Over the last few years we have been struggling to keep bots and spam out of our sites, so we decided to explore whether we could use AI to find out when those are happening, to let the moderators know as soon as possible so they can take action.
We have been testing and iterating this idea over almost a year now both in simpleplanes.com and here, and the tests have been really successful, resulting in @AutoModerator.
We have decided to expand on this by having it check all comments, bios and posts, both public and unlisted, so we can have a heads-up on potential trouble right as it happens. This has lead us to find some issues in our site we didn't even know about.
Hopefully we will be able to have a better experience free of scam bots now, and we will be able to combat bullying and nsfw content more effectively.
So, it essentially 'flags' things like caustic language, and inappropriate content, "before anyone reports it (said Pedro)? I mean, if it is, that's great. I just wanted to allude to the fact that I hope it doesn't flag unnecessary or safe things. Thank you for clarifying @SupremeDorian.