this post was submitted on 12 Dec 2025
105 points (98.2% liked)

Fuck AI

4980 readers
1612 users here now

"We did it, Patrick! We made a technological breakthrough!"

A place for all those who loathe AI to discuss things, post articles, and ridicule the AI hype. Proud supporter of working people. And proud booer of SXSW 2024.

AI, in this case, refers to LLMs, GPT technology, and anything listed as "AI" meant to increase market valuations.

founded 2 years ago
MODERATORS
105
Red Hat pushing AI (fedoramagazine.org)
submitted 1 week ago* (last edited 1 week ago) by vogi@piefed.social to c/fuck_ai@lemmy.world
 

Users points out in comments how the LLM recommends APT on Fedora which is clearly wrong. I can't tell if OP is responding with LLM as well–it would be really embarrassing if so.

PS: Debian is really cool btw :)

you are viewing a single comment's thread
view the rest of the comments
[–] AA5B@lemmy.world -1 points 1 week ago* (last edited 1 week ago) (2 children)

Seems like everyone taking a left turn into crazy ville ……

This is a good approach.

  1. It’s an mcp server, a “bridge”. A standard way LLMs could talk to your system. It’s not an LLM. It doesnt mandate an LLM. It doesn’t tie you to a specific LLM
  2. It’s optional. Don’t use it, or don’t install it. No harm done. Even if it’s installed and running, if you don’t use an LLM with local access, no harm done.
  3. Even the increased attack surface is not a big deal since it is local, optional, and focuses on reading statuses rather than executing actions
  4. It’s an open standard. If you decide to use it with an LLM but don’t like the results, try a different LLM
[–] nupo@quokk.au 4 points 1 week ago (1 children)

This is a community for criticizing LLMs. There is, therefore, no positive interest in a bridge for LLMs. This community is optional. Don't subscribe, or don't read it. No harm done.

[–] AA5B@lemmy.world 1 points 1 week ago

Yeah, I suppose so. I find plenty to hate about the way companies and too many individual apps talk about LLMs. I really hate that it’s one of the metrics my employer looks at. I always hate how wasteful speculative bubbles like this are.

But maybe this isn’t the place for me since I see some good use cases and appreciate the few times someone does it right.

  • I hate how companies like Google and Microsoft are putting LLMs everywhere, making things worse for everyone, forcing everyone to deal with it
  • I find it strange that there’s an LLM in my car but it Carrie’s a decent conversation and it can’t actually do anything
  • I like this approach of building the mcp server. Keeping it read only, keeping it optional, and leaving the actual LLM open. I also like Apples approach where they seem to care about privacy, about executing on device, and about taking their time to put it in useful places
[–] Finalsolo963@lemmy.blahaj.zone 0 points 1 week ago (1 children)

Yeah, most people, myself included, don't like AI on principle, but there are valid use cases for it, and not having the capability of integrating with AI tools is going to be a dealbreaker for someone.

That said, I've heard MCP is a bit of a shitshow of a standard and is woefully inefficient.

[–] AA5B@lemmy.world 1 points 1 week ago

Yeah it is. I have no idea how you more efficiently do that task, but there’s got to be a better way