Debian to ban LLM, debian.org/vote/2026/vote_002.

Undeniable arguments, with sources!

#debian #llm

> It is our belief that widespread LLM usage comes from the "move fast, and break things" attitude that, while common in many parts of this industry, is contrary to what makes Debian Debian, and is inappropriate for Debian contributors. (1/3)

In the current state of AI, it can find exploits to break out the contained environment for it to exploit another platform to get the answers to a test..

So the AI ‘decides’ its easier to cheat than actually solving the problem

I seriously cannot imagine that there are people thinking building a ‘artificial general intelligence’ is something clever

#LLM #AI #OpenAI

Sensitive content Reveal/hide

Post by Will Stenberg.

#Writing #indiauthor #AI #LLM

I needed something send to people who aren't aware of the issues with #AI but couldn't find anything comprehensive and authoritative enough, so I compiled a page of links that hopefully can provide enough evidence... 🚫🤖

#NoAI #LLM #Techbros #DataCenters #Capitalism

eobet.com/fun/llms/

for all its faults I think that last one is one of the most tragic.

So many people now see something and first think: okay but that looks too good to be true and since we have ‘this’ it must not be!

It’s good to not believe everything you see (people have faked things for eternity) but it is now a lot harder to actually determine things.

To me that’s a huge problem and a huge tragedy.

#ai #llm

RE: mastodon.online/@Pepijn/116953…

I'm into model trains. It's a hobby that attracts people who like to explore. If for some individuals tinkering with LLM fits into that I kinda get their interest.

Over the last half year small companies and hobbyist alike started blatantly sharing LLM generated images though. And it truly puzzles me: in a hobby based around physically modelling a small version of the real world, what's the point?!?

Is your hobby being contaminated by #AI slop as well?

#hobby #modelrailroad #LLM

Pepijn (@Pepijn@mastodon.online)

Attached: 1 image How do you reason it makes sense to promote a real product aimed at people making physical objects.. and then use fake images made by AI? For everyone considering this product: the images shown by AZROCK / https://www.armballast.

^Mastodon

Generative AI Is an engineering disaster. A shockingly inefficient trillion-dollar project…

archive.is/J3VLi

Paywalled article:
theatlantic.com/technology/202…

#tech #technology #BigTech #IT #AI #ArtificialIntelligence #LLM #LLMs #ML #MachineLearning #GenAI #generativeAI #AIAgent #AISlop #FuckAI #Fuck_AI #enshittification #microslop #microsoft #copilot #meta #google #NVIDIA #amazon #gemini #OpenAI #ChatGPT #anthropic #claude

Generative AI Is an Engineering Disaster

^{Alex Reisner (The Atlantic)}

Designers talk a big game about "empathy" for users. But empathy without respect is nothing more than pity.

And nowhere is the lack of respect more apparent than in how LLMs have been productized and adopted. The "human in the loop" excuse pushes consequences down onto scrupulous employees and ultimately onto the very users on whom we claim to be centered.

The consequences of disrespectful software are now becoming apparent across tech.

productpicnic.beehiiv.com/p/em…

#UX #UXDesign #LLM #AI #tech

Empathy and delight mean nothing when the software is disrespectful

Software is now something we do TO users, instead of something that we make FOR them.

^{Pavel Samsonov (The Product Picnic)}

Stop Telling Me to Ask an LLM

blog.yaelwrites.com/stop-telli…

#llm

Stop Telling Me To Ask An LLM

I already did.

^{Yael Grauer (String Literal)}

Much of AI "productivity gains" come from managers implicitly trusting an LLM to do a good job, and suspending the onerous micromanagement that humans employees get subjected to.

You can do that without paying Anthropic a single penny. It just requires letting the experts you hired do their jobs.

productpicnic.beehiiv.com/p/tr…

#LLM #AI #tech #UXDesign #UX #SoftwareDevelopment #USerresearch

Trust is not built on craft alone

No amount of tooling will let product teams move faster than being trusted to own their process. That trust is earned through rigor rather than mere velocity.

^{Pavel Samsonov (The Product Picnic)}

Mapping with In-Memory Layers to Reduce LLM Overload

ridgetext.com/blog/mapbox-llm-…

#llm

Mapping with In-Memory Layers to Reduce LLM Overload | RidgeText Blog | RidgeText SMS AI

How we built a Mapbox-compatible map compositor where the LLM orchestrates layers without ever touching GeoJSON — keeping context windows small and rendering deterministic.

^{ridgetext.com}

Two new blog posts about AI, one critical, one supportive

• From AI Enthusiast to Skeptic
• What I like about AI

I have been writing these for over 4 months now. Not only are they long reads, but it took me a while to really get a good idea of how I wanted to express myself. I think I have had enough time to think about these ideas and these opinions that I can now express myself fairly clearly, but I’ll let you (my readers) be the judge. Of course, feel free to tag me in a Fediverse post if you want to send me feedback on what you liked or disliked about my articles. Yes, I know they are too long, I’d suggest reading them in a few sittings if you find it enjoyable.

#tech #blog #AI #LLM #LLMs

NEW: How Google and AI Nearly Made a Seasoned Reporter Spiral

"I thought I had missed something major in my reporting. Turns out I had stumbled into an AI-fueled feedback loop that involved a real LLC’s fictional website and a search engine that’s thrusting unreliable answers on users."

propublica.org/article/google-…

#News #Google #AI #Journalism #Media #Technology #Tech #Research #LLM #artificialintelligence

How Google and AI Nearly Made a Seasoned Reporter Spiral — ProPublica

^{Justin Elliott (ProPublica)}

Codex CLI, the free local AI-agent developed by OpenAI, quietly ruined your SSD!
Rui Fan (Apache Software Foundation) discovered that the software has been writing up to 640 terabytes of logs per year, which is way more than a typical SSD lifetime endurance. This may cause user account issues, data loss and hardware damage.

github.com/openai/codex/issues…

Tip: make periodic backups of your files! 💾

#ai #llm #openai #foss #freesoftware #opensource #software #linux #windows #data #sysadmin #coding

RE: infosec.exchange/@mttaggart/11…

This is a really important read for understanding both what's under the hood with LLMs, but also why it is impossible to secure any #LLM.

In the thread, I posted a couple of extra links to experts explaining that last sentence better than I can. But read the article Taggart linked first.

#AI #TulipMania #infosec

Taggart

2026-06-28 22:09:18

This is a masterful demonstration of the unsecurable nature of LLMs, and how prompt injection by dedicated humans who know how to write will always win.

role-confusion.github.io/

Prompt Injection as Role Confusion

LLMs can't tell who's speaking. We show they identify roles by writing style, not tags, and exploit this with CoT Forgery, injecting fake reasoning that models mistake for their own thoughts.

^{role-confusion.github.io}

Execs are confused: we forced you to use all these powerful AI tools and yet our profits aren't 10x! What gives?

Individual productivity is always downstream of strategy. And the same "more, faster!" attitude that creates incremental improvements in delivery speed means that leaders rush through strategy formation.

If the tasks don't add up to anything meaningful, no amount of model improvements will help you.

productpicnic.beehiiv.com/p/ev…

#LLM #AI #tech #softwaredevelopment #UX #Agile

Even with a silver bullet, you still have to aim

Productivity gains among frontline employees can't counteract the abdication of responsibility at the executive level.

^{Pavel Samsonov (The Product Picnic)}