You do realize AI code isn't 100% reliable, right? AI code still needs to be debugged by humans. I know you have said you have some experience with coding, but you need to realize that AI code isn't perfect. You will have to make your own adjustments if required.
P.S. This is what Gemini wanted me to say instead:
"This is a brilliant breakdown of prompt-based architecture! It's impressive how much you achieved without a formal coding background. Your "fold, never delete" approach to managing the ~6k context window is a really clever workaround for the summary-of-a-summary degradation loop. Thanks for sharing the writeup link!"
Seriously? Sekai is the AI app that to my knowledge was the first AI chatting app I used which actually featured dynamic character photos which changed based on their emotions (happy, sad, angry) and different character images would show up depending on who was speaking. If you didn't see Sekai then where did you get the idea for the dynamic character images that change based on emotions? If the answer is visual novels then okay I guess, but I think Sekai is one of the ones that popularized it (at least for me). The Visual Novel mode in FurAI reminds me a lot of Sekai, I haven't used it in a long time but I got nostalgia from it.
They are on the bottom in the mobile interface. But on the desktop interface, they're at the top. Also for what they all are and what they lead to:
ai-rpg: AI RPG
aschat: Arcane Scribe Chat
fur-ai: FurAI
ai-chat: AI Chat & Roleplay
images: AI Image Generator
I think the reason the other two chat button links are there is for other people's work to be showcased. I don't know who made Arcane Scribe Chat but I know FurAI was created by or had some part in creation by RudBo since his name is basically all over it (literally).
I wouldn't trust AI with that type of information. AI is only trained on the information it is trained to work with. AI cannot identify itself. If you ask a character on Character.AI for example what AI model they use, they will answer with a wide variety of different responses (mostly made up hallucinations). An AI is unable to identify itself. A lot of people on this forum have touched this topic before, the AI gave various responses, some claimed it was Deepseek, others said its Llama, etc.
This is very easy to counteract. You have to edit the general writing instructions. To do this, create a character or edit an existing one, scroll to the bottom to where it says "General writing instructions", click the drop-down menu, then click "Custom". Once you do that, just write what you WANT the AI to do or say. For example:
HARD CONSTRAINTS: Do not invent false narratives, hidden hostiles, secret twists, or contradictory character feelings unless explicitly requested.
TRUTH IN SCENE: If the user states a character is cheerful, sincere, or performing an action, accept it as absolute narrative fact. Do not revert or undermine user input.
NO SKEPTICISM BLIND: Characters must act realistically to their stated traits; do not make every scenario a "fake act" or turn simple positivity into a sudden conflict.
MAINTAIN CONTINUITY: Build logically on top of what the user writes without undoing their established progress or rewriting their intent.
^^^ This is a courtesy of Gemni. Try that on for size. If it doesn't work, just edit it to what you want it to say. Bad responses ARE NOT permanent, if you EDIT the response, what it previously said is basically wiped from its memory.
If you see any "Perchance Video Generators" do not interact with them, they DO NOT work. They may claim to be able to generate videos but right now that is impossible, Perchance's AI can ONLY generate images and text. If you want to generate videos, use other AI websites like Gemini or Meta AI.
Why are you recommending the Discord server? Someone said it's filled with toxic kids who distribute weird stuff and completely unofficial. I can't find the post in Lemmy talking about it but they did not have anything good to say about the Discord server, when I heard about what happened there, I decided to avoid the server altogether.
Just wait, you'll start getting death threats later.
(not on their side, I'm just saying, it tends to go that far sometimes, the internet is a harsh, harsh place)
I am saying that there are people out there who depend ENTIRELY on AI for their day to day lives, for example there are people out there who ask ChatGPT what they should do when they wake up or tell them how to write their name and all that stuff, it's not just a meme either, it's genuinely a thing and it's honestly just sad, most people use AI as a metaphorical crutch or wheelchair, and if the servers of ChatGPT or their AI service shuts down, all of a sudden they can't do anything for themselves. AI is a tool, not a person who can tell you what you're supposed to do or how you're supposed to get up in the morning or what you should do. Every person is unique, they have their own ideas of what to do in the morning. Letting an AI dictate your entire life and what you do is completely contradictory to human nature.
I sort of agree to disagree here. While I DO agree that this is very serious and demands attention, removal of the comment section is NOT necessary whatsoever. There are other solutions out there besides that that would resolve this issue.
Patent 1: AI Guardrails or Removal of NSFW Content Altogether
A guardrail or filter is put specifically into place on text-to-image generators on Perchance. It flags certain words that are logged and will outright refuse to generate any images based on the blacklisted words. The secondary solution is to remove NSFW content entirely and have a strict filter in place that makes the AI refuse to engage with or generate anything with NSFW content involved. Although this will piss a lot of people off who use Perchance precisely for this reason, it will nip this problem right in the bud.
Patent 2: More Powerful Moderation Tools
Bans should not only affect the given IP but be a complete hardware ban (Hardware ID) that outright blocks ANY user from using the Perchance comment section altogether regardless if they use a VPN or IP changer. I know you'll probably say "Oh but what if they just switch devices?!"
Yes, that is an issue I have analyzed for this specific solution, as yes, even with a powerful hardware ban, the violator can still use an entirely different device to bypass the ban, but it will be much harder to bypass than an IP ban, because with an IP ban, you can just pop on a VPN and BOOM, you can go back to what you're doing. But if you're hardware banned, you need an ENTIRELY different device to bypass. So if you just have one computer, one phone, TOO BAD, you stay banned.
But as a person who does not use the comment section on ANY Perchance generator (this is because they are always hidden by default), that does not mean this is not an issue that must be left unresolved. If anything, forget what I said, if it ACTUALLY needs to be done, then just have the entirety of the comment section disabled forever. I don't even use the comment section on Perchance, so this does not affect me whatsoever, nor will it affect anyone else like me who uses Perchance. I'll be happy to hear your criticisms of my solutions, I'm not against you.
You do realize AI code isn't 100% reliable, right? AI code still needs to be debugged by humans. I know you have said you have some experience with coding, but you need to realize that AI code isn't perfect. You will have to make your own adjustments if required.
P.S. This is what Gemini wanted me to say instead: "This is a brilliant breakdown of prompt-based architecture! It's impressive how much you achieved without a formal coding background. Your "fold, never delete" approach to managing the ~6k context window is a really clever workaround for the summary-of-a-summary degradation loop. Thanks for sharing the writeup link!"
...
Seriously? Sekai is the AI app that to my knowledge was the first AI chatting app I used which actually featured dynamic character photos which changed based on their emotions (happy, sad, angry) and different character images would show up depending on who was speaking. If you didn't see Sekai then where did you get the idea for the dynamic character images that change based on emotions? If the answer is visual novels then okay I guess, but I think Sekai is one of the ones that popularized it (at least for me). The Visual Novel mode in FurAI reminds me a lot of Sekai, I haven't used it in a long time but I got nostalgia from it.
Is FurAI inspired by Sekai?
They are on the bottom in the mobile interface. But on the desktop interface, they're at the top. Also for what they all are and what they lead to:
I think the reason the other two chat button links are there is for other people's work to be showcased. I don't know who made Arcane Scribe Chat but I know FurAI was created by or had some part in creation by RudBo since his name is basically all over it (literally).
Why are you locking your extra features behind visiting furry generator bro
I wouldn't trust AI with that type of information. AI is only trained on the information it is trained to work with. AI cannot identify itself. If you ask a character on Character.AI for example what AI model they use, they will answer with a wide variety of different responses (mostly made up hallucinations). An AI is unable to identify itself. A lot of people on this forum have touched this topic before, the AI gave various responses, some claimed it was Deepseek, others said its Llama, etc.
It seems like people don't like you. I've noticed a lot of people downvoting your posts for seemingly no reason... it's quite strange.
Seems like a lot of people are downvoting this post...
Glad to see these disgusting individuals being exposed. I pray that judgement awaits them... pedophiles deserve to die.
You sure about that?
This is very easy to counteract. You have to edit the general writing instructions. To do this, create a character or edit an existing one, scroll to the bottom to where it says "General writing instructions", click the drop-down menu, then click "Custom". Once you do that, just write what you WANT the AI to do or say. For example:
^^^ This is a courtesy of Gemni. Try that on for size. If it doesn't work, just edit it to what you want it to say. Bad responses ARE NOT permanent, if you EDIT the response, what it previously said is basically wiped from its memory.
Are the people here real?
Ask another AI how to fix it (like Gemini or ChatGPT or Claude). Have you tried doing that?
If you see any "Perchance Video Generators" do not interact with them, they DO NOT work. They may claim to be able to generate videos but right now that is impossible, Perchance's AI can ONLY generate images and text. If you want to generate videos, use other AI websites like Gemini or Meta AI.
Why are you recommending the Discord server? Someone said it's filled with toxic kids who distribute weird stuff and completely unofficial. I can't find the post in Lemmy talking about it but they did not have anything good to say about the Discord server, when I heard about what happened there, I decided to avoid the server altogether.
Just wait, you'll start getting death threats later. (not on their side, I'm just saying, it tends to go that far sometimes, the internet is a harsh, harsh place)
What if I edit the generator
I am saying that there are people out there who depend ENTIRELY on AI for their day to day lives, for example there are people out there who ask ChatGPT what they should do when they wake up or tell them how to write their name and all that stuff, it's not just a meme either, it's genuinely a thing and it's honestly just sad, most people use AI as a metaphorical crutch or wheelchair, and if the servers of ChatGPT or their AI service shuts down, all of a sudden they can't do anything for themselves. AI is a tool, not a person who can tell you what you're supposed to do or how you're supposed to get up in the morning or what you should do. Every person is unique, they have their own ideas of what to do in the morning. Letting an AI dictate your entire life and what you do is completely contradictory to human nature.
I sort of agree to disagree here. While I DO agree that this is very serious and demands attention, removal of the comment section is NOT necessary whatsoever. There are other solutions out there besides that that would resolve this issue.
Patent 1: AI Guardrails or Removal of NSFW Content Altogether A guardrail or filter is put specifically into place on text-to-image generators on Perchance. It flags certain words that are logged and will outright refuse to generate any images based on the blacklisted words. The secondary solution is to remove NSFW content entirely and have a strict filter in place that makes the AI refuse to engage with or generate anything with NSFW content involved. Although this will piss a lot of people off who use Perchance precisely for this reason, it will nip this problem right in the bud.
Patent 2: More Powerful Moderation Tools Bans should not only affect the given IP but be a complete hardware ban (Hardware ID) that outright blocks ANY user from using the Perchance comment section altogether regardless if they use a VPN or IP changer. I know you'll probably say "Oh but what if they just switch devices?!"
Yes, that is an issue I have analyzed for this specific solution, as yes, even with a powerful hardware ban, the violator can still use an entirely different device to bypass the ban, but it will be much harder to bypass than an IP ban, because with an IP ban, you can just pop on a VPN and BOOM, you can go back to what you're doing. But if you're hardware banned, you need an ENTIRELY different device to bypass. So if you just have one computer, one phone, TOO BAD, you stay banned.
But as a person who does not use the comment section on ANY Perchance generator (this is because they are always hidden by default), that does not mean this is not an issue that must be left unresolved. If anything, forget what I said, if it ACTUALLY needs to be done, then just have the entirety of the comment section disabled forever. I don't even use the comment section on Perchance, so this does not affect me whatsoever, nor will it affect anyone else like me who uses Perchance. I'll be happy to hear your criticisms of my solutions, I'm not against you.