from the mod log:
The output from the LLM was provided as proof that someone needed to be banned.
I didn’t want to do this but my hand was forced by their dissembling, minimization and bullshit. Note how the output includes a nice big logo with a slogan and ascii graphics - this was not a one-off experiment, this is an app they were proud to call their own.
The thing is, nothing is automated. From what I can see it only flags potential posts for a mod/admin to verify and act on.
It’s not like this is a bot that scans everyone’s posts and automatically bans them, it’s an entirely manual process to download a comment history and AI ctrl-f for stuff.
There has been some deceptive information spread around what this tool seems to actually do. Which is especially concerning after the recent (proven) misinformation about Nazis infiltrating our instances.