushed by their creators in the initial couple of weeks of creation but are then quickly abandoned when the creator finally realizes neither they nor their AI can maintain the project. Essentially these kinds of projects flood the internet with low quality abandonware.

A newly created open source project coded using AI would be fine if the creators could prove that they understand how to maintain the project and keep it going beyond the initially promoted code and that kind of proof can only come with time.

The big problem is the sheer volume of low quality projects that get abandoned will cause the projects that actually have potential to become needles in haystacks.

You are either ignorant or lying. There are many positive points with AI. Just because the negatives outweigh the positives for you doesn't mean it "literally" doesn't have a single positive point.

The Zig language has a no AI contributions policy. That's for the language itself, but a lot of it's community have similar views.

Codeberg has a no AI policy for projects hosted on its servers, so anything hosted there should meet your requirements.

+1 for the option self hosted went with - AI use must be labeled. It allows those of us that don't want to see it to easily filter it out.

I saw that million lines of code in 3 months but also when I saw the post title I really just had a feeling that's what it was to begin with. On one hand it's kinda cool that just about anyone can make a program that does what they want, but on the other side of it I understand LLM generated code and bot submitted pulls are literally killing beloved FLOSS projects. This greatly saddens me. A good friend of mine quit contributing to projects she used to enjoy helping develop because rather than read her code, an LLM the lead dev fed it through claimed her code was AI generated. It's fucked and causing brain rot of the worst kind.

But I also want to acknowledge that in the hands of a professional developer/coder LLMs are a valuable tool. But a person needs to be able to read and understand and verify the generated code.... All this to say is that I think we need a sane and nuanced approach. Let's not start a witch hunt, let's just put labels on boxes and let the community decide what it wants to view or filter out.

A newly created open source project coded using AI would be fine if the creators could prove that they understand how to maintain the project and keep it going beyond the initially promoted code and that kind of proof can only come with time.

Prove to whom, and why?
Who entitles anyone to demand "proof of investment" for a piece of work they have zero contribution to?

The big problem is the sheer volume of low quality projects that get abandoned will cause the projects that actually have potential to become needles in haystacks.

They already are. I don't have numbers, but pretty sure the ratio between useful and useless repos on github is... Not high

I am convinced that, with the right tools, it's possible to damage 40 years of work within a few weeks.

Are you ok, mate?

I thought it was "messy, generated code with little cohesion and only barely functional", but really it's the cohesion part.

@CrypticCoffee@lemmy.ml if you have any questions on those rules, please let me know.

The disclosure was needed because there is a wide variety of LLM use, which you can see in the explanation post as well. There is a rather wide difference between "I used an LLM to summarize my changes in the readme" and "I generated everything here from a single prompt".

The disclosure addresses that part, while the tags allow people to choose what to interact with or what to filter.

Proof to the future user-base they are trying to attract by posting about their project. Proof to the community they are trying to be part of that they are not wasting everyone's time.

Before posting about your project just show that you are making an effort to maintain it, show that you're willing to engage with people and not just pass everyone's questions into an LLM and regurgitate the response.

Don't foster engagement if you're not going to do the bare minimum to maintain the project.

Yes there are tons of repos out there that don't deserve any attention from the community, not just AI slop. But there are far more AI slop projects being posted to communities online where the project is instantly abandoned.

Yup, LLMs are wonderful.

They provide anyone with access to knowledge that only about half the time is misleading in some way, but is just as persuasive as an expert explanation of the truth.

By automating writing, coding, and data analysis, LLMs remove any need for humans to understand how to express nuanced thought, how to clearly and safely execute a technical process, or how to extract insights from large amounts of information.

By automating customer‑service tasks, LLMs remove the last vestige of a human who can understand why a corporate process is failing a customer and find a way for the system to meet their needs. And as a bonus the LLM can confidently declare 'facts' that aren't true.

LLM‑driven tools can assist clinicians with literature summarization (reinforcing training-data bias, eliminating the actual point of scientific papers, and making lots of errors), patient communication (eliminating any human contact and care towards vulnerable patients), and preliminary triage (with extra training data bias and distracting the doctor from reading the not-written-down actual human presenting symptoms in front of them)

Artists, writers, musicians, and game designers use LLMs as co‑creators, generating drafts, ideas, or entire works that rehash old works and remove any pesky 'artistic expression', as the LLM is much faster at making a polished-looking end product, and it's not like people ever come up with new insights during the revision process.

Advanced multilingual models translate and transcribe minority languages, ensuring that no one ever need learn the language of their cultural heritage, as they have an English approximation that's totally good enough.

And the best part is this can scale to billions of people, meaning that no one will ever need to be knowledgeable, or know how to write, code, analyze data, do customer service, read a scientific paper, talk to a patient, observe someone to understand their current state, or have to learn anything about other languages or cultures. Just trust in the often wrong but-who-cares-stop-complaining LLM!

Sorry about the downvotes, Lemmy has a big coding community (naturally, in this comm), and that seems to be the sector where people are the most hypocritical about "AI" usage.

messy, generated code with little cohesion and only barely functional

Okay, you've described my code. Now what about the AI?

What happens if AI gets good enough in like 5-10 years to (largely) no longer fit the created definition of slop? Do we revise the definition to be more restrictive, allow the content, or block AI content all together.

The answer to this question determines if it's really about the quality of the content or if it's more about the fact it is computer generated.

• someone with zero experience using AI to blindly create something barely functional, feeding in errors until it gets a working version. No checking for vulnerabilities, testing, error catching, no thought on architecture or scalability.
• someone carefully using AI in a controlled manner to create open source software. They spend days making sure that is reliable, scalable, secure, and that the project is quality.

You believe these two are them same? That both are equal slop? Quite frankly, I don't care if an AI made most of the code for an open source project as long as it is made well, and allows me to get away from awful proprietary alternatives. If the output of an AI is distinguishable from a quality programmer, idk how that is also slop.

If you adopt a hardline anti ai stance you will just have to reverse it or become an unused backwater.

E: tagging is also stupid because a tag is self imposed and doesn’t discriminate between a person who is careful or otherwise.

Completely vibe coded - yes\
AI assisted - no

Vibe coded code whose "author" understands the language and actively challenges the agent for his decisions in the interest of long term window is OK. Vibe coded coded that are one-off without any future prospects should be 100% banned.

thats mine too

Yes please. It's dispiriting to regularly go through the cycle of seeing an interesting project pop up here, oh wait it's vibe-coded crap, nevermind...

Please