Who cleans up after the vibe-coding party?
This is an excellent and approachable article on the destructive impacts of AI coding.
Who cleans up after the vibe-coding party?
This is an excellent and approachable article on the destructive impacts of AI coding.
Können wir den Ausbau von Rechenzentren eindämmen, bevor es zu spät ist?
In unserem Artikel erfährst du, wie Rechenzentren finanziert werden und welche Auswirkungen ihr Ausbau auf uns hat.
Dazu stellen wir Rahmenbedingungen vor, die wir einführen sollten, um Rechenzentren zu bauen, die Mensch und Umwelt weniger schaden.
Spoiler: Wir glauben daran, dass eine andere Zukunft für Rechenzentren möglich ist! 🫰
#innovation #llm #datenzentren #nachhaltigedigitalisierung
Rechenzentren gefährden die Energiewende, verbrauchen kostbare Wasserressourcen und schaden der menschlichen Gesundheit. Können wir ihren Boom verlangsamen, bevor es zu spät ist?sarah-indra (Digital for Good | RESET.ORG)
Looking for the best Linux distribution to run local LLMs?
I compared Ubuntu, Fedora, Pop!_OS, Linux Mint, and Arch Linux, highlighting who each distro is best suited for.
The article also includes a visual decision tree to help you choose the right one quickly.
opensourcefeed.org/best-linux-…
#Linux #OpenSource #LocalAI #LLM #Ubuntu #Fedora #ArchLinux
Looking for the best Linux distro for Ollama or other local LLM tools? Compare Ubuntu, Fedora, Pop!_OS, Linux Mint, and Arch Linux to find the right fit.OpenSourceFeed
Emily Bender ( @emilymbender ) hat wieder einmal einen sehr hilfreichen Text zur Frage "was ist #KI/ #AI" veröffentlicht.
Dieser könnte auch für die #Lehre / #Schule ein sehr hilfreicher Standardtext sein.
Hier geht es zum kostenlosen preprint: dair-community.social/@emilymb…
faculty.washington.edu/ebender…
Direkt:
faculty.washington.edu/ebender…
#Medienkompetenz #Medienbildung #KünstlicheIntelligenz #LLM #Digitalkompetenz #Informatik #Unterricht
Preprint available from my publications page: https://faculty.washington.edu/ebender/publications/Prof. Emily M. Bender(she/her) (Distributed AI Research Community)
«Jadepuffer — Die erste KI-Ransomware-Attacke verbaselt nicht nur den Schlüssel:
Forscher dokumentieren den ersten LLM-gesteuerten Ransomware-Angriff. Doch die KI scheiterte an grundlegenden Fehlern. Eine Mahnung für Admins.»
Das die KI mittlerweile scheinbar "alles" hackt ist auch ein Teil ihres Marketings. Denn durch Unsicherheit lässt sich einiges besser als "Die Lösung" verkaufen.
🙄 heise.de/news/Jadepuffer-Die-e…
#ki #unsicherheit #jadepuffer #llm #itsec #itsicherheit #ransomeware #kifehlerquote
Does anybody else feel like LLMs completely destroyed not only the job market, but the job hunting process itself? I apply to dozens of jobs and it's like I just don't exist. Nobody even notices me jumping in the back yelling, "Ooh, me! Pick me!" because of the throng of LLMs auto-applying and the other throng of LLMs doing the actual candidate selection. The game has changed. IDK how to play it anymore, and I'm literally going hungry as a result.
let's not see this as unrealistic. The oligarch class are behind the "enhanced games" where athletes are allowed to use drugs to enhance performance.
Why would they do this?
It's not hard to understand. They've managed to foist #LLM on people who are willingly rotting their brains in the service of dubious benefits even to their employers. But dumbing people down makes them easier too control and enslave.
Capitalism is at heart a way to enslave people.
Oh hello... Anthropic just extended the Fable 5 trial for Pro users until July 12th.
Translation? They’re sweating that 90 EUR/month Max tier and desperately trying to convert people. Nice try, Anthropic, but my wallet is staying closed.
Tatsächlich fühlte ich mit beim Mitlesen der Texte an altertümlich formulierten #Aislop erinnert: Besonders angetan hat es mir der „Flurbereich“, den man ja eher in einem Ikea-Katalog erwarten würde.
Abgesehen davon, war es schön, wieder einmal einem Chor mit Klavierbegleitung zuzuhören.
Aber immerhin, es bebet das Gesträuche.
#Liebeslied #klassischemusik #LLM #AiGenerated
Aquí os dejo una explicación sobre cómo empujar a un LLM con capacidad de "razonamiento" a que haga cosas indebidas porque "las voces en su cabeza le dicen que queme cosas".
hackaday.com/2026/07/02/chain-…
#llm #ia #inteligenciaartificial #cotforgery
Researchers [Charles Ye], [Jasmine Cui], and [Dylan Hadfield-Menell] have shown that AI Large Language Models (LLMs) can fail to correctly distinguish between different instruction sources because …Donald Papp (Hackaday)
Who owns the code that Claude wrote?
oreilly.com/radar/who-owns-the…
AI-generated code copyright explained for builders.Sena Evren (O’Reilly Media)
The stories of large sites selling their data to #OpenAI then seeing a complete crash of their traffic and userbase is both sad and a blatant tale of #FAFO that I struggle to feel empathy about.
StackOverflow went from 'partnering' with OpenAI to an utter wasteland in ONE YEAR.
What did they think would happen??
#AI #LLM #programming
Oh here's someone saying that LLMs/AIs should be included in a CoC & Diversity goals. 🔥🔥🔥
There. It crafted a proper welcome message. No manipulative animal theming. Even comes with a suggestion to close the browser tab.
It seems the people running these shitty things never think of this, nor accept the need for them.
"Recommendation: Preferably, do not use this website and instead use proper websites like Wikipedia and school books for accurate information. Consider closing this browser tab."
Maybe it's the hashtags I follow. The past months I keep on seeing fedi bots promoting terrible AI generated "learning experience" websites aimed at children.
For some reason most of these sites include dodgy #LLM chat options. Often cutesy themed after animals or historical figures.*
If you're a parent: are you actually ok with your still developing child being manipulated by these things?
*Here's a Dutch one. The chat is unprotected leeralles.nl/waffel/
Kennt ihr #SLM für Alt-Textgeneratoren?🔎
Oder #LLM die nicht #Azure von #MS nutzen (wie zB der von barrierefreies.design) sondern wirklich #datenschutzkonform sind? 🤖
Und auf einem nachhaltigen Rechenzentrum betrieben?💚
Freu mich über Empfehlungen!
#AI #KI #Alttext #Barrierefreiheit
You are totally correct.
Many people confuse the benefits of automation with a perceived need for AI or LLMs.
They don't want 'intelligence' to do things for them. They want to implement a set of rules that they themselves define, to be done automatically instead of having to do it themselves. That is automation.
#AI #LLM #Automation
Auch wenn es keine konkreten Zahlen zum Energie- und Wasserverbrauch von Rechenzentren gibt: Das Problem, dh KI und LLM, für die Umwelt ist schon längst bekannt. Da könnten viele ja auch von selbst mal versuchen, kürzer zu treten und KI und va LLM nur zu benutzen, wenn es wirklich notwendig ist, und nicht, weil es lustig und bequem ist.
AI slop hurts your soul.
It makes you: distrustful, dismissive of creative works, uncreative, derivative, imaginatively stunted, unreflective, dependant, it robs you of the satisfaction and growth granted by the creative process and it poisons you with all the same cultural biases taken from the training data.
AI robs people of imagination and the drive to create and instead produces output that is pastiche on steroid, uninspired unchallenging slop, every little detail crafted without personality or opinion with nothing to say except "this might vaguely remind you of something good that already existed"
I am also particularly concerned about the rise of chatbots pretending to be humans for political purposes.
alextheward.com/blog/llms_are_…
I've seen a bunch of generally reasonable people say some pretty weird things about LLMs and I want to talk about them.alextheward.com
@frutiger @f0bb1d04 @cairobraga
github.com/tutao/tutanota/disc…
A high traction repo post requesting an explicit ban on LLM coding has had no official response in over a month
#LLM
Tuta has been a staunch fighter in resisting AI features in their application, as they have posted on their blog in several instances or also on their Mastodon. However, upon scrolling the repo, I ...GitHub
Good, LLMs are a net negative to society, the erosion of truth, impersonation, scams, misinformation, propoganda, denigration of the arts and human expression, deskilling, ai psychosis, intellectual dependance on technology leading to stunted personal development, decrease in literacy rates
When seeing your local GP ...
How sensitive information can be gleaned from medical AIs
Medical records could be revealed by AI training-data vulnerability (podcast)
nature.com/articles/d41586-026…
Knolle, M.A., Menten, M.J., Jungmann, F. et al. Disparate privacy risks from medical AI. Nature (2026). doi.org/10.1038/s41586-026-106…
nature.com/articles/s41586-026…
#privacy #PrivacyRisk #MedicalAI #DataSparcity #AI #LLM #MembershipInferenceAttack #health #MedicalRecords #consent #prosumer #PatientPrivacy #NSW
Hear the biggest stories from the world of science | 24 June 2026Petrić Howe, Nick
firststreet.org/research-libra…
"Approximately 54% of #global #datacenter capacity operates in markets facing elevated chronic #heat or #drought stress, while 79% is exposed to significant acute hazards such as #flood, #wind, or #wildfire. For many markets, #climate #risk should now be considered part of the base case rather than a tail-risk scenario."
#AI #GenAI #LLM #climatechange #water
First Street™ research is transparent, peer reviewed, and proven against real environmental events. Use the Library to access First Street Research.firststreet.org
The #EuropeanCommission advertises as "open-source" an #LLM which -by their own call- need only provide, "where applicable, the release of open weights, .. OR .. technical documentation" [1].
So the result is open-weights at most, and would not even satisfy the OSI Open Source AI Definition (#OSAID 1.0).
This is not the first time the EC's claims don't survive scrutiny.
An institution funding "sovereign AI" at scale should know the difference.
Still more reasons the AI boosters are full of as much shit as the AI itself.
threads.com/@jenatlionra/post/…
I just wanted to update my resume. Instead, I accidentally proved how a multi-billion-dollar AI tool hallucinates a glass ceiling for women. I changed a single variable: My name. Here is what happened when "Jennifer" became "Jeff.Threads
In a policy statement¹, @conservancy said:
> “#FOSS projects should not shun contributors who choose to use LLM-gen-AI systems.”
@dalias' reply:
>> “LOL WTF NOPE. 🤡”
While it was surely unintentional, your reply is quite similar to the cruelty of traditional #FreeSoftware rhetoric — wherein we shunned people for *using* #Apple & #Microsoft. Such users deserve sympathy and help toward more software freedom.
Same goes for #LLM-backed generative #AI users.
¹ sfconservancy.org/llm-gen-ai/l…
The Software Freedom Conservancy provides a non-profit home and services to Free, Libre and Open Source Software (FLOSS) projects.sfconservancy.org
Kriegsziele per LLM sloppen - was könnte schief gehen?
Anthropic CEO Dario Amodei findet es patriotisch, dass Claude im Krieg eingesetzt wird. Am Ende entscheide immer ein Mensch. Die Erfahrung zeigt: der Mensch tut einen Scheiß.
Das ist jetzt weniger ein Problem der Technik und eher ein Problem von Regimen, denen Menschen egal sind. Da passt dann einfach nur eine Technik wie Arsch auf Eimer, der die eigenen Ergebnisse auch egal sind. Und dem CEO ist es auch egal. Hauptsache das Geld kommt rein.
youtube.com/shorts/BKqbCfK0ViA
#llm #ki #ai #anthropic #aislop
Posting something insane an AI executive said every day in 2026…until I run out of crazy stuff. Day 168: Dario Amodei, CEO of Anthropic (Claude), claims Clau...Panodime (YouTube)
(URL replace addon enabled for X, YouTube, Instagram and some news sites.)
Analysis shows that indiscriminately training generative artificial intelligence on real and generated content, usually done by scraping data from the Internet, can lead to a collapse in the ability of the models to generate diverse high-quality out…Nature
Hab duck.ai gefragt, was die Politik der US Regierung auszeichnet 🤔
Fokus auf Ausbau sozialer Programme und Klimaschutzmaßnahmen
stärkere Regulierung großer Tech‑Firmen, betonte Maßnahmen für Datenschutz und algorithmische Transparenz
Einwanderung & Grenzschutz versucht, Balance zu finden — humanitäre Reformansätze kombiniert mit Durchsetzungspraxen
Ausrichtung auf Wiederherstellung internationaler Bündnisse
Betonung auf Gleichstellung, Bürgerrechten, Diversity‑Förderung
That's why you don't read anymore.
You feed the legacy text through #LLM and read the bullet point summaries 💀
Joking not joking
Die Kinder wissen, dass #Siri dumm ist und beömmeln sich darüber, was für witziger Quatsch passiert, wenn sie Siri was fragen.
Also macht #Apple ja jetzt Gemini überall rein und dann wird das alles in Ordnung kommen.
ODER?
Heute morgen habe ich die Kids beobachtet, wie sie (den Sprachmodus von) chatGPT 5.5 verarscht haben.
Sie haben erst nach dem stärksten Tier gefragt und da kam dann sowas wie Löwe oder Blauwal.
Kinder: „Aber eine Ameise sei doch auch stark…“
... 🧵 1/3
Haha shameless #llm use. I could have made these in Ai myself but now I can go outside and enjoy #nature
Just how I’ve been looking at the old platforms since finding the #fediverse
Aus Sicherheitsgründen wurde dein Tröt zurückgestuft. Angezeigt wurde kein Hinweistext. Mastodon verfügt über »Sicherheitsmaßnahmen«, »die Nachrichten zu den meisten Themen rund um Cybersicherheit oder Biologie markieren«. Dabei können auch sichere, normale Inhalte markiert werden, was offenbar der Fall war.
Das KI-Modell Mythos sei zu mächtig für die Öffentlichkeit, hieß es. Jetzt gibt es eine angeblich sichere Version. Wer sie testet, stößt schnell an Grenzen.Eva Wolfangel (DIE ZEIT)
claude.ai/share/e3ce7d3e-3992-…
Die Freude war leider kurz. Die Fascho-Regierung hat das neue LLM-Modell Fable 5 aka Mythos kurzerhand für Nicht-US-Bürger abschalten lassen.
Aber hey, ist ja gerade Fussball und wird wohl niemand so richtig stören…
Hello Dave...
HAL 9000 is the real perpetrator!
cariagiovannib.wordpress.com/2…
#InfoSec #AISecurity #LLM #Fable5 #ModelFingerprinting #HAL9000 #AI #ironic #satire
The Fable incident, the US government shutdown, invisible AI model fingerprinting, and the real architect behind all of it — whom nobody has identified yet. AI Security Threat Intel Satir…cariagiovannib (Caria Giovanni - Security Blog)
Introducing ChatSLM. A small language model with a big heart.
▪ 0 tokens per query
▪ 0 L cooling water
▪ runs on a single ARM64 core in Nuremberg, hosted on FreeBSD.
▪ no telemetry, nothing sent anywhere
▪ consistent, reproducible answers every single time
Give it a try. It's often quite helpful.
chat.rootshell.cx
#AI #SelfHosting #Sustainability #LLM #satire