Midnights in Valhalla | VRChat's Nightly Bartender
@MidinVal
Owner of the Midnight Bar in VRC, Canadian🍁, Twitch Partner, VCAN Twitch team lead | Art: #ElaisseArt | 3D: @_Mari_Art | 2D: @7chewkaesa | PFP: Aikun2012
Hey Regulars, Spooktober is here and I'm going to continue like I did last year. The week starts on the 5th as I took a break on Thursday ruling out the past week :)
October 5th, 8th, 12th, 15th, 19th, 22nd, 26th, and 29th will be streaming days and check my discord for events!
One day haha, I remember the time when I posted with some friends Zen's world with all the Gecko's and she saw it and retweeted it. It was a good time, I still know some friends of hers in VRC too.
Would love to meet her, I just don't know if she'll ever see this haha
@ImDeadArc@Velouria_Yoshev@VRChat@Steam After doing some research after some users mentioned is that they do not have to disclose it, since the out of the box experience does not have AI on it. The community making it using AI tools doesn't fall under that policy. https://t.co/5hy1kclOrg
@MidinVal VRChat won't have to disclose AI use on Steam because Steams policy is specifically for developer use of AI in games. This doesn't seem to extend to user generated content for use with in a game.
@debugempty Well for general games that use it, it seems they do to some degree https://t.co/enP3sNkOh3
I have seen games before (the general use case of AI) where they didn't disclose it and got removed. It does happens more often than you think.
You know, in some way, they're not wrong. With AI being used on VRC, they'll have to disclose this on their Steam page or else they can get in serious trouble.
@WildWanderer_vr Oooh that's true, it's a gray area because since out of the box it doesn't appear to have any use of it from themselves, and since the community is the one producing the content, it wouldn't really be considered part of the policy.
Still it'd be nice to disclose something.
🎈I've just released my FREE Static Balloon prefab!🎈
It's a balloon that attracts Physbones when it gets close!
Just like a real balloon it will pull at your hair! Or any other Physbone you bring it near!
It also makes some suitably annoying balloon squeaks.
#VRChat
@ShadowriverVR@AikatsuyuriVT One thing I could've pointed out, if they add a recommended/not like steam does with a comment box as to WHY you rated it + or -, then you can find out if it's just people not liking the person then you can avoid it. Also a desktop option to comment would be nice if that happened
I feel like there will always be people willing to dedicate their time to identifying and labeling AI worlds, regardless of whether VRChat implements an official system or not. Most people can already recognize AI-generated content the moment they enter a world, so a lot of this is going to come down to trusting whatever downvote system gets introduced.
Honestly, I think we already knew this was the direction VRChat was going to take. During this year's Spookality event, there were multiple AI-generated worlds featured, which felt like a clear attempt to normalize AI-generated content, avoid scaring away users, and encourage broader adoption of AI tools among future world uploaders. I hesitate to call some of these people "world creators," because in many cases they didn't actually create the assets themselves. They simply generated content, assembled it, and uploaded it.
The unfortunate part is that my immediate reaction to a heavily downvoted world would be, "Is this an AI world, or is it just a world that people don't like?" I don't want to waste my time checking something out if it's just another soulless AI-generated world whose only purpose is to farm donations.
I also worry that downvotes will end up being used for more than just AI content. Worlds that fail to meet expectations, creators with vocal haters, and anything that becomes controversial could all get caught in the crossfire. If there were a built-in AI disclosure system from the start, it would at least provide some context. Then I'd know whether a world was being downvoted because it was genuinely low quality, because it relied heavily on AI, or simply because people disliked it for unrelated reasons.
Another question is how the voting system will actually work. Can users downvote a world immediately, or will they only be able to do so after they've visited it? That distinction feels important, since unrestricted voting could easily be abused without anyone ever engaging with the content they're judging.
I think the community will naturally police this over time. If creators are perceived as using AI deceptively or maliciously while presenting their work as entirely human-made, people will likely call them out publicly. Even without an official labeling system, those worlds and creators could end up effectively blacklisted through community discussions, recommendations, and word-of-mouth.
This also raises an interesting question about Steam's AI disclosure policies and how they might apply to VRChat. The platform has been in Early Access for nearly a decade, and AI-driven features are becoming increasingly integrated into the user experience. At some point, does that mean VRChat itself would need to disclose AI usage on its Steam page? If Steam eventually categorizes it as an AI-enabled product, I'd be curious to see how that affects discoverability, public perception, and long-term growth. It may not have a major impact, but it's definitely something worth considering as the platform continues moving further in that direction.
@KindaKuro If at all can you clarify on this? What I'm saying is that if AI is being accepted to stay in some regard, then in turn, VRChat should disclose on their steam page that AI is being allowed to be on the platform as AI-assisted or at least therof at the bottom of the description.
@Grumpstargamer It's crazy how many people were concerned about the future of it. I know AI is gonna stay but I am pro disclosure of having what tools were used for it. I do not like lazy people doing everything AI and no little effort in doing checks and fixing what the AI didn't do.
So instead of what Steam offers in disclosure of what AI is used, we get likes and dislikes instead. Which, I feel could be worse?
I probably missed a few things from this post but I’m still on the fence with it.
A few weeks ago, we asked the community to submit feedback on how they felt about AI in VRChat.
Here's our takeaways and next steps:
https://t.co/egpByRfeRD
@AikatsuyuriVT@Velouria_Yoshev@ShadowriverVR mentioned it being anonymous likes and dislikes too so you might not even see them but I still think that's pointless. Because then they'll notice because it has AI in it, they're just going to mass dislike it.