#FuckLLM

TIL there are IPTC tags to add metadata to images indicating that they are made with AI/LLMs, who prompted them, and what system was used: iptc.org/std/photometadata/doc

This is actually good news. It means that

  • people/companies who generate images this way can be held accountable/identifiable by name
  • you may nag them continuously about tagging their shit
  • properly tagged images can be filtered out from search systems / image boards

#TIL #NoAI #NoLLM #FuckAI #FuckLLM

Oh my god I just tried iocaine for the first time and this is so much easier to configure than Anubis.

You literally just need the binary or container running and then point your reverse proxy to its port.

I love how its docs feel human instead of corporate.

And most importantly I don't need to configure OpenGraph for my link thumbnails to work!!!!!

#iocaine #Anubis #FuckAI #FuckLLM #NoAI #NoLLM

me·ta·phil, dermetaphil@chaos.social
2026-02-23

Warum heißt es „mit einem probabilistischen Sprachmodell generiert” und nicht „schwätzungsweise”?

#TypoDesTages #fuckLLM

@da_667

So I'm not the only one!!

Google Assistant has been enshittified! Every single advance into "AI" for Google Assistant past the initial genius ML model has been worse for simple use-cases.

Set a timer for [x].
Remind me to [y] at [z; time]
Remind me to [y] at [z; place]

What's the movie where Robin Williams gets his dick bit off? (Real life example that actually worked)

Assistant used to be fucking amazing for stupid shit like this. Things that notably helped my ADHD ass. Who's it better for now?

Burn our world. Eat every scrap of resource. For what?! An oracle of lies. A golem with no purpose.

#GoogleAssistant #GoogleGemini #FuckGemini #FuckLLM #FuckChatGPT

@alice

Les capsules du prof Lutzlutzray@mamot.fr
2025-02-27

How funny... a tutorial site (geeksforgeeks) presenting an example for numpy.vectorize() WITHOUT EVEN INCLUDING THE CALL IN ITS CODE... no human writer would have made this error.

#FuckLLM #enshitification #numpy #python

Client Info

Server: https://mastodon.social
Version: 2025.07
Repository: https://github.com/cyevgeniy/lmst