Forbidding generative AI and LLMs on Pepper&Carrot
Published on
Following many Free/Libre and Open Source projects over the past week, last night I added a "Use of Generative AI" chapter to our Code of Conduct, clearly stating that we do not allow the use of Generative AI or LLMs. This also affects all our repositories on Framagit.
→ Link to the new no-LLMs statement chapter in our COC.
I copied the concise and efficient text word for word after discovering it thanks to Jeff Fortin T. (風の庭園のNekohayo) who added it to GNOME Calendar, and posted about it on Mastodon. Many thanks to Sophie Herold for writing it, and also for posting it under a CC0 licence.
Illustration sources: Rage against the Slop Machines
206 - 
375
28 comments
DrJLecter@beige.party
absolutely love that drawing. Fuck those AI slop machines.
★Poslovitch@wikis.world
After the Firefox-fighting-the-huge-Chrome-flying-ball, this is going to be my new wallpaper!
★tragivictoria@catgirl.cloud
@Poslovitch I need to see that wallpaper!!
★vitordematos@floss.social
Caroooooooot! Vai-te a eles!! Nem são inteligentes nem são artificiais! São parte de um plano concertado e programado de "merdificação" para nos retirar poder e nos controlar!
★cloudskater@bark.lgbt
I really appreciate you making your stances on "generative" AI clear and being so loud and proud about it. You really inspire me, David, in that regard especially. With the news about Lutris (a game launcher I used up until this week) beginning to use and intentionally obfuscate "generated" code, as well as becoming aware of some Linux kernel devs rubbing shoulders with AI slop themselves, this puts a smile on my face.
@mrmasterkeyboard I think you might like this :3
★BigShellEvent@toot.cat
carrot go get 'em
★Khrys@mamot.fr
Thanks 🙏
3 ★kimcrawley@zeroes.ca
David, I thought you were really cool. But it turns out you're not really against Gen AI after all.
🖼️ ebe088c7b19ea88b.jpg ★davidrevoy
@kimcrawley It starts by a clear "no" , right? It also says i disapprove it, but it has a nuance for the computer science and FLOSS projects who use the technics (not with any proprietary AI of course, crawled on data without consent), with just their in-house models trained on small in house ethical datasets, for statistic, scientific research or image algorythm (eg. Filters, Sharpenîng, img search local). Sorry for not being in absolute purity and have a nuance. I' m like that on many topics.
12 ★eltheanine@the.teabag.ninja
Having an actual nuanced thought on a topic? How dare you.
People are the problem, as with any tool. I can't believe we live in a world where obviously bad people get to do obviously bad things with no repercussions, but here we are.
★
kimcrawley@zeroes.ca
There's absolutely no such thing as ethical Gen AI. I'm not talking about types of machine learning that existed before 2022. That's not Gen AI. I'm talking about *generative* "AI."
And there's no such thing as "ethical" Gen AI, nor "it'll get better in the future!" Even your "what if it's just local, on device!" is a bullshit excuse.
Gen AI is inherently fascist.
Gen AI is inherently brainwashing and mindwarping.
The massive environmental destruction and training off off massive amounts of people's creations without permission isn't the only problem with Gen AI.
And I'm sick and fucking tired of people's obsession with finding "nuance" in inherently evil shit. Take a strong stance against evil, or be a coward.
If only you had the moral courage of Anthony Moser, rather than the cowardice of a tech bro.
https://mathstodon.xyz/@gregeganSF/116219772468880168
🖼️ ffa72dd2e8dad4f4.png 🖼️ 85d70e9fd35340ad.png 🖼️ 7dc176160f5c8989.png 3 ★davidrevoy
@kimcrawley Ok, I think we just have a problem of definition here. I'm just talking about the coding method, the small neural networks home made for Character Recognition for scanner, built on a limited dataset of scans (with autorisation); or a filter that auto clean line-art, like on this paper https://www.davidrevoy.com/article642/scientific-paper-sketch-to-line-art-by-waseda-university I participated in 2018. Yes, things that existed before 2022.
I think we agree on all gen AI from surveillance capitalism by corpo. No need to jump on the big words or insults.
3 ★davidrevoy
@kimcrawley Also, I concede my wording on my FAQ (that's probably was written around 2022/2023) could be better and more precise. I'll note on my TODO to try to disambiguate this as much as possible.
Edit: updated with an 'addendum': https://www.davidrevoy.com/static2/about-me
2 ★aloxe@mast.eu.org
ça me fait trop penser à Kaonashi dans Spirited away. Ce personnage est d'ailleurs une belle image des LLM actuels, qui bouffent tout et tout le monde mais fascinent tant.
davidrevoy
@aloxe Oui, une grosse grosse inspiration, un mix entre ça et le sans visage. 💜
dalias@hachyderm.io
I like the approach of addressing it in the CoC, since the problem is fundamentally a human behavioral problem of misrepresentation, deception, time-wasting, etc.
6 ★lumi@snug.moe
@dalias it is also mostly ethical and social, rather than technical
★Whimsy@beige.party
The cat’s face is purrfect :blobcatlove:
★blind_mapmaker@eldritch.cafe
Good text, good illustration. Keep up the good fight!
gmarcosanti@mastodon.uno
that's not what an angry ready-to-attack cat looks like at all, but I like the concept.
regendans@todon.eu
"Rage Against the Slop Machines"
Nice! 💫
★FediThing@chinwag.org
👏
★voxel@infosec.space
I'm very frustrated and concerned about AI – not because it will "take over the world" or some conspiracy like that, but because it's going down the same path as blockchain and other technologies regarding cryptocurrency, tokens, etc.
It's a good and interesting technology, but the problem is that it's used as a marketing term and for all kinds of "bullsh*" which ruins the reputation of the technology as a whole and burns everyone in the process.
(The Em dash is not there because I used AI, but because I'm trying uncommon and almost "dead" but useful text elements like the semi-cologne ; for example, while keeping things language accurate)
★mangeurdenuage@shitposter.world
Pourquoi bannir tout LLM lorsqu'il est évident que les violation morale/éthique en la matière sont fais par des LLM privateur de liberté ?
La FSF fut claire sur le sujet que la raison pour que ces outils soit non éthique et utiliser a des fin nocifs est du a leur nature privateur.
Si vous n'avez pas vue la video de la FSF au FOSDEM 2025 sur le sujet je vous encourage a prendre le temps (si possible) de la regarder.
https://archive.fosdem.org/2025/schedule/event/fosdem-2025-4818-fsf-s-criteria-for-free-machine-learning-applications/
davidrevoy
@mangeurdenuage Cette présentation n'apporte rien de concret dans les usages pratique: juste un objet théorique avec des critères pas mis en application... la FSF quoi.
J'invoque avec ce genre de règle catégorique et simple une protection contre des usages déjà là. Ca donne au projet un motif de ban ou d'avertissement pour bannir les MR d'un agent Claude/Copilot/Mistral/Llama à vue, ou gronder un bleu qui penserait qu'on pourrait vibe-coder un truc pour le site et le soumettre.
(1/2)
davidrevoy
@mangeurdenuage Ensuite, si vient un jour l'objet théorique de la FSF, genre le truc génial, fiable, pas privateur, etc... et que c'est utilisé dans tout les FLOSS, hè bien on verra peut-être à bâtir des exceptions... Mais on en reparle dans 10 ans à mon avis 🤣 . (2/2)
mangeurdenuage@shitposter.world
Les gens/corporation disais ça du projet GNU en 83. Rome ne ses pas bâtit pas en 1 jour.
raymaccarthy@mastodon.ie
I love Pepper & Carrot* and also your Avian Intellegence series.
[* Though I've not read much of it]
★Post a reply
The comments are synchronised every 1h with the replies to this post on Mastodon:How to use this? (click here to unfold)
Open a new Mastodon account on the server of your choice. Then, Copy/Paste the adress above in your Mastodon 'Search' field. The post will appear and you'll be able to fully interact with it. You'll have full control of your posts: edit, remove, etc. After that, your message will appear here.
Just please note that it may take up to 1 hours for your changes to be reflected here.