- cross-posted to:
- technology@lemmy.ml
- cross-posted to:
- technology@lemmy.ml
You must log in or register to comment.
Reddit is fucking awful at this sort of thing and I would not be at all surprised if it punishes people who didn’t harass anyone in part because it’s broken and in part because it’s being gamed by trolls to harass people because it’s broken while also openly harassing people and getting away with it because, again, it’s broken and reddit won’t bother to fix it just like the current system.
Calling it now.
Worked exactly like that with their anti evil operations “team”, which I was convinced was either a LLM or some foreign labor worker with barely any English knowledge.
Nah, it was just a collection of shitty bots and scripts.
That was their “admin” team… The anti evil team was something worse than that.
They were all pretty useless. There was one guy who openly admitted to abusing botnets to mess with reddit on reddit, then he got me banned, saying in a post that’s what he was doing.
After I got banned, it took 5 days for the appeal to process and the guy who did it didn’t get banned until a month later, only after multiple reports were filed by various people (not to mention, it was cited in the appeal).
I’m sure they turned right around and made new accounts right away too.
Reddit is horribly managed and it’s only going to keep getting worse as far as I can tell: I’m just happy there’s Lemmy now.
“The filter is powered by a Large Language Model (LLM) that’s trained on moderator actions and content removed by Reddit’s internal tools and enforcement teams,” reads an excerpt from the page.
Eww, gross. I was never a moderator, but that would annoy me.
We all knew they would use bots for moderation eventually, I didn’t think it would happen this quickly though.
While I’m skeptical AI technology is ready for this, I actually think it’s one of the better changes they’ve proposed. A truly impartial AI moderator can enforce polite discourse instead of flamewars.
Of course I don’t trust Reddit to do it right, but theoretically I dig it.
Yep. Conceptually, it actually sounds like an appropriate application of the technology, but I expect Reddit to faceplant on the actual implementation. I mean honestly, the only way they were able to make a mobile app that people enjoyed using is by basically outlawing 3rd party clients.
What are the odds that the AI model actually promotes harassment?
Ask a question on a big sub and the bot will downvote you and tell you to gargle battery acid
Battery acid story time!
Long, long ago, I worked at a Jiffy Lube. I was under the hood, some pothead noob was in the pit below. We’re supposed to call out what we’re doing, for a bunch of reasons, one of which was safety.
I called out “Checking battery, bay 3!” because I was going to pull the caps off of a non-sealed battery, when pothead noob comes right underneath to say “What?” – and then he starts screaming, because he got a drop of battery acid in his eye.
I raced down the stairs, grabbed him by his shirt, dragged him over to the eye wash station, turned the water on, and shoved his face in it. Fucking Mike.
And what’s their plan for using it to fight power-tripping mods?
Reddit planning on having even fewer moderators or paid admins, in their typical cheapskate fashion.
