← Feed

Posts

• View post

  I’ve read today a contributor to an open source project says, in substance: "If generated code is accepted in the project, I will stop contributing because I don’t want to spend time trying to understand the code in order to contribute just to see it changed the next generated commit". Accepting generated code, even with mild rail guards, is a Pandora box way scarier than you think. It is not a question of purity, it is a question of keeping your project alive. https://ploum.net/2...

• View post

  RE: https://mastodon.social/@pojntfx/117425605598388134 Je n’ai jamais utilisé les passkeys car je n’ai jamais vraiment compris comment ça fonctionnait. Et si je ne le comprends pas intuitivement, cela signifie que je ne peux pas faire de backup, que je peux facilement perdre un accès à un compte ou dépendre d’un fournisseur. Comme je suis loin d’être une bille en informatique, je sais que si je ne comprends pas facilement un truc que l’industrie tente de pousser de toutes ses forces, c’est...

• View post

  If I follow correctly, "Artificial Intelligence" is so powerful that will soon destroy humanity but you can’t tax it or impose ecological constraints because that would destroy it. Also, you need to talk to it gently.

• View post

  A software always runs on a computer. A software always do what its creator wants it to do. If the software doesn’t do what its creator wanted, it is a mistake from the coder, a bug. If the software runs without the consent of the computer’s owner, it is a malware/virus. Stopping a software is as easy as shutting down the computer on which it is running. Terminator and Mission Impossible are hollywood fiction, there’s no such thing as "rogue software".

• View post

  AI people are starting to get tired and want AI to: 1. Check for clear mistakes 2. No First-Person Output 3. More Non-Natural-Language User Interfaces 4. Better Reproducibility https://blog.glyph.im/2026/09/serious-ai-product.html That’s a good start, dude. You are nearly ready to learn the UNIX command line.

• View post

  RE: https://neuromatch.social/@jonny/117325137427138728 Just take some time to understand that a lot of the "magic" and "creativity" of chatbots is basically hardcoded. Thousands of humans wrote it. The AI hype is basically ELIZA 2.0 powered by a nuclear plant.

• View post

  I make the hard prediction that every project/company/institution that keep a clear NO without exception to any use of AI will have a huge competitive advantage in the next 5 years. This also applies to intellectual workers/specialists. You just need to survive until that point. https://ploum.net/2026-09-09-ai_policy.html It looks like @kde@floss.social is on a good track: https://kdeforpeople.com/ I said 5 years but it may come very fast. Read books! Write! Learn! Resist!

• View post

  Faut-il fuir le numérique ? Ou bien tenter de le reconquérir ? https://ploum.net/2026-09-21-fuir_ou_reconquerir.html

• View post

  "Ploum, you should stop being an extremist! You see all in black and white but you have to grow up, accept compromises and see the other facet of things. If some people want to burn the planet, enslave people, commit genocides, promote dangerous ideologies, destroy education and eradicate mankind, you cannot just reject them. You have to listen and find a middle ground."

• View post

  To become independent from X, @EUCommission@ec.social-network.europa.eu is betting on … #WSocial, a sort of revamped BlueSky with exactly 3388 users with at least one post. @elena@aseachange.com is trying to alert people about this comical disaster: https://blog.elenarossini.com/tag/w-social/ Which is now even more nonsensical as they inadvertently invented BoratCasting ! https://aseachange.com/@elena/statuses/01M2MQXYZ177XDMX349PBFG7Z2 #WBoratCasting Is Very Nice European Technology for Ma...

• View post

  Just want to remind people that when Anthropic installed an "AI soda vending machine" in the Wall Street Journal office, it ended as stupidly as you could imagine. Anthropic said it was the future and when the journalist managed to have a moment of lucidity and genuinely asked if it wasn’t a stupid idea, Anthropic answer was: "Well, you should probably prepare for that world" https://ploum.net/2025-12-19-prepare-for-that-world.html The same stupid people are now shouting...

• View post

  Soyez humains, communiquez ! Pourquoi le moindre humain écrira toujours mieux que le chatbot le plus perfectionné. https://ploum.net/2026-09-14-ecrire_et_ia.html

• View post

  J’ai parfois l’impression sur le fait qu’il faut enfoncer le clou encore et encore : l’Europe est un laboratoire incroyable de développement des technologies numériques. On est juste mauvais à en faire des monopoles commerciaux qui exploitent les utilisateurs. Et c’est une excellente nouvelle ! On ne veut pas d’un Google européen mais de participation aux communs. https://ploum.net/2026-01-22-why-no-european-google.html

• View post

  I sometimes feel the need to repeat again and again that the EU is very good at developing technology but is bad at creating commercial monopolies out of it. And it is a f*** good thing! We don’t want an "European Google" https://ploum.net/2026-01-22-why-no-european-google.html

• View post

  Si vous avez en tête des personnes européennes actives dans le monde du logiciel libre que vous voulez voir mises à l’honneur, c’est le moment de les soumettre : https://awards.europeanopensource.academy/nomination-process

• View post

  My AI policy is "I don’t want any kind of interaction with that thing." And this should be all you need to know. https://ploum.net/2026-09-09-ai_policy.html

• View post

  J’ai finalement installé un human.json sur https://ploum.net https://codeberg.org/robida/human.json Je ne suis pas convaincu par le format mais tout outil pour établir un réseau de confiance entre "humains" me semble bon à prendre par les temps qui courrent.

• View post

  I eventually added a human.json to https://ploum.net https://codeberg.org/robida/human.json Thanks @contact for the initiative. Hoping to extend the human network of trust!

• View post

  Si on expliquait aux programmeurs que Claude est en fait un prénom féminin, je suis sûr qu’ils trouveraient soudainement tous que le code produit est de la merde et que les LLMs ne comprennent rien à la programmation, que programmer à la main est vachement mieux.

• View post

  2018 : here’s a very good, thin, light and modern smartphone with an audio jack. Here’s it for 300-500€. 2022: Well, small smartphone or audiojack, you have to choose, you can’t have both. And it will costs you 800€. 2026: Small smartphone ? Nope. Audiojack ? Nope. What are you smoking ? Anyway, if you want a "good" one, you are in for more than 1000€. #enshitification

• View post

  Unless proved otherwise, I now default to considering an unknown Github project younger thant 2024 as vibecoded shit. No, I don’t care about your number of stars or whatever. If you are still using Github and don’t have a clear community record, you will have to earn my trust.

• View post

  Un manifeste Offpunk https://ploum.net/2026-08-03-offpunk_manifesto.html Traduction du Offpunk Manifesto de @ideiasdechirico #offpunk #technopunk #smolnet #smalltech

• View post

  La ville de #biscarosse littéralement encerclée par les flammes ! Pensées pour celleux qui s’y trouvent et qui perdent tout aujourd’hui… Source : arcgis.com

• View post

  Automated combat drones are AI killer app.

• View post

  petit bravo au centre culturel d’Ottignies-Louvain-la-Neuve qui, propose, pour ses spectacles, 4 tarifs : - plume : pour les petits budget - juste : pour payer sa place normalement - soutien : pour soutenir les spectacles - solidaire : pour soutenir et compenser le tarif plume voire pour permettre des invitations aux personnes défavorisées. Alors c’est encore timide (quelques euros entre les tarifs) mais c’est vraiment un truc que je veux encourager. #olln #ottignies_louvain_la_neuve

• View post

• View post

  Just learned today about the Xteink X4 device, which seems really nice pocket e-reader (with physical buttons) gemini://axxuy.com/blog/2026/i-also-got-an-xteink-x4.gmi And there’s an Open Source firmware: https://github.com/crosspoint-reader/crosspoint-reader Has anybody tested it?

• View post

  For many years, I’ve been using Debian :debian: but installing Ubuntu :ubuntu: for others, fearing that Debian was too geek. I spent hours "desenshitificating" Ubuntu. After the Nth failed upgrade, I reconsidered. Oh boy! The Debian 13 GNOME live image is clean, polished, easy and perfect ! Install was a breeze, even with complex partitions. I only added a couple of gnome-shell extensions : dash2dock and desktop-icons. Debian just works like I dreamed it should 25 years ago! Than...

• View post

  Cette canicule vous est offerte par Total-Énergie. Par respect des sponsors, merci de ne plus dire "canicule de juin 2026" mais "Canicule Total-Énergie 2026". ( celle d’il y a un mois étant la "Canicule Bolloré 2026" ) https://ploum.net/2026-06-26-sponsor-canicule.html

• View post

  Vous avez des conseils pour un écran avec projecteur intégré compatible Linux :debian: ? Je parle de ces écrans qui se déplient pour occuper une grande surface et le projecteur est juste au pied de l’écran, projetant une image déformée pour qu’elle s’adapte à l’écran.