It does NSFW for me, yet I think that whatever change happened needs adjustments to sound more natural. The Dev has released no statement about the recent text changes being due to a new model. People keep saying that without proof.

I agree, I've never paid to chat either other than Gemini pro for coding. I think given some time the Dev will improve the performance of the Text gen. As he looks at feedback he slowly irons out a lot of issues.

I agree, I've tried but I have yet to find a way to make it work well. Before the change happened to Text, I was really enjoying the results for the most part.

Privately owned websites that have chatrooms are not recommended for the faint of heart. That's just the truth. The file uploader has a filter for content, your claims that people are using the website to share CSAM is a loaded claim.

"user-generated outputs within these specific generators contain explicit logs of extremely questionable fantasies" Other than galleries that have images the generations that we users create are images that we save from our own browsers. What logs are you talking about? The information on every user's browser is different. It is clear that you do not know how this website works. If you are looking at anyone's logs it would indicate that you are spying on other people's browsers somehow.

"Internal moderation on this site is entirely absent and non-functional" That is an obvious lie, there have been bans in the chatrooms and there are filters on what can be uploaded.

"A few witness testimonies can be found here" Only one of the images is visible for me the links don't work for the others.

"large population of underage users completely exposed to dangerous individuals without any active safety oversight or intervention."

The parents of any children should monitor their children's internet activities, this is a website owned by a private citizen, it has filters, moderation, warnings on pages for adult oriented generators. You have lied more than a few times in this post and made baseless claims. This website is not a business that charges the users money or a project that is sustained by donations.

"Please, remove the comments of the site; or the site itself immediately to prevent further illegal activity!" You're not providing proof of your claims. That's the problem.

A British guy, no surprise. Britain is essentially the poster child for internet censorship these days.

"Chat logs, obviously It’s helpful if you do me the honors of merely asking instead of jumping to conclusions"

So, no material, no actual content, just chat rooms, that people can choose to enter, avoid, or leave. You can find similar people in chat rooms literally throughout the whole internet. Yes, people are reported and there is moderation contrary to your false claim.

"Relative to the gen that we’re referencing, please name me at least one person that’s been actually banned for egregious behavior by a moderator"

If you searched this Lemmy forum you could find posts about that very subject. https://lemmy.world/post/49043595

"Someone keeps mass reporting me." - MeowerMisfit817

But according to you there is zero moderation and reports are therefore impossible. I'll take your words and hold you to them.

Do the due diligence of typing your links correctly, it is expected that a reader will simply copy and paste them.

"And what specifically would you like me to provide evidence for? I’ve already literally provided the formal case ID, and witness testimonies; what specifically do you want me to prove?"

If you make the claim that the website is one where people are sharing CSAM and all you have are some dodgy screenshots of chat messages, then no, you have no evidence. The chatters in these chatrooms are anonymous people and you could easily fabricate scenarios by sending these messages. Chatrooms are littered throughout the entirety of the internet and ever since they were created children were warned not to join them. The same is true with VRchat. Are you going to try to shut VRchat down as well?

Perchance has chat rooms, moderation, and report functions. You want this to be an account based website and it is not that type of website as a privately owned website that is not a marketplace there is a certain degree of freedom that it has. The images uploaded are filtered, nsfw chat bots cannot be linked. Where is the CSAM that you claim is on the website?

"Have you been paying attention to absolutely anything I’ve been saying? There are entire chat rooms literally filled with the exchange of child sexual material (CSM), child sexual abuse material (CSAM), and footage of other distasteful materials"

Is what you claim. That content would be shared with filesharing links like catbox, why would any of you click links from anonymous strangers that you don't know? You have no proof of any footage, you're making claims at this point and honestly it seems like the possible scenario is that people are posting links from other websites that have that content in them, this is a problem that is inherent to chat rooms with anonymous users.

Which, again, draws back to my primary point

No it does not, because it shows that your claim that moderation is entirely absent is a lie. The options in blocking anonymous users are limited, since there are no accounts, only ISPs can be used to block a user, and obviously they can use VPNs. But, this is the risk of any anonymous chatroom or anonymous image board. There are not accounts, people cannot be perma banned by account because the users themselves are largely anonymous. So, yes, there is moderation, but since there are no accounts and the users are anonymous which is an extremely common feature in chatrooms not much can be done to utterly ban a user. Just say that you want accounts. People have to practice caution in this respect, because you are in fact dealing with anonymous users who can post any link in the chat and they will not be found. This has been the risk of chatrooms since forever.

Yes, and I’ll absolutely stand firm in this. Hell, ask literally any present in the chatroom at any given time; or ask people who actually fled the original site for proper human moderation (You can find some of them here: https://perchance.org/safe-ai-chat)

I, again, have a physics, federal case with the NCMEC; I’ve sent the ID above. That isn’t the only evidence we have. Like I said earlier, multiple users there catalogue clips, direct screenshots, and occasionally report the materials to accredited law agencies in ‘exposed’ channels.

You have chat logs, because when I asked you about any material that Perchance was hosting on their side you did not give me any reply that they did. You're talking about anonymous people in chatrooms sending links from filesharing websites that do not care about what they archive on their servers. I am laser focused on the content on the Perchance website, I am not referring to what anonymous strangers are linking in the chatrooms. There are safer options to find friends than anonymous chatrooms. There are a multitude of websites where the users are tracked by accounts and can be perma banned and this website is not an account based website and all of you in the chatrooms have full knowledge that you are talking with anonymous strangers who at most can get ISP banned and you take that risk.

I want the site itself to at least be minimally moderated, don’t put words in my mouth.

How is it putting words in your mouth if you said something that is objectively untrue? If there is entirely no moderation as you claimed, there would not be reports or bans. So, you emotionally lied about it and lying does not make you believable especially when you are making very big claims about this website as a whole no less.

The images cannot be filtered, the upload module itself has no filter. The most it can do if it encountered anything nsfw, is put an 18+ emoji and remove the https://, which again only happened fairly recently.

I've tested the upload filter and had images blocked, so, you are in fact lying again. You do not understand how the website functions, that's obvious. There is a NSFW warning but there is also an upload filter, it will even block uploaded text content at times.

I can’t genuinely redirect someone to the specific channel names, given that it’s legally risky, but I’d be more than happy to provide the information to an actual moderator / developer to prove my own point of exchange regarding both csm, csam, and more

Considering that you have lied a number of times in this thread shows that you are not even trustworthy in general, so, you made a lot of claims, all you have are chat logs. So, I don't know what to make of that. The Dev I am sure has his own opinions about all of this.

“Internal moderation on this site is entirely absent and non-functional”

How do they ban and unban people if they have no moderation?

Sorry about that - fixed. It was a false positive in the moderation system, and I’ve tweaked things to reduce the chance of this type of issue coming up again. Also thanks for the tip on the unban bug in the other thread - that should be fixed now (let me know if not ofc).

"And thanks for the prompt length cutoff issue in reproduction in the other thread. That should be fixed tomorrow. Hopefully prompt weights too. But probably not negative prompts - I’m still trying to fix that. The “CFG burn” issues and related tendency to push stuff towards cartoon/anime won’t be solved immediately either, but I am working on it." - https://lemmy.world/post/42343571 https://lemmy.world/u/perchance?page=2&sort=New&view=Overview

Certain generators like https://perchance.org/question-this no longer follow instructions correctly, it will not abide word count, give me the number of answers that I input into the menu, or expand the answers as it always has done since I started using this website, it has always functioned even if at times it had long load time when the website undergoes updates or outages.

There has been no official statement about the recent performance of the text generation on the Lemmy site, Reddit, Perchance or otherwise. So, in that regard I won't make any assumptions. The only word on new developments that I will take seriously will be from the Dev or Vionet.

I keep reading a lot of claims about the changes, but functionally the recent changes have broken even some of the most reputable generators that rely on specific instructions regarding word length, style, and expansion of ideas.

Vionet had created some very good generators like Question This that have largely been essential to Roleplay, a big draw to this website, when good generators no longer follow inputs, that is a serious error. Instead of following the instructions, it simply engages in inexplicable behavior, instead of 16 answers it decides I should have 4 or 5 instead, ignoring the menu instructions entirely. I hope they fix it.

Furthermore, the language adaptability and word choice seem to not pick up accents or dialects as effectively as it did last week, for those who prefer a more flavorful speech in their chatbots, this lessens creativity and adopts generic language across the board. The only people that can understand this aspect of it are those that intentionally type their character descriptions in different accents. The capability to adopt and hold accents based on character descriptions and not resort to formal, overly eloquent language has been a core feature of the roleplay on this website. It has never been a CAI copy, it has unique nuances that allow the whole host of generators to function properly and it will need to be fixed for the continued viability of those generators, because in some instances as it stands presently certain generators that require clear instructions from the user are malfunctioning and simply ignoring the user's instructions.

I am not your mate. That is a queer thing to say to me. Find a lady, find a woman to be your mate. You are making the claims, so the burden of proof lies with you. My experience with the filter is that I have personally had images and lore books blocked because of their content. Users have talked about that on Reddit and probably on this Lemmy forum too. I did not make the claims that the website has no moderation and no filter on uploads. I've taken citations from the forums before to disprove your false claim about the moderation and I will keep an eye out for people who have had NSFW file uploads filtered, it's honestly not my responsibility to do research on your behalf whilst you make definite claims about how the website functions.

Chatrooms themselves are this way, no one should click links in chatrooms, why would any of you go into a room full of anonymous users that you cannot hold accountable especially if they use VPNs? Why click on unknown catbox links? This is stupid behavior. Chatrooms themselves are neutral, the people in them are not. The same is true with VRchat or 4chan; neither one of those has been shut down yet. Perchance is not at fault for simply allowing chatrooms on their private website, they have a moderation system even if you don't like it. That is no reason to lie about them and claim moderation is entirely absent. Avoid the chatrooms, enter at your own risk. The same advice was true with the first internet chatrooms created and it is still true now.

I'm not your mate or your dude. No. None of that "come on dude" stuff with me. I don't like the fact that you lied and said moderation was entirely absent on this website

Under federal laws like the STRIKE Act and strict international regulations, platforms have zero immunity when it comes to child sexual abuse material (CSAM) or severe exploitation. A platform cannot legally contract out of criminal liability by telling users to “enter at their own risk.” If illegal content is being facilitated or hosted, the platform is legally obligated to remove it, regardless of any disclaimers. If a platform’s moderation system routinely allows severe, systemic exploitation to occur without intervention, the system is functionally non-existent for those violations.

"There is no enacted federal law officially designated as the "STRIKE Act." However, labor-related legislation, such as the Empowering Striking Workers Act of 2025 (S. 2731 / H.R. 5206) and the Tax Cut for Striking Workers Act, have been introduced in Congress to provide unemployment benefits and tax relief for striking workers." - Gemini AI

Are you talking about British laws? Because we do not have that here in the US.

Courts increasingly view broken, unstaffed, or purely performative moderation systems as evidence of willful blindness or gross negligence. Private ownership does not grant a license to ignore federal crimes.

What are you even citing? Because you have yet to prove anything except show some chat logs, you can report the users. You have yet to prove that Perchance has committed or ignored any crime.

Get out of the Wild West mentality, bystanding does nothing now

You're citing an article or something. I don't care about slogans, I care about substance, You have made substantial claims and lied a few times about both the absence of moderation and filtering which I have tested myself and had both image and text blocked by the upload filter, there are posts about that function on this Lemmy forum. Your trustworthiness was shot when you lied in the first post.

You made the claim, "Internal moderation on this site is entirely absent and non-functional" and therefore you have to prove that moderation is entirely absent as you stated. If there are no accounts how can those users be perma banned? So, Perchance bans the ISP, that is all that can be done in anonymous chatrooms. So, yes there is moderation, people get reported and banned. So your claim "Internal moderation on this site is entirely absent and non-functional" is a complete lie. If your point is so important and so urgent why are you lying to us to make that point?

From what I have read it seems like it is done with IP addresses.

I asked Google Gemini for help on the code of Question This and it resulted in some better functionality in my opinion, you can try this one https://perchance.org/questionthisrevamped

This generator is very good for lore and world building purposes.

You claimed that moderation for Perchance is "entirely absent", so, it has to be one or the other. If there is moderation which there is, it means you lied in your post. Why change your words. If people can make reports, moderation is not absent.

Yup that's how it seems. For the first problem I had with VioneT's Question This generator I asked Gemini to give me new code for it and I made one that works better and follows my instructions, here's the link to that https://perchance.org/questionthisrevamped

For the dialects in speech I put this on the first line of my custom writing instructions in the ai char chat "Improvise my dialect based on character description."

So, it seems like it needs more direction now than it used to.