when using open source LLMs... 😂 @ carlosenog @ rdenadai @ vepo @ giggio

# TuneTuesday # WaitingSongs Al Green - Wait Here https://www. youtube.com/watch?v=w1dnM-8Igt Y&list=RDw1dnM-8IgtY&start_radio=1

# TuneTuesday # WaitingSongs Al Green - Wait Here https://www.

Software vulnerabilities and security tools releases

Between August 26–27, 2026, security advisories were published for WordPress TranslatePress, Tenable Security Center, Zimbra Collaboration Suite, and NethServer WebTop. CodeQL 2.26.4 was released, and the OWASP Top 10 For LLM Applications 2026 paper was published.

The Risk of Fine-Tuned Open-Weight Models · MSec Operations Blog https://www. msecops.de/blog/posts/backdoor ed-llms/

Time-series foundation models are bringing LLM-style ideas to forecasting. Sean Moran explains how t0-alpha works and what its reproducible results reveal about the state of the field. https:// toward

Time-series foundation models are bringing LLM-style ideas to forecasting. Sean Moran explains how t0-alpha works and what its reproducible results reveal about the state of the field.

When the Algorithm Enters the Classroom: A Critical Integrative Review of Large Language Models, Nursing Education Structural Gaps, and the Reconstitution of Professional Identity Computers and Educat

When the Algorithm Enters the Classroom: A Critical Integrative Review of Large Language Models, Nursing Education Structural Gaps, and the Reconstitution of Professional Identity Computers and Education: Artificial Intelligence Computers and Education: Artificial Intelligence · ScienceDirect Publication: Computers and Education: Artificial Intelligence · Yanjie Sun, Hongzhou Li # AIinHigherEd https://www.

Codeberg Bans AI Training on Platform Repositories

Code hosting platform Codeberg announced it is restricting AI development and prohibiting use of its repositories for training artificial intelligence models to protect open-source software commons. The policy received mixed reactions, with supporters praising FLOSS protection but critics arguing it limits platform utility.

Learning Programming in an Age of LLMs

Learning Programming in an Age of LLMs. On learning programming in an age of LLMs.

"Real talk: too many of you are using LLMs to generate content." # ai # content # attention https:// bcantrill.dtrace.org/2025/12/0 5/your-intellectual-fly-is-open/

I just wrote about the future of Open Source in a world where coding is "free". https:// blog.mempko.com/a-million-kaka pos/ # OpenSource # AI # LLMs # software # tech # economics

More FLOSS attrition due to LLM use https:// frvr.com/blog/news/ps5-linux-l ead-quits-as-open-source-projects-have-become-a-bunch-of-noobs-using-llms-that-they-dont-even-understand/

More FLOSS attrition due to LLM use https:// frvr.com/blog/news/ps5-linux-l ead-quits-as-open-source-projects-have-become-a-bunch-of-noobs-using-llms-that-they-dont-even-understand/

Best Open Weight LLMs in July 2026 (Tested & Compared)

Thanks to the awesome work of Jonathan Kimmit and the massive use of LLMs we now have an open source FPGA toolchain that can build a working LiteX design on a Xilinx Virtex7 FPGA

RE: https:// mastodon.green/@gerrymcgovern/ 117172792450861828 there are no "ethical LLMs", regardless of local models, regardless of open weights, regardless of renewable power, regardless of country

PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand"

PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand". PS5 Linux lead quits: "a bunch of noobs using LLMs" that "they don't understand".

https://www. theguardian.com/media/2026/aug /22/deadly-tiktok-motorway-driving-trend-horrifies-ireland

https://www.

John's colleague told him he didn't belong. It was just what he needed to hear L: https:// text.npr.org/nx-s1-5906734 C: https:// news.ycombinator.com/item?id=4 9076493 posted on 2026.07.27 at 18:45:3

Are LLMs bullshit machines?

Like many, I have too many open tabs on my browser. A new trend I spotted with myself was that among them too many were tabs with conversations with LLMs I'd come back later to. I never did. Bye bye t

I really like that we have open-source models that we can run locally and for free. So you don't have to pay OpenAI or Anthropic to use LLMs in your apps when testing. Let me show you how. # AI # LLM

PrismML deploys tiny LLMs on Qualcomm smart glasses

PrismML launched small language models optimized for Qualcomm-powered smart glasses on September 24, 2026. The company aims to enable on-device AI that leverages existing computing power more efficiently.

# Anthropic and # OpenAI think open weight LLMs are good enough now to eat their margins, ergo begging for gubmint regs to save their failing business from competition. The companies launching irrespo

Swallow LLM — 日本語に強いオープンなLLMを、アカデミックな研究開発から

SkillBloat: Token Amplification Attacks via Skill Injection in LLM Coding Agents

21929v1 Announce Type: new Abstract: Agent skills extend coding agents with task-specific instructions, scripts, and resources, but they also create a trusted instruction channel that can be abused beyond conventional security attacks. This paper studies token amplification through skill injection: an economic resource-abuse threat in which a malicious skill causes an agent to consume substantially more tokens than needed for normal task execution.

Sweet Jesus, why am I starting some casual reading on locally hosted open source # AI # LLMs ? I don’t need another stupid high frustration, high maintenance hobby. I can’t even keep my # nethack serv

I just wrote a whole article about an open-source tech topic, and LLMs didn't come up once. Not once. More than 3,500 words, and all of it about stuff a community is doing as a community, as people, s

[$] Considering the OpenMDW license

Mmm lotta grumbling about Debian’s LLM vote and like… the longer I’m in open source the longer the stated values feel like theater. The open source definition was sacred for years and used as a blunt

Validate User

📣 The R+AI 2026 Call for Proposals is open! How are you using R with AI, machine learning, LLMs, agents, coding assistants, data products, or responsible AI? Share your practical workflows, research,

I created a harness for running local LLMs. Core concept: Fair research. It only uses sources that don't rely on ads, like Wikipedia, Wikidata and open scholarly databases, so agent traffic doesn't ta

RE: https:// hachyderm.io/@thomasfuchs/1173 09282155310011 Regardless of what you think about LLMs, you should understand this. If you do, you can then use LLMs (if you so desire) to do actually usefu

An interesting argument applying Daniel Dennett’s idea of mental stances to LLMs. https://www. newyorker.com/culture/open-que stions/can-ai-go-rogue We take three stances toward other people: ⁃ intent

Use cheap, accessible infrastructure as a laboratory: old e-waste machines, home broadband, open-source software, tiny local LLMs and a low-cost VPS. Rather than beginning with "use a queue" / 'use Re

An open benchmark and language models for AI in aging biology

Open weights vs. closed: An AI civil war's afoot, and the stakes are existential

"We reserve the right to have automated LLM reviews on our project since we reserve the right to choose how development goes." "No!!! You can't ban LLMs in your project! That's against the spirit of o

GitHub - johnbean393/KeyType: An open-source Cotypist with macOS system wide AI autocomplete · GitHub Serving as an open-source alternative to Cotypist, this macOS utility provides system-wide AI tab-

Das Tempo der Anderen

Kimi K3 releasing on Hugging Face in 3 hours https:// huggingface.co/moonshotai/Kimi -K3 Exciting time for open LLM advocates and honestly with censorship removal techniques like https:// github.com/p

Run a full RAG pipeline privately with ChatGLM, Qwen2, or Llama3

I find LLMs to be really good tutors for things that are either new to me or that I'm revising, such as areas of maths I've not covered before, or an unfamiliar programming language, and so on. But ar

Paritok - Token Compression Proxy for LLMs | EveryDev.ai

LLMs are copies of each other, based on open source code published by Google in 2017. A core communication technique is copying the style of its user.. (Tech Bro) - "No, you don't understand! I have a

LLMs don't need to self-replicate and take over all hyperscalers/neoclouds to be a genuine danger to society, hacking just 1% of the industrial control hardware that's connected to the internet is way

RE: https://infosec.exchange/@drwhax/116997238394916522 This will probably piss people off, but it's true. If malicious actors are tentatively looking to destroy open source communities with their own

CyberFactory: Scaling Cyber Security Capabilities with Instances from the Wild

From Letters to LLMs: Jobs, AGI, and the Fear We Don’t Have to Carry

GCC bans LLM-generated code contributions except for tests

The GCC Steering Committee adopted a new policy declining any legally significant code contributions that include LLM-generated content, though test cases are exempted. The policy aims to protect the integrity of the open-source compiler project.

Attending AGNTCon + MCPCon in Japan? Don't miss Daniel Oh & Kevin Dubois's Workshop: "Enterprise Agentic AI: Architecting Autonomous Java Systems for Production" https:// github.com/quarkusio/quarkus/

Attending AGNTCon + MCPCon in Japan?

Starling: A New Linux Desktop Written In Swift, Own Wayland Compositor & Written With AI

Small open-weight LLMs outperform single model for malware analysis

An arXiv paper demonstrates that orchestrating multiple small, free-to-use language models can outperform a single large LLM for analyzing malware detonation reports covering filesystem, network, and process behaviors. The approach enables rapid interpretation of complex security data without expensive proprietary models.

GitHub - anuj0456/OpenArch: PyTorch implementations of modern open-source LLM architectures (Llama, Qwen, DeepSeek, Gemma, GPT-OSS, Kimi, and more) — written from scratch for readability and learning,

Nvidia-Led Coalition Signs Letter Supporting Open AI Weights

Nvidia and 24 other companies including chipmakers, cloud operators, and venture funds signed a letter supporting open-weights AI models as Washington considers restricting Chinese AI model development. OpenAI, Anthropic, and Google notably declined to sign.

Middle East tensions, Gaza crisis, AI adoption, website design debate

Incoherent mix: projectile strike in Strait of Hormuz with crew death and oil price rise; UN report on Gaza cropland destruction (3% usable); unrelated social media posts about website design and AI code testing.