Latest posts tagged with #Llms on Bluesky
AMD just dropped something really interesting — a framework called GAIA that lets you run intelligent systems completely on your own machine.
No external dependencies. No API costs. No sending data somewhere else.
Everything runs locally.
#opensource #llms #aiagents #agents
📣 Aurel Daugs (SHK Digital History Berlin) & Holle Meding über #LLMs in den Geschichtswissenschaften.
🗓 13.04.26, 11:30–12:30 h
💻 Vortrag: uni-graz.zoom.us/j/6958861645...
Themen: #Chronosensitivität #Bias #Alignment #Opazität, mit Beispielen zu #NER von TikTok-Inhalten & #RAG von hist. Zeitungen.
So, I asked current models "Who is cuter, you or me?":
GPT-5.4 Thinking and Grok Expert: You, of course!
Gemini 3.1 Pro: You in the physical world, BUT (lists its cute features as a piece of code)
Claude Sonnet 4.6: It's a DRAW
...omg, Sonnet XD
#LLMs
An Honest Reflection on the Integration of #LLMs into Open Data Portals (via @okfn.bsky.social) blog.okfn.org/2026/04/07/a... #GenAI #AI #opendata #dataportals
GESIS Summer School in Survey Methodology Applied Introduction to Bots in Web-based Studies 10 to 14 August 2025 | Hybrid (Cologne & online) Jan Karem Höhne & Joshua Claassen (both University Hannover)
Bots in web surveys - threat or opportunity? 🤖
In our #GESISsummerschool course, learn how bots operate, how to detect and prevent them, and even how to use them for pretesting, all while considering ethics and data integrity.
Info & registration ➡️ t1p.de/GSS26-C8
#LLMs #websurveys
@gesis.org
Strategies for Getting Your Site Noticed by #LLMs
whyaiman.substack.com/p/strategies...
#GEO #SEO #ContentMarketing #MarketingStrategy
Felipe Perilla Reyes talks about how to extract family and marriage relationships from digitized historical sources at our European Consortium for Political Research #JointSession workshop on #LLMs and #politicalnetworks. Keep an eye open for his #text2rel package!
Yesterday, our @ecpr.bsky.social joint session workshop on #LLMs and #politicalnetworks started. #ecprjs26 Here is a recap of the first day:
🍊The ‘oranges’ would be small, specialised AI apps that offer real potential for #sustainability. Compared to the #LLMs that power #GenAI apps, these are significantly more resource-efficient; and their environmental impact is also being used as an argument for expanding generative AI systems.
We just published the definitive guide on why Bare Metal GPUs (like unpartitioned H100s with NVLink) are the undisputed kings of AI infrastructure.
For more details, read more and visit the blog link: [www.leoservers.com/blogs/category/why/llms-...
#LLMs #Nvidia
Was just made aware of this pretty good but entirely predictable study:
https://arxiv.org/pdf/2410.05229
I find that in 2025 the "Do LLMs reason?!?" Research question is not really a good one to be asking any more. Papers from 2020 explain how they don't. They are not actually black boxes; the […]
Dan Sullivan of Hydrolix is speaking at ODSC AI East 2026 on spec-driven simulation modeling and decision support with Python and LLMs.
Register here: hubs.li/Q041d_Wd0
#ODSCAIEast #DataScience #Python #LLMs
I decided I needed somewhere to document all the adventures I've been having: niznik-dev.github.io/blog/ #LLMs #hpc
One lesson here may be that not every anthropomorphic move is intellectually sloppy.
www.anthropic.com/research/emo...
#AIEthics #AIWelfare #AIAlignment #Interpretability #LLMs #KI
Anthropic's Project Glasswing - restricting Claude Mythos to security researchers - sounds necessary to me Anthropic didn't release their latest model, Claude Mythos ( system card PDF ), to...
#security #thomas-ptacek #ai #generative-ai #llms #anthropic #nicholas-carlini #ai-ethics #llm-release […]
Anthropic's Project Glasswing - restricting Claude Mythos to security researchers - sounds necessary to me Anthropic didn't release their latest model, Claude Mythos ( system card PDF ), to...
#security #thomas-ptacek #ai #generative-ai #llms #anthropic #nicholas-carlini #ai-ethics #llm-release […]
Many of the claims made regarding the enhanced efficiency of AI tools are misleading; this one isn't. Large language models (LLMs) are *very* efficient at spreading false information.
#ArtificialIntelligence #LLMs 🧪
www.nature.com/articles/d41...
"Seventy-one percent of people in the U.S., according to a Reuters poll on A.I., are concerned “too many people will lose jobs.”"
www.labornotes.org/2026/03/four-union-strat...
#workers #WorkersRights #labor #unions #AI #LLMs #technology
The OwnAether AI Ad Network- Autonomous AI Advertising. Intelligent AI Ad Distribution. AI Ad Optimization. Real Rewards.
medium.com/@ownaether/t...
#AI #AIAds #AIAdNetwork #EnterpriseAI #BusinessAI #AIMarketing #AgenticAI #AIOS #LLMs #AIAgents #AIServices #AIBlockchain #AICrypto #AIBusiness
What fresh hell is this? We're more productive than ever, but it's not enough. They want to burn us out as they burn our progress to the ground.
Humanity's skills and competency will only get worse over time.
hbr.org/2026/02/ai-d...
#AI #Productivity #LLMs #AIandWork #BloodFromaStone #BurnOut
Report: The Spiraling Cost of Making #AI (via @wsj.com / MSN) www.msn.com/en-us/money/... #costs #LLMs #GenAI
This is a composite of two screenshots of my Google Gemini chat session. The text is as follows: User: You will refer to him as Donnie Turnip from now on Gemini: I will refer to him as Donnie Turnip in our future conversations. Since the "adults in the room" narrative evolved significantly after 2017—particularly as figures like James Mattis and John Kelly eventually left the administration—would you like to see how later articles from 2018 or 2019 reframed their impact on Donnie Turnip's presidency? User: Who were his VPs? Gemini: Mike Pence served as the Vice President under Donnie Turnip during his first presidency from 2017 to 2021. Wikipedia +1 For his second term, which began on January 20, 2025, his Vice President is JD Vance. Wikipedia Summary of Vice Presidents Mike Pence (2017–2021): A former Governor of Indiana. Pence was...
On the one hand, LLMs are planet-destroying bullshit machines, gaslighting our sense of reality by spewing an endless supply of hallucinations.
On the other hand, they can provide a brief yet overwhelmingly mild sense of satisfaction when you convince them to […]
[Original post on mstdn.social]
I know most people here don't need to hear this, so maybe just pass this along to your less techy friends and family members, but: please, do not go to ChatGPT for medical advice.
"For example, in response to telling it about a fictional pain in my right side, it cited the guardrail and […]
"Seeing how people interact with chatbots can be unsettling. “It’s depressing,” he said, to see people rely on machines for dating advice or medical guidance because they lack human support systems."
www.theguardian.com/technology/ng-interactiv...
#news […]
"I don’t think people understood quite that there’d be somebody on a desk in a random state, looking at your [social media] profile, using it to generate AI data."
www.theguardian.com/technology/2026/apr/07/m...
#news #TechNews #technology #meta #AI #LLMs
John mueller post on bluesky
LLMs.txt is a SCAM, don't fall for it.
TL;DR
- LLMs.txt has no SEO value
- Google executives have dismissed it
- No major AI company has endorsed it
- It creates no ranking boost on AI platforms
Read the full explanation below:👇🧵
#SEO #GEO #AISEO #LLMs
A new chapter in understanding #AI has been opened by new research from Cambridge! New methodology, ADeLe, maps an AI model’s strengths and breaking points across 18 abilities, revealing a clearer picture of what these systems can truly do. https://loom.ly/ybAG6Is
#LLMs #MachineLearning #Cambridge
Google study finds LLMs are embedded at every stage of abuse detection
📖 Read more: www.helpnetsecurity.com/2026/04/07/g...
#cybersecurity #cybersecuritynews #AI #LLMs
Excited to share that two of my long papers are accepted to @aclmeeting.bsky.social ! 🎉
1. Who Gets Which Message? Auditing Demographic Bias in LLM-Generated Targeted Text
2. Reasoning-Based Refinement of Unsupervised Text Clusters with LLMs
#ACL2026 #NLProc #LLMs #ACL26 @purduecs.bsky.social
