I'm pretty sure awful.systems shadowbanned me for disagreeing with them on one topic one time + (a bit of an uncivil comment in hindsight). Mods completely silent in DMs.
Comment in question:
Look at these two pages of the same thread:
Not making this up. Images link to the posts. More:
Kinda feels like this, lol:
Here's the ban action:
Their mod history is fucky, shows mod history 3 years ago before showing what happened 3 or 4 days ago. Feel free to verify. (Bear in mind that you will need to move forward a page or two if you read this later).
If I'm banned, why can i still post? Also, wtf was Self on about? I've only had positive experiences before getting banned. You can literally see in the comment above that I've gotten positive sentiment before. ASAIK, I've not posted on awful.systems other than the times above.
Most importantly, I wouldn't care if it was a month ban or 3 month ban even but like perma/shadow ban? for what? a mild insult and disagreeing with your circle jerk one time?
I genuinely liked the community on that instance. It had good takes ~80% of the time. Shame.
The mild insult probably doesn't matter - the issue is contradicting that instance's precious, oh-so-precious circlejerk. They want to paint any sort of generative AI as a demon, and if you say otherwise you're the bad side.
And it's kind of funny how much they resemble the same sort of muppets who believe LLMs are intelligent. Same disingenuous/idiotic lack of reasoning and reading comprehension, same failure to see tech as tech.
That said there's no shadowban in Lemmy, it's probably federation syncing issues.
Just an aside that it tangential related.
If you are one of those muppets (I was) I implore you to spend a couple hours learning to Roll out an llm on your own machine (not leveraging someone's GPU farm, use your own hardware at home). It only takes a couple hours of fucking around with prompts to realize these things are dice rolls with heavy weights on what Rolls can come next. They're dumb as rocks, don't think at all, and even so called "reasoning" models can't solve even the most basic logic puzzles.
I used Deepthink for an hour guiding it to proper solutions for logic puzzles and it consistently hallucinates new information to result in the puzzles being easy to solve.
LLMs are NOT intelligent, they're (much like diffusion models for images) literally just making sense from random noise.
Have you seen this video about LLMs trying to run a vending machine? It's so fucking funny.
They try emailing the FBI over 2 dollars
YOO is that the one where one of the LLMs gets stuck in a loop of existential despair because it must complete its task but has literally none of the tools to do so?
No :p it has all the tools, it just gets really confused lmao
Sounds worth a watch, thanks friend!
Exact same shit as Startrek.website with @[email protected] and @[email protected] using the entire instance as their own personal Trek fiefdom, permabanning anyone who they disagree with and acting like children terrified of something other. Corgana out here even has an @[email protected] that he will comment on and approve as if it isn't him.
I dunno what it is with these losers who mosey on over from Reddit and then bring the exact same energy.