Bot/LLM posts/threads on STH

Notice: Page may contain affiliate links for which we may earn a small commission through services like Amazon Affiliates or Skimlinks.
Wasmachineman_NL

Wasmachineman_NL

Wittgenstein the Supercomputer FTW!
Aug 7, 2019
2,361
877
113
What does STH's staff team think of them? I personally feel like they're cringe and unwanted as it does nothing but pollute the forum.

Especially people who verbatim post LLM output as threads and do zero effort in making their posts 'human' or do not disclose the fact they used an LLM to begin with.
 
  • Like
Reactions: nabsltd and CyklonDX
M

MBastian

Active Member
Jul 17, 2016
337
99
28
Germany
I do not mind the wall of text some users paste from an AI to explain their issue ... because tl/dr
What I do mind are the attempt to promote some more or less shady software, like the thread necros in the last weeks. One tried to push some vibe coded firewall software with an high version number and multiple releases on his Github page, despite the first commit being only a few month old.
Servethehome has a high google ranking and with nearly every tech related request I see Codex searching this forum.
 
Last edited:
Patrick

Patrick

Administrator
Staff member
Dec 21, 2010
12,638
6,056
113
Right now (and this changes) my thought is that we usually spam clean the ones that are just trying to promote links. Also, we filter a ton pre-registration. If they are just repeat comments, I am not a fan.

I do not mind the translation ones, because English is not everyone's first language.

Folks do use AI to set things up, and that is super useful. I mean, even the hyper-scalers have engineers that are using AI to do real engineering work, and then they get long summaries that they would never write themselves. I am not opposed to this per se because this is just the output of crazy useful tools.

On the other hand, I took a photo of the side of the APC rack at DTW and some folks thought I made the image using AI. APC might have, but it was a photo of the side of the rack that was at the show. I worry about folks not being able to tell the difference and being overzealous on the spam cleaning. The delta between human and AI is continually compressing. APC PowerForge with Dell and NVIDIA at DTW 2026 - ServeTheHome
 
  • Like
Reactions: nexox, Marsh and marcoi
nabsltd

nabsltd

Well-Known Member
Jan 26, 2022
815
619
93
Patrick said:
Folks do use AI to set things up, and that is super useful.
Click to expand...
Before the AI invasion, first posts from users were much more likely to be replies to existing threads. Now, it seems like every first post from a new user is a very specifically detailed problem that maybe can't be directly solved by normal Internet searches, but at least some headway could have been made if the user had done a search and tried some of the suggestions.

This makes me believe that most are just attempts to get responses so the poster can see how good their AI chatbot is.
 
  • Like
Reactions: Ralph_IT
You must log in or register to reply here.
Top Bottom