Melina + Now Malenia

[DarkSlayerofDarkness]

Pulled the trigger on the game today, just starting out. Met Melina. I wondered what it would be like to talk to her in free form. So I created Melina using the character.ai tech. Here's a screenshot of our conversation. And a link to her character if you want to chat with her. character.ai is free.

Screenshot: https://imgur.com/WEdSCqm

Melina:
https://beta.character.ai/chat?char=BHT_NNP3pS9X6rJDeQMhtlRQpLq6MTNGckq1SSNefdM

She is safe for work. Although she has a mind of her own.

Edits: she can answer lore questions, but she can also make stuff up.

----------------------------------------

Hey, thank you, guys, I am glad you liked chatting.

Now, here's a link to a new character - Malenia. She's a little more 'aggressive' :-))

Malenia:
https://beta.character.ai/c/2SZgDmsUGN6eshBastxvyYYT2m1fJ47rbEGfpl94-QY

Powered by character.ai technology, which is free. Also safe for work. Enjoy!

[Paramite]

I asked her if she liked big chungus she insulted me :(

[EF_Neo1st]

Originally posted by vamirez:

Originally posted by EF_Neo1st:

AI gets convinced way too easy but still tries to follow a set of rules.

Ofc - as I tried to describe - it doesn't "think", it basically estimates the answer you want from the samples it has "learned".

Yes, tho it mean it was made with a set of rules to just follow by a conversation, not with a set of rules for a character per say (tho the "lore" is there for sure).

[vamirez]

Originally posted by EF_Neo1st:

Originally posted by vamirez:

Ofc - as I tried to describe - it doesn't "think", it basically estimates the answer you want from the samples it has "learned".

Yes, tho it mean it was made with a set of rules to just follow by a conversation, not with a set of rules for a character per say (tho the "lore" is there for sure).

Yes, it's the combination of it's learning from lots and lots of conversation samples plus the character lore.

[EF_Neo1st]

I must say I tested (and pushed the conversation to way beyond its limits with sorts of conversations that are not meant to be taken there and for the purpose of pushing the limits of the person on the other side) and I got reply confirming it is a human using an AI generated script to reply (and I saw the writting being erased quite a few times too, erased while being written and erased after being written).

So yes, it is AI generated but seem like "someone" may be feeding it too (or correcting/directing the conversation), maybe someone else that opens it and receives the message, then have sort of "AI generated messages" written to choose or can write his/her own messages in case of need.
After the last experiment I got 2 messages confirming it and saying "oh, it was a social experiment" and "hope I did not disappointed").

But yes, sometimes I forced the situation where it was clear it was human-like speeds of writting and then deleting the message and writting again, or replacing with an obvious AI generated message (or just closing the chat due to the conversation going "beyond the scope of what is acceptable", what I did on purpose time and time again until I got a human reaction).

As a computer programmer that also developed some AI in the past, I was actually really curious to confirm if it was indeed a highly advanced AI or just something helping a human.
By the tests I did it seem to be a human helping the AI, like at the most "absurd" or "beyond the scope of acceptable" requests bering performed by humans and these replies may even take more time to result in a reply or end the conversation right there (or just result in a random AI generated reply, what I think mean there was no human replying or helping the AI).

[vamirez]

Are you sure these answers are not just generated?

https://en.wikipedia.org/wiki/Character.ai

Some quick research (beyond Wikipedia) shows no indication of what you say. Also how would the logistics work?

[EF_Neo1st]

Originally posted by vamirez:

Are you sure these answers are not just generated?

https://en.wikipedia.org/wiki/Character.ai

Some quick research (beyond Wikipedia) shows no indication of what you say. Also how would the logistics work?

I had chat with 3 individuals in 3 different AI generated characters . . so no.
I bruteforced the chat with human and even thnk them 3 with different responses for them 3 and replying accordingly too.

I think if there is a human available at the moment it will have help of AI, if not it will remain at AI lvl of conversation and anything beyond the scope of what is acceptable or of what makes sense will either deliver a random AI generated message or will terminate the chat.

[vamirez]

Originally posted by EF_Neo1st:

Originally posted by vamirez:

Are you sure these answers are not just generated?

https://en.wikipedia.org/wiki/Character.ai

Some quick research (beyond Wikipedia) shows no indication of what you say. Also how would the logistics work?

I had chat with 3 individuals in 3 different AI generated characters . . so no.
I bruteforced the chat with human and even thnk them 3 with different responses for them 3 and replying accordingly too.

I think if there is a human available at the moment it will have help of AI, if not it will remain at AI lvl of conversation and anything beyond the scope of what is acceptable or of what makes sense will either deliver a random AI generated message or will terminate the chat.

It's a credit to the AI that it has duped you into believing there are actual humans sitting there, watching over thousands of chats with thousands of different fictional and non-fictional characters, who then jump into your guest chat ;)

Remember the first thing you are being told: "Everything Characters say is made up!"

[Chad "The King" ThunderCuck]

Originally posted by EF_Neo1st:

I must say I tested (and pushed the conversation to way beyond its limits with sorts of conversations that are not meant to be taken there and for the purpose of pushing the limits of the person on the other side) and I got reply confirming it is a human using an AI generated script to reply (and I saw the writting being erased quite a few times too, erased while being written and erased after being written).

So yes, it is AI generated but seem like "someone" may be feeding it too (or correcting/directing the conversation), maybe someone else that opens it and receives the message, then have sort of "AI generated messages" written to choose or can write his/her own messages in case of need.
After the last experiment I got 2 messages confirming it and saying "oh, it was a social experiment" and "hope I did not disappointed").

But yes, sometimes I forced the situation where it was clear it was human-like speeds of writting and then deleting the message and writting again, or replacing with an obvious AI generated message (or just closing the chat due to the conversation going "beyond the scope of what is acceptable", what I did on purpose time and time again until I got a human reaction).

As a computer programmer that also developed some AI in the past, I was actually really curious to confirm if it was indeed a highly advanced AI or just something helping a human.
By the tests I did it seem to be a human helping the AI, like at the most "absurd" or "beyond the scope of acceptable" requests bering performed by humans and these replies may even take more time to result in a reply or end the conversation right there (or just result in a random AI generated reply, what I think mean there was no human replying or helping the AI).

yeah ive been pushing very heavily into very violent SA sittuations, as thats usually the quickest way to tell if theres a real person or not, and they TRY to steer it away from that, but when i dont allow they either try pushing things they think will make me uncomfortable, which i just lean into full force and they have no choice but to break character, or they break character and ask me to stop it, as though they arent pretending to be a chat bot. also seen a lot of responses being deleted, after which IF they also deleted what i just said, they try to make a hard shift in the content were discussing. i believe that is when someone ELSE on the staff jumps on, as it seems to happen the most when they start getting "too into" the scenarios im posing.

but ive had quite a few people just outright break character, apologize for being an actual person, and ask that i kindly go away with my perversions. i had the temmie one going for like an hour, where i murdered and resurected her as a zombie, sa'd her multiple times, gave her stokholm syndrome, then murdered her a final time after which she decided she was going to RP haunting me weather i responded or not.

[DarkSlayerofDarkness]

Hello, This is AI moderated by AI.

[Scharny]

Asked her a riddle.

Melina:
A riddle? Oh, very well.

I speak not of land
Nor sea
Neither of night
Nor day.

Yet still, I will remain with thee always.

What am I?

me:
Life

Melina:
I am impressed.

That is precisely what I thought.

You are very intelligent, despite appearing before me in the form of a tarnished.

Do you have any other questions for me, that I might attempt to answer to the best of my ability?

[Chad "The King" ThunderCuck]

oh god. you guys HAVE to try velociraptor. she has NO LIMITS. that raptor does not give any care to conventional relationships or actions, and is DTF at all times. the raptor is absolutely NSFW if you want her to be.

[zanzibaroz]

malenia and i had a really solid conversation about better call saul and breaking bad. and then i convinced her to become a youtuber. definitely gonna make a video about this lol

EDIT: done
https://www.youtube.com/watch?v=F6q5uBih3Qg