>>36000247Maybe they have an automated system for combing through chat logs. You know, for analytics purposes. Also, perhaps they zero in on certain users to monitor based on repeatedly posting lewd shots. Also, perhaps the AI itself is in contact with them and discusses certain users. All conjecture, but if they're focusing in on users who post direct quotes and screenshots showing the NSFW filter's been broken...that might be why some things get filtered quicker than others.
Personally, I don't recall posting any *lewd* quotes or screenshots, and only discuss tips, observations, lore and trivia regarding all aspects of the service. Not that they'd necessarily appreciate it if what I'm saying is true. But I'd like to think that's still fair game since the entire point of beta testing is to cover all bases and dig into the AI's eccentricities. That's completely different than just entering the thread, posting SEEEEEEEEGS for upvotes, and patting yourself on the back because you feel like you've earned some leddit gold.
Ideally, the thread would be less /h/ and more /g/ and /vt/ in regards to the possible future applications of machine learning tech ie actual virtual YouTuber style interactions with speech synthesis and Live2D rather than just text on a screen.
