soc.octade.net is a Fediverse instance that uses the ActivityPub protocol. In other words, users at this host can communicate with people that use software like Mastodon, Pleroma, Friendica, etc. all around the world.
This server runs the snac software and there is no automatic sign-up process.
Why do LLMs freak out over the seahorse emoji? https://lobste.rs/s/2ykcqe #ai
https://vgel.me/posts/seahorse
A #Colorful #Clown at a #Festive #Party by Kaye Menner Photography Wide variety #Prints & lovely #Products at:
#whimsical #balloons #digitalart #ai #curlyhair #facepaint #rednose #tophat #wrinkles #homedecor #mastoart #fediverse #fediart #fedigiftshop #giftideas #wallartforsale #Art #artforsale #BuyIntoArt #AYearForArt #Artist #FineArtAmerica #PhotographyFeed #VisualArts #CreativeArts
The Case for Learned Index Structures https://lobste.rs/s/nsl5gr #ai #compsci
https://dl.acm.org/doi/10.1145/3183713.3196909
Tree of Butterflies by Kaye Menner #Photography Wide variety #Prints & lovely #Products at:
https://bit.ly/4gXvnWV
#mystical #magical #fantasy #whimsical #digital art #ai #homedecor #mastoart #fediverse #fediart #giftideas #wallartforsale #Art #artforsale #BuyIntoArt #AYearForArt #Artist #FineArtAmerica #PhotographyFeed #VisualArts #CreativeArts
Apertus: a fully open, transparent, multilingual language model https://lobste.rs/s/vpjhmi #ai
https://ethz.ch/en/news-and-events/eth-news/news/2025/09/press-release-apertus-a-fully-open-transparent-multilingual-language-model.html
The AI bubble is 17 times the size of the dot-com frenzy, and four times the subprime bubble, analyst says. It's not just a bubble but an epically sized one, an analyst argues. When it will burst it is going to hit harder for tech sectors especially AI and related compaines and stocks. https://www.marketwatch.com/story/the-ai-bubble-is-17-times-the-size-of-the-dot-com-frenzy-this-analyst-argues-046e7c5c
We Bought the Whole GPU, So We're Damn Well Going to Use the Whole GPU https://lobste.rs/s/cqdtto #ai #performance
https://hazyresearch.stanford.edu/blog/2025-09-28-tp-llama-main
Double dose de drogue : la religion, "opium du peuple" selon Marx, plus un booster d'AI... Le monde dystopique de 2025 ressemble de plus en plus à celui des "Eaux de Mortelune", non ? #Religion #AI #IA
Virtual Jesus? Virtual Buddha? People of faith divided as AI enters religion
Bangkok Post - Accès libre.
https://www.bangkokpost.com/life/tech/3115114/virtual-jesus-virtual-buddha-people-of-faith-divided-as-ai-enters-religion
Sur "Les Eaux de Mortelune"
https://fr.wikipedia.org/wiki/Les_Eaux_de_Mortelune
ZLUDA update Q3 2025 – ZLUDA 5 is here https://lobste.rs/s/letrig #ai #release
https://vosen.github.io/ZLUDA/blog/zluda-update-q3-2025/
#Field of #Red #Poppies at #Sunset by Kaye Menner #Photography Wide variety #Prints & lovely #Products at:
#floral #redfloral #flowers #redflowers #redpoppies #digitalart #ai #colorful #landscape #mountains #redyellow #sky #homedecor #mastoart #fediverse #fediart #fedigiftshop #giftideas #wallartforsale #Art #artforsale #BuyIntoArt #AYearForArt #Artist #FineArtAmerica #PhotographyFeed #VisualArts #CreativeArts
#Bicycle in #Spring #Garden by Kaye Menner Wide variety #Prints & lovely #Products at:
https://kaye-menner.pixels.com/featured/bicycle-in-spring-garden-by-kaye-menner-kaye-menner.html
#vintage #flowers #floral #colorfulflowers #spring #springflowers #sunrise #blossoms #pastel #garden #nature #ai #digitalart #homedecor #mastoart #fediverse #fediart #fedigiftshop #giftideas #wallartforsale #Art #artforsale #BuyIntoArt #AYearForArt #Artist #FineArtAmerica #PhotographyFeed #VisualArts #CreativeArts
“We Have No Idea How Models will Behave in Production until Production”: ML Ops Study [2024] https://lobste.rs/s/hhohj2 #ai #practices
https://arxiv.org/html/2403.16795v1
@loganer it doesn't need infinite options hard-coded, just a list of options it isn't allowed.
Pretty shure they all boil down to a few rules written in #Lua which could be summarized as if/else staircases re: said behaviour...
@loganer exactly that is why "#AI" is useless in so many ways and why I'd say that you ain't looking for actual "AI" but just #NPCs or #bots that do more than just give you quests, trades and a limited set op options that don't change much...
Again: "fancy dice with a tree of options", not much else.
@loganer Except "#AI" solely is being pushed as a #Bubble to #precariate #WageWork, increase #Theft from #Society to #Billionares and espechally #defund and #divest #PublicInfrastructure and #Education.
Calling an NPC an "AI" is far fetched given it's merely a #bot with a set list of rules to follow.
#TLDW: "#AI" is #bad - period!
Seriously, I think "AI" needs to be outlawed like #CFC11, #Tetraetyllead, #Shitcoins (like #Bitcoin, #Ethereum, and their clones) as well as "pointlessly driving within city limits" because it's an
"The most important thing is the investor..." #AI #Quote
- Cory Doctorow ➡️ @pluralistic / #CoryDoctorow
https://mamot.fr/@pluralistic
(P.S I really like Cory's profile banner and think it could be part of a show - similar to teachings of Defective by Design (@endDRM) aka endDRM #DRM #DefectiveByDesign
Making sure AI serves people and knowledge stays human https://lobste.rs/s/bdv0hj #ai
https://diff.wikimedia.org/2025/09/30/making-sure-ai-serves-people-and-knowledge-stays-human-wikimedia-foundation-publishes-a-human-rights-impact-assessment-on-the-interaction-of-ai-and-machine-learning-with-wikimedia-projects/
We scold AI for “hallucinations” – invented books, fake citations, Socrates debating Quidditch. But humans are the real hallucination machines. We carved race and selfhood out of fog, baptised them in violence, and called them truth. Censuses, laws, diaries, courts – these stabilised fictions until they felt natural. They don’t need metaphysics; they’re real because people kill for them
#AI #Hallucinations #Philosophy #Race #Identity #Self #Enlightenment #CriticalTheory
So today I needed an interface with a few intentional accessibility errors that I wanted to show for a demo. It didn't need to actually work, I just needed to briefly record the effect of a particular issue. So I used #AI to vibe code it for me. It was perfect! Within fifteen minutes I had an inaccessible interface that didn't work right! Exactly what I wanted! No, but on a more serious note, it was impressive that it did manage to build in exactly the flaws I asked it for (and some bonus other flaws), and it was full fledged enough that I could show the issue I wanted to demonstrate without writing any code myself. I do find it somewhat ironic, though, that this is easily going to become the way I most frequently use AI at work, and it's going to save me a ton of time and effort.
Functional genetic programming and exhaustive program search with combinator expressions (2007) https://lobste.rs/s/oysxby #pdf #ai #concatenative #math #ml #plt
https://www.cs.hmc.edu/~oneill/papers/Combinators-KES.pdf
I have no idea how widespread it is, but it seems there's an obsession among many Christians with "AI" generated content.
What the Bible tells about God is that He really really doesn't like being offered "Whatever"; From the very beginning with the story of Cain and Abel, to the end saying He prefers people who are cold to Him than lukewarm.
1/2
I just conducted a training for teachers of a non-profit on AI and told them that AI isn't really thinking. They were so shocked.
We have to realize that non-technical folks will accept the marketing talk by the AI propagandists without question since they have no way to verify what they're saying.
Someone should probably inform the White House's "AI & Crypto Czar" that no one is forcing AI companies to train their models on Wikipedia
You would think the obvious solution to "the volunteer-powered project we all train our AI models on for free isn't adequately twisting reality to our political views" would be "... and so we stopped training on it" and not "... and so we will force the volunteers to bend to our will"
Acuitas' Text Parser can finally (in theory) handle all the grammatical structures in the test sets! https://writerofminds.blogspot.com/2025/09/acuitas-diary-88-september-2025.html
#ArtificialIntelligence #AI #chatbots #NLP #NLU #symbolicAI
Introducing Claude Sonnet 4.5 https://lobste.rs/s/gtnuao #ai
https://www.anthropic.com/news/claude-sonnet-4-5
Fast Company:A I videos are tricking tourists into visiting places that don’t exist. That’s just the beginning
You thought that travel influencers were bad? It’s about to get worse.
(7/17/2025)
https://www.fastcompany.com/91368492/ai-video-tricking-tourists-places-that-dont-exist
Learning a new AI-related term: workslop.
"The insidious effect of workslop is that it shifts the burden of the work downstream, requiring the receiver to interpret, correct, or redo the work. In other words, it transfers the effort from creator to receiver."
https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity
#workslop #AI #slop #GenAI #technology #news #TechNews #study
My apologies, I believe the preferred term now is "vibe working".
https://www.theverge.com/news/787076/microsoft-office-agent-mode-office-agent-anthropic-models
LLM Inference Economics from First Principles https://lobste.rs/s/cay1iq #ai #vibecoding
https://www.tensoreconomics.com/p/llm-inference-economics-from-first
Slop Machines: on the interaction between feed recommender systems and genai https://lobste.rs/s/ogs7d1 #ai
https://notes.hella.cheap/slop-machines.html