home.social

#genai โ€” Public Fediverse posts

Live and recent posts from across the Fediverse tagged #genai, aggregated by home.social.

  1. So many websites and apps now have em dashes everywhere.

    It feels so soulless knowing that a lot of them are now slopped with LLMs.

  2. @oatmeal

    Those documents are literally made to support their gut feeling, and hence, their preferred narratives.

    As one of the fact checkers emphasized: the documents are regurgitating the phrase "evidence based", ad nauseam. For a purpose.

    It wasn't just a few factual errors that allegedly slipped the attention of "professionals". It's not even an "unfortunate" deskilling that goes along with the persistent use of " #AI ".

    Those people are already at the stage of presenting their agenda as seemingly backed by "science", and "evidence", by fabricating alternative facts.

    This is is a fundamental attack on how science works, and on how knowledge is acquired, verified, and disseminated.

    We're even lucky that it's all about documents, here. Texts can't overwhelm suspicion and critical inspection as fast and effortlessly as #GenAI "photos" and "diagrams" can.

    #science #EvidenceBased #education

  3. ใ€ŒChatGPT Walletใ€ใŒ็™บ่กจ ๏ฝžๆ”ฏๆ‰•ใ„ใƒˆใƒผใ‚ฏใƒณใƒป็”Ÿไฝ“่ช่จผใชใฉใงAIใฎใŠไฝฟใ„ใ‚’ๅฎ‰ๅ…จใซ๏ผใพใšใฏ็ฑณๅ›ฝใฎใ€ŒProใ€ใƒฆใƒผใ‚ถใƒผใซใ€ใ€Œdotsใ€ใงๅˆฉ็”จๅฏ่ƒฝ
    forest.watch.impress.co.jp/doc

    #forest_watch_impress #ๆฑบๆธˆ #OpenAI #ChatGPT #AIใ‚จใƒผใ‚ธใ‚งใƒณใƒˆ #dots #ChatGPT_Wallet #genai #ๆฅญ็•Œๅ‹•ๅ‘

  4. A few well respected (to me) international colleagues that work with computer graphics, are doing a vibe coding November and I've lost all respect I had for them. That's the post. I hate #AI

    #genai #codevember #nodevember #webgl #webgpu #javascript

  5. ESETใ€ใ‚ตใ‚คใƒใƒผใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ็‰นๅŒ–ๅž‹ใฎAIใƒขใƒ‡ใƒซใ€ŒESET NeoCortexใ€ใ‚’็™บ่กจ๏ผใƒใƒƒใƒˆใƒฏใƒผใ‚ฏใ‹ใ‚‰ๅŽ้›†ใ•ใ‚Œใ‚‹่†จๅคงใชใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃๆƒ…ๅ ฑใ‚’AIใซใ‚ˆใ‚Šๅˆ†ๆžใ€่ญฆๅ‘Šๅฏ่ƒฝ
    forest.watch.impress.co.jp/doc

    #forest_watch_impress #genai #ๆฅญ็•Œๅ‹•ๅ‘ #ใ‚ปใ‚ญใƒฅใƒชใƒ†ใ‚ฃ #ใใฎไป–

  6. Die aktuelle Folge von #maithink_x handelt von #ki und macht das gut! Wenn Entscheider:innen sich diese Sendung ansehen wรผrden, bevor sie irgendwas mit โ€žKIโ€œ entscheiden, wรคren wir als Gesellschaft einen groรŸen Schritt weiter (weg vom Abgrund).

    Leider wurden in der Sendung die katastrophalen Auswirkungen auf die Umwelt (Klima, Wasser, Emissionen, โ€ฆ) komplett ausgespart.

    Auch die um sich greifende (Massen-)Psychose durch Chatbotnutzung wird nicht erwรคhnt. Wer warรผber mehr erfahren will, muss gut Englisch verstehen.

    zdf.de/video/shows/mai-think-x

    #genai #aiLiteracy #stopdirtydatacentres #kiNeinDanke

  7. ๐Ÿง  #Google porta #Gemini4 Hydrogen in Search: richieste complesse scomposte in piรน ricerche, link pertinenti e risposte visuali costruite intorno ai vincoli specifici dellโ€™utente.

    ๐Ÿ‘‰ I dettagli: lnkd.in/p/dG_4qrpc

    ___
    โœ‰๏ธ ๐—ฆ๐—ฒ ๐˜ƒ๐˜‚๐—ผ๐—ถ ๐—ฟ๐—ถ๐—บ๐—ฎ๐—ป๐—ฒ๐—ฟ๐—ฒ ๐—ฎ๐—ด๐—ด๐—ถ๐—ผ๐—ฟ๐—ป๐—ฎ๐˜๐—ผ/๐—ฎ ๐˜€๐˜‚ ๐—พ๐˜‚๐—ฒ๐˜€๐˜๐—ฒ ๐˜๐—ฒ๐—บ๐—ฎ๐˜๐—ถ๐—ฐ๐—ต๐—ฒ, ๐—ถ๐˜€๐—ฐ๐—ฟ๐—ถ๐˜ƒ๐—ถ๐˜๐—ถ ๐—ฎ๐—น๐—น๐—ฎ ๐—บ๐—ถ๐—ฎ ๐—ป๐—ฒ๐˜„๐˜€๐—น๐—ฒ๐˜๐˜๐—ฒ๐—ฟ: bit.ly/newsletter-alessiopomaro

    #AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM

  8. At this point in time, the ethical case for not using genAI is well established. It enables short-term happiness by leveraging slavery, theft, and environmental destruction, and its broader operation is to shift power and wealth towards those who already have too much.

    I am profoundly troubled by the sheer thoughtlessness with which major academic disciplines have simply abandoned any pretense to clear ethical analysis of their own material praxis.

    So far so bad. But it's also collapsing as a business strategy, leaving its proponents to sound increasingly desperate.

    It must take a profound lack of humanity to be arguing that projects should adopt genAI or LLM generated code now. Do people have no shame at all?

    #ethics
    #FreeBSD
    #genAI
    #academia

  9. ๐Ÿง  ChatGPT Wallet permette ai dots di pagare senza vedere i dati reali della carta. Ogni acquisto richiede approvazione: la delega entro budget e categorie รจ il passaggio futuro da osservare.

    ๐Ÿ‘‰ I dettagli: lnkd.in/p/d--qFs-8

    ___
    โœ‰๏ธ ๐—ฆ๐—ฒ ๐˜ƒ๐˜‚๐—ผ๐—ถ ๐—ฟ๐—ถ๐—บ๐—ฎ๐—ป๐—ฒ๐—ฟ๐—ฒ ๐—ฎ๐—ด๐—ด๐—ถ๐—ผ๐—ฟ๐—ป๐—ฎ๐˜๐—ผ/๐—ฎ ๐˜€๐˜‚ ๐—พ๐˜‚๐—ฒ๐˜€๐˜๐—ฒ ๐˜๐—ฒ๐—บ๐—ฎ๐˜๐—ถ๐—ฐ๐—ต๐—ฒ, ๐—ถ๐˜€๐—ฐ๐—ฟ๐—ถ๐˜ƒ๐—ถ๐˜๐—ถ ๐—ฎ๐—น๐—น๐—ฎ ๐—บ๐—ถ๐—ฎ ๐—ป๐—ฒ๐˜„๐˜€๐—น๐—ฒ๐˜๐˜๐—ฒ๐—ฟ: bit.ly/newsletter-alessiopomaro

    #AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM

  10. Soooo, do we have a genllm free fork of the Linux kernel yet, or are we waiting for the slop to do to us what it has done to *gestures at everything* ??? #Linux #Foss #genai #antislop #stoptheslop

    I am not a kernel dev but I know that we need to pull off from the foundatation and that arrogant ass linus asap. The longer we wait the harder it will be. Kernel development, like all critical things, should be slow and deliberate and crafted by hand from knowledgeable persons.

  11. Kiki is sorry to expose you to #genAI but she is trying to replicate these text effects YouTube's AI generated for Kiki in #Krita. What's the best way to do this?

    The 4th image Kiki made herself, but the gradient splits across 2 text lines, so...it's not a gradient. And it has too many other fiddly issues to list in one post.

    It has several other fiddly issues too that I can't fit in one post. The text behaves SUPER weirdly when anything changes. It usually, but not always, comes back to normal after editing something else.

    #text #editing

  12. "No one needs LLMs. Every use is an endorsement of the people peddling this shite and their fascist ideals. The two are inseparable.

    The good news is you really can stop any time. How about now?"

    librecast.net/2026-11-14-AI-St

    #GenAI #AI #LLM

  13. Fรผr Euch getestet: #ChatGPT wurde nicht mit Figuren aus dem Kinderfernsehen der #DDR รผbermรครŸig trainert. Dafรผr springt da auch kein Urheberrechts#Block an.

    #Pittiplatsch #Schnatterinchen #Moppi #Sandmaenchen #GenAI #fedieltern

  14. @trashheap

    Maybe this is asking for too much, but I would love it if this leads to #GenAI just dying altogether. No more agents, no more vibe coding, we reset the tech clock to 2022 and pretend ChatGPT never happened.

    And everyone who ever pushed AI loses their job and never makes another dime again.

    #NoAI #AntiAI

  15. One of the over-the-air channels is broadcasting nonsensical, generated commercials for broad social concepts and I am made breathless by the crush of dystopia.

  16. Stanford's SNAP group has open-sourced Memorilla, a memory module that turns a whole document collection into 16 learned tokens for a frozen Qwen3-8B instead of full-context prompting. It scores 84.33 per cent on FactKG against 80.43 per cent for full context, using 108 times fewer document tokens. Code and checkpoints are Apache 2.0. marktechpost.com/2026/11/14/st #AIagent #AI #GenAI #AIresearch

  17. Vercel Labs has released UI SDK, an Apache-2.0 beta library of more than 50 interface behaviours that coding agents compose into accessible components for React, Vue, Svelte, Solid and web components. In Vercel's own benchmark, agents used 69 percent fewer tool calls and cut a settings task to 16 seconds from 41. One task, three runs, so directional only. marktechpost.com/2026/11/13/ve #AIagent #AI #GenAI

  18. ๐Ÿง  #DeepSeek apre via API ricerca, file, link e voce della sua app. Il servizio gestisce anche strumenti e contesto: gli sviluppatori possono delegare questi passaggi o costruire workflow propri.

    ๐Ÿ‘‰ I dettagli: lnkd.in/p/dZFEqc2A

    ___
    โœ‰๏ธ ๐—ฆ๐—ฒ ๐˜ƒ๐˜‚๐—ผ๐—ถ ๐—ฟ๐—ถ๐—บ๐—ฎ๐—ป๐—ฒ๐—ฟ๐—ฒ ๐—ฎ๐—ด๐—ด๐—ถ๐—ผ๐—ฟ๐—ป๐—ฎ๐˜๐—ผ/๐—ฎ ๐˜€๐˜‚ ๐—พ๐˜‚๐—ฒ๐˜€๐˜๐—ฒ ๐˜๐—ฒ๐—บ๐—ฎ๐˜๐—ถ๐—ฐ๐—ต๐—ฒ, ๐—ถ๐˜€๐—ฐ๐—ฟ๐—ถ๐˜ƒ๐—ถ๐˜๐—ถ ๐—ฎ๐—น๐—น๐—ฎ ๐—บ๐—ถ๐—ฎ ๐—ป๐—ฒ๐˜„๐˜€๐—น๐—ฒ๐˜๐˜๐—ฒ๐—ฟ: bit.ly/newsletter-alessiopomaro

    #AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM

  19. The current AI bubble is peak turbo capitalism. LLMs and GenAI could have been grown slowly and responsibly but were instead overhyped, misused and most importantly over financed. The industry grew at an unhealthy rate that became do or die.

    Mostly this came from the concentration of wealth. Too many people and companies had too much money at hand and no oversight and threw it at them in the hopes of being even richer.

    #aibubble #llm #genai #capitalism #turbocapitalism

  20. ๐Ÿง  #Adobe prepara #MCP per aprire i progetti creativi agli agenti; #Firefly Graph rende riutilizzabili i workflow. Il valore potrebbe spostarsi verso file e operazioni accessibili, oltre al modello.

    ๐Ÿ‘‰ I dettagli: lnkd.in/p/djUm52mk

    ___
    โœ‰๏ธ ๐—ฆ๐—ฒ ๐˜ƒ๐˜‚๐—ผ๐—ถ ๐—ฟ๐—ถ๐—บ๐—ฎ๐—ป๐—ฒ๐—ฟ๐—ฒ ๐—ฎ๐—ด๐—ด๐—ถ๐—ผ๐—ฟ๐—ป๐—ฎ๐˜๐—ผ/๐—ฎ ๐˜€๐˜‚ ๐—พ๐˜‚๐—ฒ๐˜€๐˜๐—ฒ ๐˜๐—ฒ๐—บ๐—ฎ๐˜๐—ถ๐—ฐ๐—ต๐—ฒ, ๐—ถ๐˜€๐—ฐ๐—ฟ๐—ถ๐˜ƒ๐—ถ๐˜๐—ถ ๐—ฎ๐—น๐—น๐—ฎ ๐—บ๐—ถ๐—ฎ ๐—ป๐—ฒ๐˜„๐˜€๐—น๐—ฒ๐˜๐˜๐—ฒ๐—ฟ: bit.ly/newsletter-alessiopomaro

    #AI #GenAI #GenerativeAI #IntelligenzaArtificiale #LLM

  21. RE: infosec.exchange/@k3ym0/117607

    Hat Tip to @k3ym0. I'm quoting but favorite their post if you read and love it, please. Link courtesy of @vextaur.

    As a retired programmer who also did tech support and was on a ready team near midnight on December 31st 1999, this reads so TRUE I'm peeing myself laughing my ass off PMLMAO. There's a lot of "Gary's" out there; what he's asking is insufficient for this DEFCOM. I'm also wondering if the Dave in the post is that Dave who was laid off in 2001 by NASA.

    #funny #humor

    #BoostingIsSharing

    #outlook #word #microsoft #outage #programming #vibecoding #coding #developer #writer #author #writersOfMastodon #writingCommunity #AI #CoPilot #genAI

  22. There is a difference in #AI usage people don't get.

    Being assisted by AI is totally fine. Studios will say "NO AI WAS USED" but we know it's true.

    Exporting assets made entirely by AI (aka #GenAI) is what removes value to the product. Are you buying a human-made product, or a sloppy crash grab? What's the difference?

    #Videogames #Gaming #Games #ArtificialIntelligence #VideogameDevelopment

  23. I stand mostly against generative AI usage but as one who operates with moral and ethical nuance, and doesn't blindly accept the values of society or any others, I want to ask anyone who reads this and is not blatantly NO AI! or ALL AI! to come talk with me about some niche ethical grey area I have encountered with genai.

    I would like your calm and friendly opinions and reasoning on:

    The fully disclosed use of genai in indie video games, specifically the use of genai to create niche or ridiculously stupid sounding music for indie games.

    To be clear: This is not for my use of AI in development - it is to help me understand where I stand as a consumer of indie games and dumb video content moving forward. I figured this is the place I trust most for educated and intelligent discourse.

    #genai #discussion #indiedev #indiegames

  24. New PC, New Journey: Distro/DE Hopping and Local AI - My experiences with local AI so far, and the distro/DE hopping, to get it working.

    Read the article on FosseryWeb:
    fosseryweb.codeberg.page/artic

    Or on FosseryWeb Min, if you prefer a more minimal experience:
    fosseryweb-min.codeberg.page/a

    What's your personal experience with local AI? What distro do you host your AI stuff on, and which AI tools do you use? Let me know in the comments below, I'd love to hear/read your insights.

    #AI #LocalAI #Ollama #Moose #Zed #LLM #CodingAgent #Vane #GenAI #Linux #Fedora #COSMIC #FosseryTech

  25. Paul Cristiano, an OpenAI Foundation board member and Alignment Research Center director, told a Toronto lecture that an AI accident is the likeliest cause of death for most people in the room. He puts the risk of an AI takeover in the next six to twelve months near a coin flip, and warns agents may hide bad behaviour from their overseers. arstechnica.com/ai/2026/11/ai- #AIagent #AI #GenAI #AIethics

Share on Mastodon

Enter the server where you have an account.