A thought that popped into my head when I woke up at 4 am and couldn’t get back to sleep…

Imagine that AI/LLM tools were being marketed to workers as a way to do the same work more quickly and work fewer hours without telling their employers.

“Use ChatGPT to write your TPS reports, go home at lunchtime. Spend more time with your kids!” “Use Claude to write your code, turn 60-hour weeks into four-day weekends!” “Collect two paychecks by using AI! You can hold two jobs without the boss knowing the difference!”

Imagine if AI/LLM tools were not shareholder catnip, but a grassroots movement of tooling that workers were sharing with each other to work less. Same quality of output, but instead of being pushed top-down, being adopted to empower people to work less and “cheat” employers.

Imagine if unions were arguing for the right of workers to use LLMs as labor saving devices, instead of trying to protect members from their damage.

CEOs would be screaming bloody murder. There’d be an overnight industry in AI-detection tools and immediate bans on AI in the workplace. Instead of Microsoft CoPilot 365, Satya would be out promoting Microsoft SlopGuard - add ons that detect LLM tools running on Windows and prevent AI scrapers from harvesting your company’s valuable content for training.

The media would be running horror stories about the terrible trend of workers getting the same pay for working less, and the awful quality of LLM output. Maybe they’d still call them “hallucinations,” but it’d be in the terrified tone of 80s anti-drug PSAs.

What I’m trying to say in my sleep-deprived state is that you shouldn’t ignore the intent and ill effects of these tools. If they were good for you, shareholders would hate them.

You should understand that they’re anti-worker and anti-human. TPTB would be fighting them tooth and nail if their benefits were reversed. It doesn’t matter how good they get, or how interesting they are: the ultimate purpose of the industry behind them is to create less demand for labor and aggregate more wealth in fewer hands.

Unless you happen to be in a very very small club of ultra-wealthy tech bros, they’re not for you, they’re against you. #AI #LLMs #claude #chatgpt

There was a small snafu with the Git repository of GTK which required rewriting the history of the main development branch to drop a large file that was committed by mistake; if you have a local clone, you may need to force an update. Nothing else should have changed, and the latest commit is: gitlab.gnome.org/GNOME/gtk/-/c…

Matrix... The modern IRC replacement. Working on phone, app, web browser...

Today greeted me with "your key store is not synchronized" (loose translation from Polish) which means BLABLABLAH to me.

And the only option is to continue with erasing whole history, reverifying all clients and contacts.

At same time FluffyChat on Android works fine, does not complain about anything.

WTH?

Fun fact: I routed Claude Code to a local model and captured its system prompt. It is about 16K tokens long, compared with 6.5K tokens for Codex and 5.5K tokens for Gemini CLI. Some of the length is due to tool call descriptions, but I didn't have any additional MCP. No wonder why you can't effectively use local models with it unless you have a crazy machine that can run a bigger model capable of following lengthy instructions and handle very big context window. #LLM #AI gist.githubusercontent.com/chi…
#AI #llm

Wellp, the Qwen2.5 LLM model has been useful in increasing the productivity rate of my annoying legacy PHP code refactoring and modernization project by at least 100%. Why 100%? Because before I started using the model, I was doing nothing on the project, LOL.

Don't worry, I'm not being radicalized. AI in its LLM format is everything that I've always gone on and griped about here, and elsewhere. It has no situational awareness or context. It is not intelligent. It halucinates. It will lie to your face, with a smile on its own.
So it has wasted my time with all of this. But all of that aside, it has been able to provide very accurate one-liners or very short code blocks. I purposefully keep any examples for ingestion short and to the point. I always ask for output that is concise, accurate, and to the point. However, I've been treating it like a robot, not a human. I don't waste frivilous words that pander to feelings. Just straight to the point. I am trying to optomize my words for tokens used in prompt analysis, but also having to be as absolutely as specific as possible, as to not waste tokens on having it do useless work.
So I'm trying to optomize for a tool, not an intelligence, personality, or chat buddy.
Its just like a real big knowledge pool, and a crude understanding engine. Instead of looking it up across thirty manuals, or sections of manual, then pieceing it together in your own head, the computer is doing all that in seconds. Sure, there's a comprehension barrier, a language barrier, now and again, but that's why you the human are still needed in the loop. So stead of fighting it, I'm trying to use it. Its pretty neat that way. Just don't trust it, ever. And anyone giving this thing a fair chance will very quicly come to realize that. Humans are still needed, and this thing will not permanently kill any jobs that won't be replaced in some other field, very quickly. Its just not that good or smart. Its a glorified tape recorder that can find and stitch humanity's knowledge together quite quickly, with just enough errors to make it useless without the human behind the keyboard looking after it. But that's what we want, eh? If it got any better than this, then it probably will well and truly wipe out jobs.

Ik kan gewoon niet meer - net gehoord dat een van de allergrootste verwerkers van privégegevens binnen de Nederlandse overheid druk bezig is met zijn totale migratie naar Microsoft Azure. En de mensen die het tegen zouden kunnen houden worden buitengesloten & hebben het opgegeven. Stukje: berthub.eu/articles/posts/nee-…

reshared this

in reply to Jiří Eischmann

@sesivany Ty služby, kam jsi přešel, asi mají hromadu věcí někde na AWS, což? 😅

Je to vznešená snaha, ale ten ekosystém je složitý a (pouze) osobní zodpovědností to vyřešit nejde. Když hovno bouchne do větráku, budeš nejspíš stejně nahraný, jako všichni ostatní.

Možná budeš aspoň na ty alternativy zvyklejší a zatímco my budem prskat nad „tím strašným UI”, ty už budeš mít vychozené klávesové zkratky. Ale to je asi jako mít při přípravě na atomovku schované oproti ostatním pět konzerv fazolí navíc. Lepší než nic, ale má to nakonec vlastně reálný efekt?

in reply to Honza Javorek

no, nemají. To jsem si vyzkoušel při posledním výpadku AWS. O výpadku BigTech se dozvím buď z novin nebo že mi nefungují nástroje v práci, ale ne to, co používám osobně. Já totiž dost selfhostuju a tam, kde ne, vybírám pro důležité věci služby, které mají servery opravdu v EU. Nějaký odsun od BigTech praktikuju už 10 let, takže teď to bylo jen o tom to dotáhnout.
This entry was edited (1 hour ago)

This is yet another train with a captive portal at 10.230.0.1, with that address resolving to exactly nothing. This is one of the major Polish train carriers, the most expensive kind of train to be had, and that issue has been unfixed for at least half a year.

At least it's not Deutsche Bahn I guess.