It’s everywhere. I was just trying to find some information on starting seeds for the garden this year and I was met with AI article after AI article just making removed up. One even had a “picture” of someone planting some seeds and their hand was merged into the ceramic flower pot.
The AI fire hose is destroying the internet.
I fear when they learn a different layout. Right now it seems they are usually obvious, but soon I wont be able to tell slop from intelligence.
One could argue that if the AI response is not distinguishable from a human one at all, then they are equivalent and it doesn’t matter.
That said, the current LLM designs have no ability to do that, and so far all efforts to improve them beyond where they are today has made them worse at it. So, I don’t think that any tweaking or fiddling with the model will ever be able to do anything toward what you’re describing, except possibly using a different, but equally cookie-cutter way of responding that may look different from the old output, but will be much like other new output. It will still be obvious and predictable in a short time after we learn its new obvious tells.
The reason they can’t make it better anymore is because they are trying to do so by giving it ever more information to consume in a misguided notion that once it has enough data, it will be overall smarter, but that is not true because it doesn’t have any way to distinguish good data from garbage, and they have read and consumed the whole Internet already.
Now, when they try to consume more new data, a ton of it was actually already generated by an LLM, maybe even the same one, so contains no new data, but still takes more CPU to read and process. That redundant data also reinforces what it thinks it knows, counting its own repetition of a piece of information as another corroboration that the data is accurate. It thinks conjecture might be a fact because it saw a lot of “people” say the same thing. It could have been one crackpot talking nonsense that was then repeated as gospel on Reddit by 400 LLM bots. 401 people said the same thing; it MUST be true!
This was honestly my biggest fear for a lot of FOSS applications.
Not necessarily in a malicious way (although there’s certainly that happening as well). I think there’s a lot of users who want to contribute, but don’t know how to code, and suddenly think…hey…this is great! I can help out now!
Well meaning slop is still slop.
Look. I have no problems if you want to use AI to make removed code for your own bullshit. Have at it.
Don’t submit that removed to open Source projects.
You want to use it? Use it for your own removed. The rest of us didn’t ask for this. I’m really hoping the AI bubble bursts in a big way very soon. Microsoft is going to need a bail out, openai is fucking doomed, and z/Twitter/grok could go either way honestly.
Who in their right fucking mind looks at the costs of running an AI datacenter, and the fact that it’s more economically feasible to buy a fucking nuclear power plant to run it all, and then say, yea, this is reasonable.
The C-whatever-O’s are all taking crazy pills.
AI crowd trying hard to find uses for AI
I think the open slop situation is also in part people who just want a feature and genuinely think they’re helping. People who can’t do the task themselves also can’t tell that the LLM also can’t do it.
But a lot of them are probably just padding their GitHub account too. Any given popular project has tons of forks by people who just want to have lots of repositories on their GitHub but don’t actually make changes because they can’t actually do it. I used to maintain my employer’s projects on GitHub and literally we’d have something like 3000 forks and 2990 of them would just be forks with no changes by people with lots of repositories but no actual work. Now these people are using LLMs to also make changes…
Why people try to contribute even if they don’t work on their codes? Ai slop not helping at all.
CV padding and main character syndrome.
just deny PRs that are obvious bots and ban them from ever contributing.
even better if you’re running your own git instance and can outright ban IP ranges of known AI shitlords.
The bots, they don’t like when you do that.
If my own mother can’t shame me, a glorified sex bot has a snowballs chance in hell of doing it.
Get that code off of slophub and move it to Codeberg.
Is codeberg magically immune to AI slop pull requests?
No but they are actively not promoting it or encouraging it. Github and MS are. If you’re going to keep staying on the pro-AI site, you’re going to eat the consequences of that. Github are actively encouraging these submissions with profile badges and other obnoxious crap. Its not an appropriate env for development anymore. Its gamified AI crap.
No (just like Lemmy isn’t immune against AI comments) but Github is actively working towards AI slop
If you want to get a programming job, you want a good looking CV. By contributing to prominent open source projects on github, github’s popularity and fancy profile system makes it look real good on a CV.
Github is a magnet for lazy vibe coders spamming their removed everywhere to farm their CVs. On other git hosts without such a fancy profile systems, there’s less on an incentive to do so. Slop to good code ratio should be lower and more managable.
This is big tech trying to kill FOSS.
Which is funny because most of them rely on it
Godot is also weighing the possibility of moving the project to another platform where there might be less incentive for users to “farm” legitimacy as a software developer with AI-generated code contributions.
Aahhh, I see the issue know.
That’s the incentive to just skirt the rules of whatever their submission policy is.
I don’t contribute to open source projects (not talented enough at the moment, I can do basic stuff for myself sometimes) but I wonder if you can implement some kind of requirement to prove that your code worked to avoid this issue.
Like, you’re submitting a request that fixes X thing or adds Y feature, show us it doing it before we review it in full.
The trouble is just volume and time, even just reading through the description and “proof it works” would take a few minutes, and if you’re getting 10s of these a day it can easily eat up time to find the ones worth reviewing. (and these volunteers are working in their free time after a normal work day, so wasting 15 or 30 minutes out of the volunteers one or two hours of work is throwing away a lot of time.
Plus, when volunteering is annoying the volunteers stop showing up which kills projects
Tests, what you are asking for are automated tests.
Can that be done on github?
Yep, take a look into GitHub actions. Basically you can make it so that a specific set of tests are run every time a PR is opened against your code repo. In the background it just spins up a container and runs any commands you define in a YAML config file.
A similar problem is happening in submissions to science journals.
A lot of programmers with thigh-high striped socks should take one for the team and take back Godot and such. Seriously!
I am a game developer and a web developer and I use AI sometimes just to make it write template code for me so that I can make the boilerplate faster. For the rest of the code, AI is soooo dumb it’s basically impossible to make something that works!
The context windows are only so large. Once you give it too much to juggle, it starts doing crazy shit.
Boilerplates are fine, they can even usually stub out endpoints.
Also the cheap model access is often a lot less useful than the enterprise stuff. I have access to three different services through work and even inside GPT land there are vast differences in capability.
Claude Code has this REALLY useful implementation of agents. You can create agents with their own system prompts. Then the main context window becomes an orchestrator; you tell it what you’re looking for and tell it to use the agents to do the work. The main window becomes a project manager with a mostly empty context window, it farms out the requests to the agents which each have their own context window. Each new task is individual, The orchestrator makes sure the agents get the job done, none of the workloads get so large that stuff goes insane.
It’s still not like you can say, go make me this game then argue with it for a couple of hours and end up with good things. But if you keep the windows small, it can crap-out a decent function/module if you clarify you want to focus on security, best practice, and code reusability. They’re also not bad at writing unit tests.
Something like speckit is necessary to make big, sweeping changes that continue past the context window
Interesting project, thanks for sharing
Yes I feel like many people misunderstand AI capabilities
They think it somehow comes up with the best solution, when really it’s more like lightning and takes the path of least resistance. It finds whatever works the fastest, if it even can without making it up and then lying that it works
It by no means creates elegant and efficient solutions to anything
AI is just a tool. You still need to know what you are doing to be able to tell if it’s solution is worth anything and then you still will need to be able to adjust and tweak it
It’s most useful for being able to maybe give you an idea on how to do something by coming up with a method/solution you may not have known about or wouldn’t have considered. Testing your own stuff as well is useful or having it make slight adjustments.
It finds whatever works the fastest
For a very lax definition of “works”…
Kind of agree with the rest of your points. Remember though, that the suggestions it gives you, for things you’re not familiar with may very well be terrible ones that are frowned upon. So it’s always best to triple check what it outputs, and only use it for broad suggestions.
Works in this case doesn’t mean the output works but that it passes the input parameter rules.
I think moving off of GitHub to their own forge would be a good first step to reduce this spam.
To Codeberg we go!
Codeberg is cool but I would prefer not having all FOSS project centralised on another platform. In my opinion projects of the size of Godot should consider using their own infrastructure.
Hosting a public code repo can be expensive, however they can run a private repo using Forgejo and mirror to Codeberg to create redundancy and have public code that doesn’t eat so much monthy revenue, if they even have revenue.
Let’s be realistic. Not everyone is going to move to Codeberg. Godot moving to Codeberg would be decentralizing.
Back to sourceforge it is then.
“Real men just upload their important stuff on ftp, and let the rest of the world mirror it.” - Linus Torvalds.
Don’t underestimate legitimate contributions from people who only do it because they already have an account.
It’s discussed in the Bluesky thread but the CI costs are too high on Gitlab and Codeberg for Godot‘s workflow.
That’s a shame. Did they take the wasted developer time dealing with slop into account in that discussion?
Before hitting submit I’d worry I’ve made a silly mistake which would make me look a fool and waste their time.
Do they think the AI written code Just Works ™? Do they feel so detached from that code that they don’t feel embarrassment when it’s shit? It’s like calling yourself a fictional story writer and writing “written by (your name)” on the cover when you didn’t write it, and it’s nonsense.
I’d worry I’ve made a silly mistake which would make me look a fool and waste their time.
AI bros have zero self awareness and shame, which is why I continue to encourage that the best tool for fighting against it is making it socially shameful.
Somebody comes along saying “Oh look at the image is just genera…” and you cut them with “looks like absolute garbage right? Yeah, I know, AI always sucks, imagine seriously enjoying that hahah, so anyway, what were you saying?”
Not good enough, you need to poison the data
I don’t want my data poisoned, I’d rather just poison the AI bros.
Yeah but then their Facebook accounts will keep producing slop even after they’re gone.
the data eventually poisons itself when it can do nothing but refer to its own output from however many generations of hallucinated data
Nowadays people use OpenClaw agents which don’t really involve human input beyond the initial “fix this bug” prompt. They independently write the code, submit the PR, argue in the comments, and might even write a hit piece on you for refusing to merge their code.
LLM code generation is the ultimate dunning Kruger enhancer. They think they’re 10x ninja wizards because they can generate unmaintainable demos.
They’re not going to maintain it - they’ll just throw it back to the LLM and say “enhance”.
Sigh, now in CSI when they enhance a grainy image they AI will make a fake face and send them searching for someone that doesn’t exist, or it’ll use a face of someone in the training set and they go after the wrong person.
Either way I have a feeling they’ll he some ENHANCE failure episode due to AI.
Do they think the AI written code Just Works
yes.
literally yes.
It’s insane
That’s how you know who never even tried to run the code.
that’s the annoying part.
LLM code can range to “doesn’t even compile” to “it actually works as requested”.
The problem is, depending on what exactly was done, the model will move mountains to actually get it running as requested. And will absolutely trash anything in its way, From “let’s abstract this with 5 new layers” to “I’m going to refactor that whole class of objects to get this simple method in there”.
The requested feature might actually work. 100%.
It’s just very possible that it either broke other stuff, or made the codebase less maintainable.
That’s why it’s important that people actually know the codebase and know what they/the model are doing. Just going “works for me, glhf” is not a good way to keep a maintainable codebase
LOL. So true.
On top of that, an LLM can also take you on a wild goose chase. When it gives you trash, you tell it to find a way to fix it. It introduces new layers of complication and installs new libraries without ever really approaching a solution. It’s up to the programmer to notice a wild goose chase like that and pull the plug early on.That’s a fun little mini-game that comes with vibe coding.
Reminds me of one job I had where my boss asked shortly after starting there if their entry test was too hard. They had gotten several submissions from candidates that wouldn’t even run.
I envision these types of people are now vibe coding.
Super lazy job applications… can’t even bother to put two minutes into vibing.
From what I have seen Anthropic, OpenAI, etc. seem to be running bots that are going around and submitting updates to open source repos with little to no human input.
You guys, it’s almost as if AI companies try to kill FOSS projects intentionally by burying them in garbage code. Sounds like they took something from Steve Bannon’s playbook by flooding the zone with slop.
at least with foss the horseshit is being done in public.
Doesn’t someone have to review those submissions before they’re published?
Can Cloudflare help prevent this?
