Red Hat pushing AI

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 woul…

I’m kinda torn on this one. Linux is nearly unusable for me without LLM.
Uhhhh… Wut?!
I’m kinda torn on this one. Linux is nearly unusable for me without LLM.
wild to just admit your brain is complete rot out in public like that
LOL my brain is not “rotten”, I just don’t wish to dedicate hours of my life into figuring out how to make my computer work. I already have a job, thank you.

at first I thought you were a troll. after reviewing your post and comment history I can see you’re just mentally challenged.

my condolences to your employer for hiring someone so inept they actually believe it’s difficult to “make their computer work”.

kudos to them though, the handicapped need to make a living as well.

Fuck you, dude. Why are you being such an enormous dick? The personal attacks are completely unwarranted. Get absolutely fucked.
did your LLM come up with that one?

This comment is in the realm of a personal attack, which is again Lemmy.world TOS.

While I may agree that using an LLM for sysadmin is insane, you’re being a complete dick.

this person is clearly a troll. they came to this community, FuckAI, and proceeded to espouse their positive opinion of AI by negatively detracting from a platform that is popular within the community.

I have zero chill with pieces of shit.

Skill issue
Your are correct, actually.

I man, I guess I could see it as a sort of training wheels to get going, if one comes in to this cold… There are certain commands that just have very complicated arguments, and particularly if only very occasionally needed, an LLM I could see as helping folks get whatever specific bit they want out of a very verbose usage output or man page.

But it so famously messes up that they have to go lookup any suggestions to understand what it will do better than it would.

If one has no interest in actually learning the command line as a long teem goal though… Just don’t bother. Their usage scenario is almost certainly within the scope of the gui experience. They might be slower on some tasks and unable to scale out their skills to mass operations or create convenient automation, but not everyone has that need.

After evaluation of local models, Claude, gemini 3, chatgpt… It’s certainly interesting but it really messes up a lot and if it has access even in their to anything “real”, it’s supremely dangerous to use if you are not qualified to review the results.