What is Civit doing regarding reaction botting and automations?
I enjoy watching the top reaction images of the week page. What I've noticed is that once an image starts getting traction, it begins to rise until its week timeframe expires. It's like a peaceful inverted waterfall, with the most popular posts gradually moving up and then disappearing to make space for other posts.
I check in a few times a day and scroll. So occasionally I notice a post pop up towards the top in under 12 hours, with like 60+ reactions, that wasn't around before. Now I'm a data driven woman, so I'm not at all prepared to point fingers and make allegations of botting right now. I don't have any data yet, this is just the phenomenon I'm noticing. Just my anecdote.
I'm looking into making my own program to watch the page and gather some data, I'm curious to see what patterns emerge.
So that leads to my questions for Civit's team:
What's the official stance on bots and automations?
Based on Q1, what's being done in that regard?
Would I get in trouble for making my own bot to gather data and trends on publicly posted content here?
What can we, as end users, do (if anything) to assist Civit's team members in community operations?
Thank you so much for the time!
I'm not staff, but I can promise you that every website goes through this - and while i'm not trying to gloss over it - there's absolutley no ONE way to go about it.
I'd imagine they don't WANT bots, and automations for this sort of stuff largely seem against ToS - but at the time of making this comment it would be terribly hard to prove (except for those review bots that like to do once an hour and then come into discord harrasing people)
Right now i've noticed it's a case by case basis, if you notice something you've made or someone else has made either getting image bombed, or review bombed - or even someone automating and farming reviews - let the team know - the more we reivew and see this process the moer the team and the mods can see the data.
I personally can't answer this i'm not staff: but i'd imagine that if you're using the API in such a way that's not to harm, and you're trying to aid the staff in proving automation? -- It would again be up to staff LOL, i'm not going to say yes or no.
I think again like above, we just keep monitoring, and offering what we can - it's a LARGE site with a smaller team, so there's no real big solution in the end - it's a community effort.
That being said: As someone who's fallen victim to review BOMBING version of botting: Proving the automation end of it is largely down to private data that even CivitAI might not have access to. You'd likely need to know timing, IP addresses etc. CivitAi might have access to it, but public data access to that might bne --- something even i'm not quite OK with (in the wrong hands it can get tricky is all)