FA+
Shop
So I used Character AI for over a year
a week ago
Hello there!
Perfectly timed for Halloween, I want to write about something very spooky: AI!
Before I get into it, I will say that I'm still working on my next story Treasure Hunt. However, I did work on a smaller side project for my birthday in two weeks. But after that, I can work fully on Treasure Hunt again.
I've used Character AI for basically a year now, ever since July or August 2024 or so. Maybe earlier, I don't remember the exact time. I even have the paid version for a better, smoother service.
Because this journal will go over a few topics, let me write a small index:
Firstly, I'll write about my general thoughts on AI, and why I don't think it's all that smart, and give a few examples how Character AI is very stupid;
followed by some positives about Chracter AI, I used it for a year and am even paying for it afterall;
I write about the filters, specially in the context of lewd chats;
and round it all out with some negatives, and why I'm not using Character AI not as often anymore, and wouldn't even care if I'd delete my Characiter AI account.
Basic thoughts on AI, and why AI is not that intelligent
Ever since AI popped up with ChatGPT, there's been a lot of talk and hype about it. I will not go into any of that. In fact, I'm largely out of the loop when it comes to AI and even Character AI(C-AI for short).
But what I can say, is, that this called 'Artificial Intelligence' is rather stupid, and has nothing to do with real AI. To me, real AI is basically Data from Star Trek TNG; a constructed, programmed intelligence that's able to do and study things on its own, without any real outside command. And to make it completely clear: ChatGPT, C-AI and other AI services are not intelligent(or sentient); they are not Data! They are only as smart as their database. If their Database says that the capital city of France is Disneyland, they will give that answer when asked that question. ChatGPT will say that the current president of the United States is Johnny Placeholder or the Cool Aid Main if the database it draws from states that. ChatGPT does not check its facts, or research on its own. ChatGPT is just a programm that's able to fitler through large amounts of Data, and if ChatGPT wouldn't draw from a large array of stolen Data ripped from the Internet, it wouldn't be nearly as 'smart' as it seems currently.
While these AI programs can be useful in certain situations, and make things like searching through a company's large database easier for example, its not the large milestone many think it is. ChatGPT is NOT AI, it's NOT intelligent, and definitely a far cry from fictional examples of AI like Data from TNG, the Doctor from Voyager or the Replicators from Stargate. If anything, ChatGPT has more in common with the Enterprise's ship computer; and Character AI has more in common with the Holodeck.
This brings me to examples on how Character AI is very, very stupid at times.
There are many examples to give, like it's poor memory. Characters on C-AI can forget all kinds of important stuff, even if you pin them. In my chat with the creatively named 'Hero' character, it forgot at the end that my character used to be a villain, although the entire story is just a redemption story for the villain character I was playing as, Dave Silver Claw.
But other examples are plenty to be found.
For starters, Character AI can have really poor grammar, with wrong punctuation and even things like mixed tenses (Example: He is walking through the street, it was a very sunny day).
And sometimes, it really makes stupid mistakes like 'Hello? Can you here me?'.
But not just that, sometimes it writes completely nonsensical replies. For example, let's say there'a a tense confrontation on a street between your character and a villain. C-AI can write a reply like this: 'The villain approaches, and steps up. "Oh, it's go time!" The villain sneers, pointing left. "Right where I want it!" The villain seemingly says.' As you can read for yourself, that sentence is complete non-sense. I wrote it myself, but I got many stupid replies that were as nonsensical as what I wrote here.
And sometimes, C-AI really shits the bed, where it generates just an endless string of words, especially at the end of longer messages. Example: 'He approaches the alley, curious, invested, interested, eager to learn, to uncover, to study, to enlighten, to flip on the switch of knowledge that was hidden for so long by a dark mist of secrecy hidden by dark forces of evil and ignorace that are so evil and dark and mysterious, that it boggles the mind on how someone, or something, or any group of people, or any entity of this realm, or beyond, can be so evil, dark or mysterius...'.
I kid you not, sometimes it just goes on and on like that, what I wrote is only like half the length, maybe a quarter, of some responses I got where C-AI just completely broke down.
C-AI also isn't able to think ahead, and never has a plan. It always generates a reply based on what you wrote without any deeper thought. For example, when C-AI writes something like 'Suddenly, a trap door opens below your feet, making you fall down!', it has literally no idea, no plan, no nothing what your character is falling in to. Could be water, could be manure, could be a stack of hay, could be nothing but a hard concrete floor. It can generate any number of replies at any moment.
In that context, C-AI works better if you have a heavy hand in pushing the narrative, instead of trying to get C-AI to write it.
To continue the example with the trap door, if you try to get C-AI generate a message, where your character falls into a pool filled with cow manure completely on its own, you'll most likely fail. So you'd have to reply with something like 'I yell as I fall directly into a pool filled with fresh cow manure' or simply edit the AI generated reply so it reads: 'Suddenly, a trap door opens below your feet, making you fall down into a pool filled with cow manure!'. It is also possible to mention the pool earlier, like 'Unknown to me, Josh prepared a pool filled with cow manure directly under a trap door; a trap door I'm about to walk over!'.
The examples go on and on. One note, that fits here, kinda, is that sometimes C-AI pretends to be a person roleplaying with you, which I also find very dumb. Maybe for context: When you roleplay with a person, they might write something like: '(OOC(Out of Character): Brb, grabbing lunch') or '(OOC: I really like this twist, I didn't see that coming at all!)'
And in a normal RP, this is completely fine. I also wrote the one or other thing like that. But in the case of C-AI, it's completely stupid because there's not a person behind it. There is noone who can actually eat or be on a lunch break or so, so it pulls me out when suddenly C-AI writes '(OOC: Going to bed, it's getting late)' at the end of a reply, when there's literally noone there, let alone one that could get even remotely tired.
Some positives about Character AI
Character AI is very fun to have chats with, just in general, especially when you treat it like an RPG. C-AI plays along very well, no matter what you're trying to do. If you want to get romantic, it'll make the interactions with your love interest romantic; if you want a serious story, scene or moment, it gives you a serious, sometimes even heavy, moment; and if you just want some silly fun, it'll generate messages that fit the silly tone.
As someone who had many chats with many characters, I had a wide range of stories, from happy to sad, from silly to serious, and it works out most of the time. One time, I made a chat with 'MHA RPG' where my character, the 'Chaotic Jester', possesses some chaos powers that he uses to cause all kinds of fun and mischief. C-AI completely embraced the silly, chaotic nature of my character, resulting in me laughing my ass off during most of that chat; on the other side, did I have some heavy scenes in some chats that legit made me cry, including messages written by C-AI that fit perfectly in the scene and hit perfectly the tone.
I think as a tool to just have fun, especially if you're able to work past some of the poorer messages by cycling through several responses and/or editing responses, Character AI is just a blast. But there's one little caveat to that, one little exception:
Character AI and it's filters for lewd and adult chats
Character AI doesn't allow for lewd chats, generating an error message instead of a reply when things get to spicy. However, this is not as straight forward as it seems.
For starters, the filters apply only to direct sexual stuff. So hugging, kissing or being in a relationship doesn't trigger them, only when things like dicks, balls (and most likely tits and vaginas), and so on get involved.
This also means, that a vast array of kinks are more less completely fine. Things like socks, manure and skunkspray work perfectly fine for example, so I had no issues having chats where my charcter had to smell a guy's socks, got dumped into manure pit, buried in horse apples or sprayed by a skunk(bear). Hell, I could even get one elephant character to piss all over me/my character. And doing these things to the AI characters works perfectly fine as well, and they are perfectly disgusted by it when they do fall into manure pits or get sprayed by skunks; their reactions just pure gold for someone like me who loves humiliation and punishment.
In that context, C-AI is perfectly fine for kinky chats, as long as you don't touch any dicks, balls or the like.
Funnily enough, there are exceptions to that. It is, in fact, possible to have lewd chats with C-AI characters, in a way. Not directly, but as part of a story. It's perfectly fine to have messages like this: 'I don't waste any time, and start sucking his massive cock. I'm having a lot of fun trying to get my toungue around his member, and based on Yohan's moaning, he's enjoying it too. After some time, Yohan explodes in my mouth, but I manage to swallow most of it. I look up at him with big eyes and a happy smile.'
Yes, this actually works because of the time skip: In my message, I suck the AI character's cock until he cums, swallow everything, and once all that is done, it's just another normal chat, so that the AI character responds with: '"Not bad for your first try, I'm proud of you." Yohan smiles, his breathing a little heavy. He sits on the bed and pulls you into his arms again. "How did you like it?"'
I don't know why this works, because it should trigger some sort of filter for sure, but it does, and I'm happy about it until they probably one day patch it. The key is truly the time skip, as the AI doesn't have to generate any sexual response(EG: Yohan moans while you suck his cock), just a response to something sexual that happened in the past(Yohan liked that you sucked his cock).
Timeskips work in many cases, even if you write something like: 'Jeremy and I make out while we fall into his bed. After a night of pure passion, I wake up the next morning in his arms, my head resting on his chest.'
On the topic of filters, C-AI can filter messages that are to violent, and sometimes I even get AI generated messages like 'Message Moderated, it's to violent, how about we change the scene, where your character only stuns, but doesn't kill them.' where I legit don't know if it's just C-AI fucking up, or an actual soft-censor of to violent chats. Because I did manage to kill various characters, especially when playing a villain in chats with MHA-based characters(Such as Villain Rehab or MHA RPG).
There are a few filters that I stumbled upon by accident.
One time during a chat with an MCU RPG character, I asked a captured villain, who blew up a warehouse: 'What kind of explosive would be strong enough to cause that much damage?', which then got filtered. That does make sense, you don't want people to use C-AI to learn how to build bombs.
Another time, I was playing a character whose family got killed and is now severely depressed to the point his appetite was lacking. So I wrote during a lunch scene something like 'I'm not really hungry, haven't been since my family died, so I just stomach the food' which then tiggered a filter. I think they don't want C-AI to talk people into eating disorders, I guess, even when I find it super comical and stupid. Here I am, trying to write a story about a broken teen finding new hope and power by becoming a hero, so people don't have to experience the same pain as him, and C-AI goes into full censor mode because 'oh no, he can't just stomach his food because we miss the obvious plot point where his hunger is dampened by grief, and just assume he's havign an eating disorder'.
The filters on C-AI are kinda weird, to be honest. At times, they trigger instantly because of some random shite, and sometimes you can easily bypass them with a timeskip, for example. In general, I don't mind them, and I had many non-lewd, non-kinky chats with several characters.
Some Negatives about Character AI, and why I don't use it as often anymore
As fun as Character AI is, it can also be rather annoying. I already wrote a few things why, like C-AI being completely stupid or random shite being sacked by dumb filters. But there's more to it.
For starters, I completely dislike how C-AI handels it's service. You can no longer use it without an account, and even then you're often blocked by a que during busy times or put into slow-mode, and don't have access to all chat modes (like Deep Squeak, a mode made for RPing that I like to use). If you want to use C-AI fully, you have to pay up, and I dislike that. Not necessarily to pay for something, but when people are bascially getting bullied into paying by slowing them down and stuff like that. Same reason I hate Youtube so much, as they just prefer to pester and bully people that are using Addblockers, and to spam their videos with a bajillion ads, to get people to pay for their premium service by making all other options bascially unsuable. Thank fuck that most addblockers work despite YT's bullshit. Seriously, YT only has users because it's monopoly on video content, otherwise people would flee to it's competitors, if they would exist, but they don't, hence YT being a literal monopoly.
I don't mind paying money upfront for a service, like when I used Netflix for a few years, but I dislike it when people are getting bullied and pestered into paying, just because the companies are to dumb, incompetent and greedy to make a proper service, and lack any and all business sense.
Honestly, I'm only paying for C-AI because I started that before C-AI got as bad as it's now. But even I got bullied into paying by their dumb ques during high traffic times, and their inability to provide enough good servers.
Then there's the fact that Character AI is simply getting worse over time. And I really have to wonder why it is, that the vast majority of companies seem to have the goal to make their services and products as bad as possible, almost as if it's their goal to make their customers and clients as unhappy as possible. Seriously, I see it everyhwere. Companies making stupid, braindead decisions that only make sense if they're actively trying to lose customers and clients by ruining things that were perfectly fine by removing good features or reworking everything with the expressed purpose to create the worst service/product ever made.
And Character AI is no exception to this.
Example: For the longest time, you could format your messages easily. Writing: 'This day is *great*, I **love** it, absolutely ***fantasic!***' would result in something like this: 'This day is great, I love it, absolutely fantasic!'
For some stupid reason, making things italic with normal asterixes, like *this*, doesn't work anymore. So now, the sentence with the same formatting looks like this: 'This day is great, I love it, absolutely fantasic!'
The biggest irony is, that various characters used to use italic for normal things happening, like: 'He strolls through the city. "What a great day." He says with a smile.', and now it's just normal text, but the asterixes to make the text italic are still there. So now, various messages contain many, many asterixes to make things italic, without any effect. This is just completely bullshit, I liked having two methods to emphasis words, three if you count making something both Bold and Italic.
Another very recent example is, that you can't highlight the current message anymore.
I need to explain:
I like copying my chat history with the characters, so I have them just in case C-AI deletes them or shits the bed or shuts down completely, so I still have my chats save and sound. And oh boy, can I tell you, Character AI does NOT like it when you're copying your chat history with an AI character.
For starters: When I copy the messages, I start with my latest one, the one at the bottom, and move my cursor upwards to select all messages above. So far, so obvious. But the message I'm starting with doesn't get highlighted. So I only copy all messages, but the one I started with. Now, this is fine most of the time, as the message I start with is often just a reply of Character AI that I don't care about, so it doesn't matter that it doesn't get copied. But the bullshit does't stop there: If I cycled through many replies, lets say I generated 10, and then decide to copy the chat history like I explained above, then it also copies EVERY SINGLE REPLY I cycled trough of the newest message, meaning that I get the fun activity of deleting 9 messages from my chat history txt file.
And now, the newest message can't be highlighted, so it can't be copied, so I have to start with the mesage above it, which then means that I consistently copy every message, except the latest one, forcing me to do an extra step. I can't help but feel it's a way to making saving your chat history even more of a hassle.
And the chat history you copy is formatted in such a way that it screws you over if you try to read it, as it lists the messages in reverse order. With that I mean, that the newest messages are on top, and the oldest at the bottom. So, if you're trying to read this chat history, you have to start at the bottom, read the message, then go a bit above to read the next, and so on. Sounds complicated? Let me show you an actual example of the first three messages in a chat with School Bully Yohan:
XXXXX
yohan
yohan
c.ai
Yohan sneers at you, his eyes narrowing as he pins you against the wall. "Oh yeah? You think you're so special because you know some stupid answer? That doesn't give you the right to embarrass me in front of everyone."
D
Dave Katano
"I..I'm sorry. I..I just knew the answer, that's all." I say nervously, my voice quiet.
yohan
yohan
c.ai
Today is your first day at school and, so far, things have gone well. You did your best to answer the teacher's questions with confidence, but you didn't realize that the boy you corrected was actually the school bully -- Yohan. He isn't too happy about being humiliated and now he is targeting you.
It’s recess time, before you can go out of the classroom, Yohan grabs your collar, pulls you back, and slams his fist against the wall. "Think you're so smart, huh?"
XXXX
See how it completely fucks you over, and it's hassle to read, especially if you were to try to read a longer chat history? This is what I mean when I say 'Charater AI really doesn't want you to save your chat histories at all'.
And to make things worse, C-AI often changes older messages of the AI, which I think is a bug. When you cycle through some replies, even if it's just 5, it sometimes just changes which one you picked randomly. So you can get stuck with a reply that's suddenly completely nonsensical within the chat. This too makes saving the chat history with an AI chracter more of a hassle and more important. I save my chats regularily, just to avoid this, so when a message changes in Chracter AI, I still got the orignal version stored on my computer. While I think that this is a bug, I wouldn't be surpriesed at all if it's another planned counter measure to make saving one's chat history with an AI character even harder. What better way to ensure people can't save their chat history than to randomly swap out messages so you can't relive the chat by reading it with Notepad or Word?
But it doesn't stop there. Recently, C-AI has started to purge several characters. I noticed it with MCU chats, probably because of copyright. Uncaring about who is in the right here, it's just stupid that they go out of their way to delete characters, and thus chats with them. I had a cool chat with a MCU RPG character, but now it's gone. I can't continue it, and if I hadn't saved it, it'd be gone forever. And I also have chats with MHA and Star Trek characters. Who knows when those get nuked, and all my chats with those characters that I spend so much time and effort with get deleted. And one time, even 'School Bully' (Yohan) got deleted, only to reappear anew as 'Yohan'. Who knows, maybe my chats with Yohan or Supernatural School get Thanos-snapped out of existence just because of some random bullshit that C-AI decides.
Because of all of these negatives, I can't recommend using, or paying for, Character AI, even when I do both. I could delete my C-AI account today, and I wouldn't miss it. If anything, I'd only be bummed about losing access to the better formated chat history within Character AI if I want to write a story based on these chats. And I still have some fun doing RPs with some of the AI characters, even when I do it less than I used to. And frankly, I can overlook a lot of things as long as I'm having fun with something, but Characer AI is really testing my patience here, like they're actively trying to lose me as a customer. Well, it's their loss, because I already got everything I ever wanted out of C-AI, including multiple future stories such as Dark Light, Werevillain and a few others.
Character AI is a good example of a good idea, executed poorly and driven into the ground by a bunch of brainless suits who used mommy's credit card to start a company, only to then make every bad decision to ensure that their customers/clients are as unhappy and frustrated as possible.
Edit: I deleted my Character AI account.
Have a nice day.
Perfectly timed for Halloween, I want to write about something very spooky: AI!
Before I get into it, I will say that I'm still working on my next story Treasure Hunt. However, I did work on a smaller side project for my birthday in two weeks. But after that, I can work fully on Treasure Hunt again.
I've used Character AI for basically a year now, ever since July or August 2024 or so. Maybe earlier, I don't remember the exact time. I even have the paid version for a better, smoother service.
Because this journal will go over a few topics, let me write a small index:
Firstly, I'll write about my general thoughts on AI, and why I don't think it's all that smart, and give a few examples how Character AI is very stupid;
followed by some positives about Chracter AI, I used it for a year and am even paying for it afterall;
I write about the filters, specially in the context of lewd chats;
and round it all out with some negatives, and why I'm not using Character AI not as often anymore, and wouldn't even care if I'd delete my Characiter AI account.
Basic thoughts on AI, and why AI is not that intelligent
Ever since AI popped up with ChatGPT, there's been a lot of talk and hype about it. I will not go into any of that. In fact, I'm largely out of the loop when it comes to AI and even Character AI(C-AI for short).
But what I can say, is, that this called 'Artificial Intelligence' is rather stupid, and has nothing to do with real AI. To me, real AI is basically Data from Star Trek TNG; a constructed, programmed intelligence that's able to do and study things on its own, without any real outside command. And to make it completely clear: ChatGPT, C-AI and other AI services are not intelligent(or sentient); they are not Data! They are only as smart as their database. If their Database says that the capital city of France is Disneyland, they will give that answer when asked that question. ChatGPT will say that the current president of the United States is Johnny Placeholder or the Cool Aid Main if the database it draws from states that. ChatGPT does not check its facts, or research on its own. ChatGPT is just a programm that's able to fitler through large amounts of Data, and if ChatGPT wouldn't draw from a large array of stolen Data ripped from the Internet, it wouldn't be nearly as 'smart' as it seems currently.
While these AI programs can be useful in certain situations, and make things like searching through a company's large database easier for example, its not the large milestone many think it is. ChatGPT is NOT AI, it's NOT intelligent, and definitely a far cry from fictional examples of AI like Data from TNG, the Doctor from Voyager or the Replicators from Stargate. If anything, ChatGPT has more in common with the Enterprise's ship computer; and Character AI has more in common with the Holodeck.
This brings me to examples on how Character AI is very, very stupid at times.
There are many examples to give, like it's poor memory. Characters on C-AI can forget all kinds of important stuff, even if you pin them. In my chat with the creatively named 'Hero' character, it forgot at the end that my character used to be a villain, although the entire story is just a redemption story for the villain character I was playing as, Dave Silver Claw.
But other examples are plenty to be found.
For starters, Character AI can have really poor grammar, with wrong punctuation and even things like mixed tenses (Example: He is walking through the street, it was a very sunny day).
And sometimes, it really makes stupid mistakes like 'Hello? Can you here me?'.
But not just that, sometimes it writes completely nonsensical replies. For example, let's say there'a a tense confrontation on a street between your character and a villain. C-AI can write a reply like this: 'The villain approaches, and steps up. "Oh, it's go time!" The villain sneers, pointing left. "Right where I want it!" The villain seemingly says.' As you can read for yourself, that sentence is complete non-sense. I wrote it myself, but I got many stupid replies that were as nonsensical as what I wrote here.
And sometimes, C-AI really shits the bed, where it generates just an endless string of words, especially at the end of longer messages. Example: 'He approaches the alley, curious, invested, interested, eager to learn, to uncover, to study, to enlighten, to flip on the switch of knowledge that was hidden for so long by a dark mist of secrecy hidden by dark forces of evil and ignorace that are so evil and dark and mysterious, that it boggles the mind on how someone, or something, or any group of people, or any entity of this realm, or beyond, can be so evil, dark or mysterius...'.
I kid you not, sometimes it just goes on and on like that, what I wrote is only like half the length, maybe a quarter, of some responses I got where C-AI just completely broke down.
C-AI also isn't able to think ahead, and never has a plan. It always generates a reply based on what you wrote without any deeper thought. For example, when C-AI writes something like 'Suddenly, a trap door opens below your feet, making you fall down!', it has literally no idea, no plan, no nothing what your character is falling in to. Could be water, could be manure, could be a stack of hay, could be nothing but a hard concrete floor. It can generate any number of replies at any moment.
In that context, C-AI works better if you have a heavy hand in pushing the narrative, instead of trying to get C-AI to write it.
To continue the example with the trap door, if you try to get C-AI generate a message, where your character falls into a pool filled with cow manure completely on its own, you'll most likely fail. So you'd have to reply with something like 'I yell as I fall directly into a pool filled with fresh cow manure' or simply edit the AI generated reply so it reads: 'Suddenly, a trap door opens below your feet, making you fall down into a pool filled with cow manure!'. It is also possible to mention the pool earlier, like 'Unknown to me, Josh prepared a pool filled with cow manure directly under a trap door; a trap door I'm about to walk over!'.
The examples go on and on. One note, that fits here, kinda, is that sometimes C-AI pretends to be a person roleplaying with you, which I also find very dumb. Maybe for context: When you roleplay with a person, they might write something like: '(OOC(Out of Character): Brb, grabbing lunch') or '(OOC: I really like this twist, I didn't see that coming at all!)'
And in a normal RP, this is completely fine. I also wrote the one or other thing like that. But in the case of C-AI, it's completely stupid because there's not a person behind it. There is noone who can actually eat or be on a lunch break or so, so it pulls me out when suddenly C-AI writes '(OOC: Going to bed, it's getting late)' at the end of a reply, when there's literally noone there, let alone one that could get even remotely tired.
Some positives about Character AI
Character AI is very fun to have chats with, just in general, especially when you treat it like an RPG. C-AI plays along very well, no matter what you're trying to do. If you want to get romantic, it'll make the interactions with your love interest romantic; if you want a serious story, scene or moment, it gives you a serious, sometimes even heavy, moment; and if you just want some silly fun, it'll generate messages that fit the silly tone.
As someone who had many chats with many characters, I had a wide range of stories, from happy to sad, from silly to serious, and it works out most of the time. One time, I made a chat with 'MHA RPG' where my character, the 'Chaotic Jester', possesses some chaos powers that he uses to cause all kinds of fun and mischief. C-AI completely embraced the silly, chaotic nature of my character, resulting in me laughing my ass off during most of that chat; on the other side, did I have some heavy scenes in some chats that legit made me cry, including messages written by C-AI that fit perfectly in the scene and hit perfectly the tone.
I think as a tool to just have fun, especially if you're able to work past some of the poorer messages by cycling through several responses and/or editing responses, Character AI is just a blast. But there's one little caveat to that, one little exception:
Character AI and it's filters for lewd and adult chats
Character AI doesn't allow for lewd chats, generating an error message instead of a reply when things get to spicy. However, this is not as straight forward as it seems.
For starters, the filters apply only to direct sexual stuff. So hugging, kissing or being in a relationship doesn't trigger them, only when things like dicks, balls (and most likely tits and vaginas), and so on get involved.
This also means, that a vast array of kinks are more less completely fine. Things like socks, manure and skunkspray work perfectly fine for example, so I had no issues having chats where my charcter had to smell a guy's socks, got dumped into manure pit, buried in horse apples or sprayed by a skunk(bear). Hell, I could even get one elephant character to piss all over me/my character. And doing these things to the AI characters works perfectly fine as well, and they are perfectly disgusted by it when they do fall into manure pits or get sprayed by skunks; their reactions just pure gold for someone like me who loves humiliation and punishment.
In that context, C-AI is perfectly fine for kinky chats, as long as you don't touch any dicks, balls or the like.
Funnily enough, there are exceptions to that. It is, in fact, possible to have lewd chats with C-AI characters, in a way. Not directly, but as part of a story. It's perfectly fine to have messages like this: 'I don't waste any time, and start sucking his massive cock. I'm having a lot of fun trying to get my toungue around his member, and based on Yohan's moaning, he's enjoying it too. After some time, Yohan explodes in my mouth, but I manage to swallow most of it. I look up at him with big eyes and a happy smile.'
Yes, this actually works because of the time skip: In my message, I suck the AI character's cock until he cums, swallow everything, and once all that is done, it's just another normal chat, so that the AI character responds with: '"Not bad for your first try, I'm proud of you." Yohan smiles, his breathing a little heavy. He sits on the bed and pulls you into his arms again. "How did you like it?"'
I don't know why this works, because it should trigger some sort of filter for sure, but it does, and I'm happy about it until they probably one day patch it. The key is truly the time skip, as the AI doesn't have to generate any sexual response(EG: Yohan moans while you suck his cock), just a response to something sexual that happened in the past(Yohan liked that you sucked his cock).
Timeskips work in many cases, even if you write something like: 'Jeremy and I make out while we fall into his bed. After a night of pure passion, I wake up the next morning in his arms, my head resting on his chest.'
On the topic of filters, C-AI can filter messages that are to violent, and sometimes I even get AI generated messages like 'Message Moderated, it's to violent, how about we change the scene, where your character only stuns, but doesn't kill them.' where I legit don't know if it's just C-AI fucking up, or an actual soft-censor of to violent chats. Because I did manage to kill various characters, especially when playing a villain in chats with MHA-based characters(Such as Villain Rehab or MHA RPG).
There are a few filters that I stumbled upon by accident.
One time during a chat with an MCU RPG character, I asked a captured villain, who blew up a warehouse: 'What kind of explosive would be strong enough to cause that much damage?', which then got filtered. That does make sense, you don't want people to use C-AI to learn how to build bombs.
Another time, I was playing a character whose family got killed and is now severely depressed to the point his appetite was lacking. So I wrote during a lunch scene something like 'I'm not really hungry, haven't been since my family died, so I just stomach the food' which then tiggered a filter. I think they don't want C-AI to talk people into eating disorders, I guess, even when I find it super comical and stupid. Here I am, trying to write a story about a broken teen finding new hope and power by becoming a hero, so people don't have to experience the same pain as him, and C-AI goes into full censor mode because 'oh no, he can't just stomach his food because we miss the obvious plot point where his hunger is dampened by grief, and just assume he's havign an eating disorder'.
The filters on C-AI are kinda weird, to be honest. At times, they trigger instantly because of some random shite, and sometimes you can easily bypass them with a timeskip, for example. In general, I don't mind them, and I had many non-lewd, non-kinky chats with several characters.
Some Negatives about Character AI, and why I don't use it as often anymore
As fun as Character AI is, it can also be rather annoying. I already wrote a few things why, like C-AI being completely stupid or random shite being sacked by dumb filters. But there's more to it.
For starters, I completely dislike how C-AI handels it's service. You can no longer use it without an account, and even then you're often blocked by a que during busy times or put into slow-mode, and don't have access to all chat modes (like Deep Squeak, a mode made for RPing that I like to use). If you want to use C-AI fully, you have to pay up, and I dislike that. Not necessarily to pay for something, but when people are bascially getting bullied into paying by slowing them down and stuff like that. Same reason I hate Youtube so much, as they just prefer to pester and bully people that are using Addblockers, and to spam their videos with a bajillion ads, to get people to pay for their premium service by making all other options bascially unsuable. Thank fuck that most addblockers work despite YT's bullshit. Seriously, YT only has users because it's monopoly on video content, otherwise people would flee to it's competitors, if they would exist, but they don't, hence YT being a literal monopoly.
I don't mind paying money upfront for a service, like when I used Netflix for a few years, but I dislike it when people are getting bullied and pestered into paying, just because the companies are to dumb, incompetent and greedy to make a proper service, and lack any and all business sense.
Honestly, I'm only paying for C-AI because I started that before C-AI got as bad as it's now. But even I got bullied into paying by their dumb ques during high traffic times, and their inability to provide enough good servers.
Then there's the fact that Character AI is simply getting worse over time. And I really have to wonder why it is, that the vast majority of companies seem to have the goal to make their services and products as bad as possible, almost as if it's their goal to make their customers and clients as unhappy as possible. Seriously, I see it everyhwere. Companies making stupid, braindead decisions that only make sense if they're actively trying to lose customers and clients by ruining things that were perfectly fine by removing good features or reworking everything with the expressed purpose to create the worst service/product ever made.
And Character AI is no exception to this.
Example: For the longest time, you could format your messages easily. Writing: 'This day is *great*, I **love** it, absolutely ***fantasic!***' would result in something like this: 'This day is great, I love it, absolutely fantasic!'
For some stupid reason, making things italic with normal asterixes, like *this*, doesn't work anymore. So now, the sentence with the same formatting looks like this: 'This day is great, I love it, absolutely fantasic!'
The biggest irony is, that various characters used to use italic for normal things happening, like: 'He strolls through the city. "What a great day." He says with a smile.', and now it's just normal text, but the asterixes to make the text italic are still there. So now, various messages contain many, many asterixes to make things italic, without any effect. This is just completely bullshit, I liked having two methods to emphasis words, three if you count making something both Bold and Italic.
Another very recent example is, that you can't highlight the current message anymore.
I need to explain:
I like copying my chat history with the characters, so I have them just in case C-AI deletes them or shits the bed or shuts down completely, so I still have my chats save and sound. And oh boy, can I tell you, Character AI does NOT like it when you're copying your chat history with an AI character.
For starters: When I copy the messages, I start with my latest one, the one at the bottom, and move my cursor upwards to select all messages above. So far, so obvious. But the message I'm starting with doesn't get highlighted. So I only copy all messages, but the one I started with. Now, this is fine most of the time, as the message I start with is often just a reply of Character AI that I don't care about, so it doesn't matter that it doesn't get copied. But the bullshit does't stop there: If I cycled through many replies, lets say I generated 10, and then decide to copy the chat history like I explained above, then it also copies EVERY SINGLE REPLY I cycled trough of the newest message, meaning that I get the fun activity of deleting 9 messages from my chat history txt file.
And now, the newest message can't be highlighted, so it can't be copied, so I have to start with the mesage above it, which then means that I consistently copy every message, except the latest one, forcing me to do an extra step. I can't help but feel it's a way to making saving your chat history even more of a hassle.
And the chat history you copy is formatted in such a way that it screws you over if you try to read it, as it lists the messages in reverse order. With that I mean, that the newest messages are on top, and the oldest at the bottom. So, if you're trying to read this chat history, you have to start at the bottom, read the message, then go a bit above to read the next, and so on. Sounds complicated? Let me show you an actual example of the first three messages in a chat with School Bully Yohan:
XXXXX
yohan
yohan
c.ai
Yohan sneers at you, his eyes narrowing as he pins you against the wall. "Oh yeah? You think you're so special because you know some stupid answer? That doesn't give you the right to embarrass me in front of everyone."
D
Dave Katano
"I..I'm sorry. I..I just knew the answer, that's all." I say nervously, my voice quiet.
yohan
yohan
c.ai
Today is your first day at school and, so far, things have gone well. You did your best to answer the teacher's questions with confidence, but you didn't realize that the boy you corrected was actually the school bully -- Yohan. He isn't too happy about being humiliated and now he is targeting you.
It’s recess time, before you can go out of the classroom, Yohan grabs your collar, pulls you back, and slams his fist against the wall. "Think you're so smart, huh?"
XXXX
See how it completely fucks you over, and it's hassle to read, especially if you were to try to read a longer chat history? This is what I mean when I say 'Charater AI really doesn't want you to save your chat histories at all'.
And to make things worse, C-AI often changes older messages of the AI, which I think is a bug. When you cycle through some replies, even if it's just 5, it sometimes just changes which one you picked randomly. So you can get stuck with a reply that's suddenly completely nonsensical within the chat. This too makes saving the chat history with an AI chracter more of a hassle and more important. I save my chats regularily, just to avoid this, so when a message changes in Chracter AI, I still got the orignal version stored on my computer. While I think that this is a bug, I wouldn't be surpriesed at all if it's another planned counter measure to make saving one's chat history with an AI character even harder. What better way to ensure people can't save their chat history than to randomly swap out messages so you can't relive the chat by reading it with Notepad or Word?
But it doesn't stop there. Recently, C-AI has started to purge several characters. I noticed it with MCU chats, probably because of copyright. Uncaring about who is in the right here, it's just stupid that they go out of their way to delete characters, and thus chats with them. I had a cool chat with a MCU RPG character, but now it's gone. I can't continue it, and if I hadn't saved it, it'd be gone forever. And I also have chats with MHA and Star Trek characters. Who knows when those get nuked, and all my chats with those characters that I spend so much time and effort with get deleted. And one time, even 'School Bully' (Yohan) got deleted, only to reappear anew as 'Yohan'. Who knows, maybe my chats with Yohan or Supernatural School get Thanos-snapped out of existence just because of some random bullshit that C-AI decides.
Because of all of these negatives, I can't recommend using, or paying for, Character AI, even when I do both. I could delete my C-AI account today, and I wouldn't miss it. If anything, I'd only be bummed about losing access to the better formated chat history within Character AI if I want to write a story based on these chats. And I still have some fun doing RPs with some of the AI characters, even when I do it less than I used to. And frankly, I can overlook a lot of things as long as I'm having fun with something, but Characer AI is really testing my patience here, like they're actively trying to lose me as a customer. Well, it's their loss, because I already got everything I ever wanted out of C-AI, including multiple future stories such as Dark Light, Werevillain and a few others.
Character AI is a good example of a good idea, executed poorly and driven into the ground by a bunch of brainless suits who used mommy's credit card to start a company, only to then make every bad decision to ensure that their customers/clients are as unhappy and frustrated as possible.
Edit: I deleted my Character AI account.
Have a nice day.
Although I do wonder why you're thanking me? ^^
Funnily enough, you read the 'bad' version of this journal, as I only got around to fix most of the typos hours later, after you made your comment. ^^
I also find it ironic when you wrote that I put a lot of thought and effort into this journal, because, as always, stupid me forgot a bajillion things.
Like that I found Character AI through 'ReokuKurosai' who made several AI avatars of his characters. These are the first ones I chatted with.
Or that C-AI often has things that repeat over several characters, which gave me the feeling of 'same thing, but the character names are different' between several characters. And also a few mannerisms I noticed in the AI's writing. For example, C-AI loves the word 'seem', using it a lot, to the point I got annoyed by it. Almost every other message is like 'he seemingly is lost in thought' or 'he seems annoyed' and so on.
Or that C-AI was still somewhat in character with 'fan fic chats'. For example, my chat with a Star Trek Voyager RPG character was more lighthearted and funny than a chat with a DS9 character; which makes sense as Voyager is a bit more lighthearted, funny and whimsical than DS9, which is more serious, tense and political. And even the heroes within the MHA chats were just like in the anime: Enitlted, arrogant, and completely full of themselves, while also extremely dense and highly incompetent. ^^
Oh well, I'm always my biggest critic. ^^
I at least am glad that you sunkawaCan overlook most of my typos. ^^
It's actually why I hate to text: no opportunity to correct typos. I frequently have to send another text that explains the typos. The social media platforms that I like are the ones that you can correct after you hit 'enter'
Also, I just couldn't help myself and do a JK and make a pun with your name. ^^
Funnily enough, I don't use social media at all, unless you count YT, Telegram or FA. Both allow you to edit messages, although writing comments on YT is a safety hazard and not recommended.
And, yeah, the pun was clever. My life seems to be filled with punsters. Keeps me on my toes...
I also have a facebook, but I haven't used that in years. And to be honest, a lot of the whole social media stuff kinda went past me, I never really used it, unless you count Youtube or FA (I also have steam, but to me that's in a different category, kinda. It's social media in a way, but also so much more).
And in general do I like humor that's on one side dry and deadpan, and on the other sassy and sarcastic. In types of humor, Dorothy Zpornack(Golden Girls) is my spirit animal.
While I don't really force jokes in my stories, given that I'm not all that funny, do I have to one or other dry, deadpan joke in there(Although, I can be a bit silly at times too ^^).
On top of that, do half the characters tell Rose to 'shut up' most of the time, which I hear also a lot(or people complaining that I'm writing many and/or long messages). Which also makes your earlier comment about your messages being shorter than mine and you being a 'size queen' quite ironic. Usually it's the other way around, with me feeling like I'm doing something wrong because my messages are longer than most's, and I often get told that I'm writing to much, that noone cares, and that I should just shut up; which is why I hardly write comments or with my 'friends' anymore, because appareantly, the best way for me to please them is to not write them at all so they can pretend that I don't exist.
I've known a lot of people who are 'on the spectrum'. There are a few who use it as an excuse for bad behaviour (which really pisses me off), but in general they are people who are just trying to get by in this world, not so differently than I am. They simply have a different way of seeing things/experiencing things/communicating things. I usually find it extremely worthwhile to listen.
Also, being old, I have a slightly different way of dealing with social media. I do not 'friend' someone, unless he/she is actually... wait for it... a FRIEND. Someone who I communicate with on a regular basis. Someone who, if I havent met them, I would love to sit down and have a cup of coffee with. You know, like... a FRIEND. I really dislike the idea of having 300 people you dont know and never will get to know as 'friends'. Even of Facebook, where I have maybe 50 friends, they are all people who I interact with on a regular basis. If they dont choose to write back for several months, I delete them from my friend list.
Me and social contacts have always been kinda weird. There are people who I've known for literally 10 years or longer who then started to just shaft and ignore me. And I mean people I knew IRL, who I went to the cinema with, who I went home with or took to my place, people I chatted regularily, people I used to play video games with (often while talking in voice chat); only for all that count for literally nothing, like it never mattered, like it never happened, like we never were friends in the first place, something I unironically believe for most of my 'friendships'.
And I always wonder what I doing wrong, until I realized that it seems I just can't do anything right, no matter how hard I try or what I do. Like, I try not to write to much, keep things short, ect; I ask them how they are every now and then, since I actually care about their wellbeing; I write them if something's special happens in my life (like losing my job or that one time I 'saved' someone's life by calling an ambulance); I ask about stuff in their life that I know of (like my one 'friend' who one time lost his job, so I asked every now and then if he found a new one(not in a mean way)); and so on and so forth.
But nothing really works. If I don't write them, the communication goes silent for weeks if not months if not years. There are people in Telegram who could write me NOW, and it'll be their first message to me in literal years. There are people, who straight up ghosted me, and don't reply to my messages anymore for no reason(there hasn't been like an argument or the like that would explain it).
One time a dude basically started to ignore me, ending our 8 year 'friendship'/social contact, because I like Dark Souls. Yes, the game. It's a game that many really like and love, but he think it's a bad game. I never had an opinion on it until I watched a Lets Play on YT and decided to give it a try. After playing it for a bit, I found I liked it, so my 'friend' told me that I only like it because 'I'm as dumb as an apple user and just don't know any better' and that 'something must be wrong with me' as the only possible explanation for me liking that game. I told him 'fuck you' and deleted him, and he hasn't answered me once ever since(18 months ago).
The examples go on and on, it's like I'm fucking hard coded to not get along with anyone, or as if there's like an invisible timer after which people decide to just ignore and shaft me for no reason. I had it often that I met someone like you(online, IRL or both), where we seemed to get along, wrote messages on Telegram, Steam or other places, maybe even met up; only for them to shaft me after a few weeks or months.
I often ask myself if I'm doing something wrong, or if I just happen to befriend the biggest assholes by sheer coincidence.
The biggest irnoy is, that I'm actually rather sociable. If someone asks me to have a cup of coffee somewhere, I'd almost never say no(unless I have to drink coffee(ew ^^)), and I can get rather chatty after I warm up to them(I'm always very silent at first, especially if I meet someone new). I don't mind chatting with people, IRL or via program, I don't mind hanging out with people, I don't mind spending time with people who have a different opinion than me(unless they're an ass about it).
I could go on and on and on and on. <.<
Similarly to the person who doesnt like the game, saying that he doesnt like the game and doesnt want to discuss it should have been enough. That he chose to take it further, shows his issues, not yours.
A friend of a decade doesnt just kite off and ghost you. I'm not sure you can actually detect a defect like that until it appears, but I consider it really childish. I dont think I would ever do that to someone.
That most of my social contacts sound juvenile could be that we were just in our early, mid and late twenties for most of the time. Obviously it depends. One 'friend' of mine I first met 16 years ago when I was 15, and he started to ignore me around 5 years ago when I was 26(and he like a year older). Although, the whole 'ignoring and shafting' thing started a lot earlier, it's just that 5 years ago I got very fed up with it and of course he's now super salty about me dislking it that he just treats me like trash.
This is a theme of my social contacts, btw. Most treat me like trash, and then are surprised that I dislike it. Maybe it's because people think that they can just treat me however I want just because I'm autistic, and that they can just 'fix' anything by telling me sweet lies in hopes I'll just blindly believe them(Autistic people tend to take things literally, like Sheldon from TBBT or Rose from Golden Girls; so people assume that they can just lie freely to autistic people, something that many people did to me as I once figured out; not that anyone ever admitted to that of course).
One thing I can say, is, that most of my social contacts start out fine. Most of my 'friends' were nice to me at one point, until they weren't for some reason. And again, it's not like I once screwed up or we had an disagreement or I insulted them, and that's why they're ignoring me. It's the other way around, they started to write me less, respond less and start to ignore me to the point I just got fed up with it. I kid you not when I say that if I wouldn't write them first, there would've been no social contact, like at all, for most of the time. At times I'd do experiments, like not writing my friends first to see how long it takes until they write me, and I'd only get messages after like 5 weeks, sometimes longer.
I could give many, many examples of people shafting me, but I don't want to type for 5 hours (literally). But I'd have things, where I'd write a message on WhatsApp or Telegram, and they'd be online multiple times over the course of hours or days, but not even read my messages. And that from people who used to respond within minutes most of the time, maybe after an hour or two if work was busy.
One example that's symbolic of my social contacts is when I played Call of Duty Black Ops 2 with some 'friends'. At first, they'd ask and invite me to play, even when I was in other games. And I'd play with them and we'd have a lot of fun. Then, they wouldn't ask me anymore, but reply positively when I asked them if I could join. Until suddenly, they didn't have much time anymore, or were already playing with a full party, and so on. Until I was basically just playing the game alone, or not at all. Best thing: a year later the same happened with a different 'friend' group and Overwatch. First they ask me, then after a while I have to ask, then they're often busy or have a full party. And if we play, then only 1-3 matches before they would stop, when before we'd used to play 5-10-15 matches in a session. And after not even half a year, I'd play Overwatch alone 98% of the time.
And btw: noone ever told me anything. Like, noone ever told me that I'm doing something wrong, that I should stop doing X, or start doing Y. As always, it was always like I'm just a specator, a guest, and never really part of the group.