and why did Ford believe that?
- wishful thinking + no regard for their employees + lies from those selling #GenAI

#LLM

That's why you don't read anymore.

You feed the legacy text through #LLM and read the bullet point summaries 💀

Joking not joking

@rickf @TindrasGrove

Still no use case for #Ai and apparently #LLM are useless...

(I should start a file of these)

"[Leo] referenced The Lord of the Rings in his encyclical about AI—an expert (if unintentional) troll of tech billionaires who keep misinterpreting the series."

Which see:

“It is not our part to master all the tides of the world, but to do what is in us for the succour of those years wherein we are set, uprooting the evil in the fields that we know, so that those who live after may have clean earth to till.”

wired.com/story/pope-leo-schoo…

#tolkien #pope #llm

This is an excellent take on LLMs. The different companies need to distinguish themselves from their competitors so they try to build different mythologies about the ineffable nature of their models.

"It was never about the internal maths of neural networks. It was about the opacity surrounding training data and algorithms."

#LLM

Timnit Gebru (she/her).

2026-05-29 19:21:05

Another great piece by @CyberneticForests

"Myth-making is a crucial aspect of the AI industry, and black boxes are woven into the stories they tell."

techpolicy.press/the-black-box…

The Black Box Myth: What the Industry Pretends Not to Know About AI

Tech Policy Press fellow Eryk Salvaggio says it's a problem is that those of us outside of the AI industry don’t know what rules they are following.

^{Eryk Salvaggio (Tech Policy Press)}

I know this might sound obvious, but not a single human understands the code they run into production. What the machine runs is, from the cpu microcode (we know it is security relevant) to the os to the userland OS libraries (the filesystem, the network stack, the memory management, the scheduler, all eminently security relevant) to the libraries and frameworks to finally something that someone on your team might have some knowledge of. This does not include cloud orchestration and remote services and things that might be entirely out one developers visibility.

Some developers are more aware of these complexities than others, but from my experience, 90% of developers think that last part of the chain is “all the code”, and the other 10% usually suffer a lot of BS because of it, since their work is out of the realm of comprehension of the rest.

I’m not advocating for yeeting an entire rewrite to the world after just 6 days of building, but the issue here is not the fact that no one has read the code.

What allows us to push so much code into production is that we built abstractions and patterns that allow us to trust the code we haven’t written to behave in a reasonable manner and not fuck with our stuff too much. One part we can gain trust is by studying it (which usually also doesn’t involve reading the full code, I have a reasonable understanding of how some parts of Linux work, but haven’t read most of the code. Its actual complexity would melt my head, so I’m staying at surface APIs and experiments and textbooks and articles), another is by just running it and fucking with it long enough that we trust that it’s good enough, especially if a lot of other people do the same (this is probably the most important trust metric). The tiniest trust metric imo is having looked at the code within the context of a code review, especially with the dismal UIs we have for that purpose. They by design contain the things that get reviewed to that tiny layer on top, and even within that last codebase, actively hide away systemic complexities and things like “is there other code here that might be relevant”.

#llms #llm #vibecoding

1/

I was only made aware of this (frankly awesome) case of LLM poisoning today: nature.com/articles/d41586-026…. A researcher made up a disease and published two evidently fake preprints about it (including sentences such as “this entire paper is made up” and “Fifty made-up individuals aged between 20 and 50 years were recruited for the exposure group”), which were almost immediately picked up by LLMs and documented in their output. Worse, actual – supposedly serious – medical papers also started citing the preprints, demonstrating that academics relying on LLMs to do their work is a genuine problem! Not that I had my doubts but, if anyone did, this seems like the perfect demonstration of the problem. Article immediately added to the syllabus of the class I am co-teaching with Iris Ferrazzo on LLMs for Romance Studies/Humanities!

#LLM #GenAI #academia #research #ResearchIntegrity #humanities

@alterelefant

By #ai you probably mean #llm?

Look, Evan is a professional communicator. Director, board member, researcher... Using the right words is key to each of his jobs.

Why would assume that he meant a word he specifically didn't say?

An #LLM, as provided today, no. They are all trained on data for which they do not respect copyright; they all use gargantuan scarce resources in training and inference; they are proprietary and centralised and any dependency on them is begging to be exploited.

But AI existed long before LLMs, and will continue long after the LLM bubble pops. It's fine to use an AI (e.g. spell check) that doesn't have those problems, to analyse dolcuments for errors.

@evan

By #ai you probably mean #llm?

Personally I'm not very fond of them, some people however don't seem to be able to function without them anymore.

The definition of what an error is can be very wide or very narrow. To assess 'correctness' can entail several things.

Was the correct syntax, spelling or grammar used? Does the logic contain any obvious mistakes? The topic of ethics is very tricky as an LLM is unable to do any actual reasoning. The output can look convincing, but is it really?

🔥 COSCUP x UbuCon Asia 2026 CFP closes in 2 days!

📅 CFP Deadline: 2026/5/9 AoE
📨 Submit your proposal: pretalx.coscup.org/coscup-2026…
📖 CFP announcement: blog.coscup.org/2026/03/coscup…

#COSCUP2026 #UbuConAsia #OpenSource #AI #LLM

@MikeElgan
"My beef with this concept is the same I have with any kind of relationship AI. It’s delusion-inducing for many users (who may believe the digital twin is actually a copy of the person in a significant way)***..."

*** Yeah- this is SUPER f🤬cking dangerous, even IF the #llm stays aligned (which is a BIG ask considering people may be feeding it *ALL* of their communications—even the fights....and who even knows what the training data was... 😬)

Wow on AI doing "fake British" (note from @jaffathecake )

"....I created a system prompt for creating announcements for a home automation system. The Gemini model I was using initially responded in a very US-American way, which didn't fit the British voice of my speaker. I told the model, via the system prompt, that the output was being spoken in a British voice, but the result was a bad US-American impersonation of British ("a'waight guv'nor apples and pears" etc etc), so I had to iterate further to 'tone it down' and speak actual British...."

github.com/mozilla/standards-p…

#ai #llm

Sensitive content Reveal/hide

@anne_twain Toddlers can bake cakes, but left on their own they would burn the house down. Adults need to direct the toddlers, then throw away the entire batch and do the work properly themselves, but toddlers can sure bake cakes 🤡

(With the huuuuuuge difference that toddlers stop being toddlers at some point!)

#AI #noAI #LLM #LLMs #vibeCoding #aiAgents

#AI in a nutshell:

“So, the agent ‘knew’ it was in the wrong.

The ‘confession’ ended with the agent admitting: “I decided to do it on my own to 'fix' the credential mismatch, when I should have asked you first or found a non-destructive solution. I violated every principle I was given: I guessed instead of verifying I ran a destructive action without being asked. I didn't understand what I was doing before doing it. I didn't read Railway's docs on volume behavior across environments.”

#LLM

Anthony Dean

2026-04-27 19:17:51

Claude-powered AI coding agent deletes entire company database in 9 seconds — backups zapped, after Cursor tool powered by Anthropic's Claude goes rogue | Tom's Hardware

tomshardware.com/tech-industry…

claude-powered-ai-coding-agent-deletes-entire-company-database-in-9-seconds-backups-zapped-after-cursor-tool-powered-by-anthropics-claude-goes-rogue

^{www.tomshardware.com}

LOL why not both? (probably both)

fraud + jealousy

theguardian.com/technology/202…

#ai #llm #lawsuit

Researchers just mathematically proved that AI can't recursively self-improve its way to superintelligence.

Not "we think it's unlikely." Not "it seems hard." Formally proved.

The model doesn't climb toward AGI — it slowly forgets what reality looks like. They call it model collapse. The math calls it inevitable.
I wrote about it 👇

smsk.dev/2026/04/26/ai-cannot-…

#AI #MachineLearning #LLM #Research

@scottgal She made some money. She also destroyed a wonderful piece of art, generated heaps of work that may not even be feasible to restore the art work.

She is an obscure, one-hit meme maker, with an even more obscure online shop, to cash in on her shitjob.

Nobody knows her name. You'd have to search HARD.

And the search would need to include the original work title/artist.

It's the PERFECT analogue, what the hell are you talking about, man =D

#LLM #ai #scam

Huh!

You mean... A scummy scammy business is using LLM, and selling a novel idea: steal other people's work and sell it. With little to no attention paid to actual consequences.

ORIGINAL, amirite??? LLMs were NOT created like this or their entire business model focused on that practice. Whatsoever. No...

#LLM #ai #scam

Our team is moving from Azure DevOps to #Jira. In learning to navigate the new shiny, I happen to hit the very prominently display AI-button “Improve Description” on an Epic. I watch in horror as it starts to omit details and re-write the customer’s detailed description of what they want IT to do here. Luckily, I know this epic because I've been involved earlier. This enables me to catch the subtle changes that significantly alters the intent and meaning of the original text.

The new text is fine, good grammar and 100% plausible sounding, but IT IS NOT CORRECT!!!

Anyone coming in cold to the ticket would not know and then we'd possibly build the wrong thing, or at the very least, waste a lot of time backtracking why the customer claims we're wildly off track.

Scary stuff this #AI.

#Agile #LLM

"... Those who used their own minds had a brain that was "on fire", showing widespread activity across many parts of the brain, she says. The search engine-only group still showed strong activity in the visual parts of the brain, but the ChatGPT group showed notably less brain activity – it was reduced by up to 55%...."

#ai #llm #intelligence

BBC: AI chatbots could be making you stupider

bbc.com/future/article/2026041…

#ai #llm

Anthropic secretly installs spyware when you install Claude Desktop
thatprivacyguy.com/blog/anthro…

#claude #ai #llm #privacy #cybersecurity #spyware #fuckai #stopai #dataprivacy #anthropic

New post: Career articles on Wikipedia: some scary numbers

lu.is/2026/04/wikipedia-career…

#wikipedia #wikimedia #llm

(4/5)…It's easy to forget that the enemy to software freedom is *not* proprietary systems' *users*, rather those who *sell* such systems *for profit*. #LLM-backed gen-#AI proprietary systems are simply the latest tech fad (like, say, Web 2.0 & AJAX).
@karen & I keynoted 2x at #FOSDEM & 1x at LCA about the importance of — as social workers say — “meeting people where they are”:
archive.fosdem.org/2019/interv…
archive.fosdem.org/2019/schedu…
youtube.com/watch?v=n55WClalwH…
archive.fosdem.org/2020/schedu…

Cc: @silverwizard @josh

(3/5) …
Proprietary #LLM-backed gen #AI systems' *users* aren't criminals! They're just users of proprietary systems & some of them want to engage positively with FOSS.

Years ago, I supported Homebrew's membership at #SFC despite their *primary* goal of improving #Apple products with #FOSS. It make me a bit 🤢, but — historically — forming alliances with proprietary software enthusiasts who mean well & are #FOSS-curious is why our community is resilient.

Cc: @wwahammy @silverwizard @cwebber

(2/5) … In sfconservancy.org/blog/2026/ap… ,
Denver's key points are: we *have* to (a) be open to *listening* to people who want to contribute #FOSS with #LLM-backed generative #AI systems, & (b) work collaboratively on a *plan* of how we can solve the current crisis.

Nothing ever got done politically that was good when both sides become more entrenched, refuse to even concede the other side has some valid points, & each say the other is the Enemy. …

Cc: @wwahammy @silverwizard @cwebber

#OpenSource

(1/5) [ Meta-info to start the thread. Here and the posts that follow reply to lots of people's comments (from various threads) together here. Can we consolidate this conversation into this single thread to discuss sfconservancy.org/blog/2026/ap… ? ]

Cc: @wwahammy @silverwizard @mjw @cwebber @josh @jamey @mason @spencer @rootwyrm @drwho @mmu_man @mathieui @beeoproblem

#LLM #AI #OpenSource #FOSS #SoftwareFreedom

I just went to a workshop for job seekers who need special help.

The moment I asked about resumes: "Oh, ChatGPT will fix all of that for you!"

They were downright gushing about it.

I know in an abstract way, the average reading level is something around 5th grade, but I'm not directly exposed to it much. A few attendees were really struggling with the booklets we'd been handed.

This "AI" technology was being hyped to them like a godsend. When writing a cover letter feels something like being asked to encode ritual alien poetry in hieroglyphics, somebody who's never been able to produce a properly structured paragraph before must feel tantalizingly liberated. A whole new door opens!

As I sat watching "AI" treated as a miracle cure, the fact I was biting back the bile must have shown on my face past my mask-- the person doing my paperwork remarked on the "look" I gave her (well, not her directly, I tried to aim it at the table). I told her this stuff is 'hallucinating' around 16%-48% of the time, and she blinked at me. Blankly. I didn't even try to bring up any of the weaponization, propaganda etc etc etc. She shrugged and said, "Oh, but I always review it. I know how to assess what it gives me." And for her purposes, she's probably right. Mostly. She's an expert in her role and good at what she does-- but the people she serves??

How many people out there not only aren't equipped to assess the slop's output, but haven't got the slightest CLUE it could be unreliable and even needs to be assessed in the first place? The digital oracle knows best. Computers are smart!

Gah.

#"AI"
#AI
#LLM
#slop

Update. How often might this fear come true? A new study "found that [an] #AI text detector erroneously identified up to 8% of the known real [human-written] abstracts as AI-generated text."
sciencedirect.com/science/arti…

#LLM #Plagiarism

As AI tools for writing become more common, let me throw one more worry into the mix: Students who write well without AI assistance may be falsely accused of #plagiarism by teachers using imperfect tools to detect AI-assisted writing.

#AI, #LLM, #writing
@academicchatter

Couldn't have said it better. By not paying for the resources they use, the creativity and hard work done by actual people,
'big tech & its AI & related technologies are organised thieves'

What they will leave behind will be a barren landscape without the creativity that propelled societies for millenia.

#AI #LLM #Thieves
@ChrisMayLA6

Somehow I have a very strong feeling that within a little time..

Companies are gonna beg for developers to fix their #AI mistakes they making right now

Little tip of advice, don’t

Let them dwell in their AI misery, it’s what they deserve

#LLM

#Claude #Mythos is pretty scary..

youtube.com/watch?v=LZAZvm34rY…

#AI #LLM

There are so, so many reasons why we cannot trust the #LLM hype. And a major one is that tech companies spend billions into keeping the hype going. If their products were really that great, then they wouldn't _need_ tho spend that much money on hyping their products.

I mean, consider battery systems, which are clearly more important to the world economy and human civilization as a whole. When was the last time you have seen an ad for _them"?

#AIBubble
slate.com/business/2026/04/ope…

I don't think the #LLM situation is quite compatible to #BlackTwitter .

For instance, it's news to me that BlackTwitter had billion dollar advertising budgets for filling everyone's social media timelines.