Broken AI Schizo Filters

I thought things were doing badly before, but Civit staff never seizes to amaze, do they?

So not only they have downgraded working training engine to half-functional AI Toolkit, not only there is still working game-like rewarding program with competing rankings for flagging content, but now AI is given first word priority to blindly decide which character is minor or not.

Apparently clanker reads first results from google-search by using the name of the model on its page, reads those results wrong (what a surprise) and preemptively censors model without human revision. This is just great. And this is also an elaborate decision, while moderation and support is still being "improved", and we, creators, have yet to get "more control over ratings".

But hey, at least now there are sparkly pretty badges and whole hall of fame, haha.
Amazing.