If someone break the rule, don't wait for admin, and ban them automatically with auto ban system.
Declined Make autobans
- Thread starter sklalforever
- Start date
How would you even implement this kind of system? There is a reason that admins exist, which is to verify that each and every report is valid. It's a thankless job which, does not have a way to automate with any kind of reliable accuracy.
I don't see how that could work. How would an autoban system know what toxic behavior would be?
There is an n word filter that mutes for racism. But people still get around it.
There is an n word filter that mutes for racism. But people still get around it.
I can't see how this could work. Even if you use AI it's still imperfect, there would almost certainly be false positives or false negatives no matter how sophisticated the prompt or model is. Currently it's impossible to do without ensuring a human also reviews each incident the AI detected first, which means it's not an auto ban anymore. I've already played around with this kind of idea and it definitely requires human review. There's also certain actions which an AI can't identify anyway.
Bottiger
Administrator
We already had voteban and people kept crying on reddit that admins banned them for being "good", so it was turned into votekick.