Jenatha
voice can be a big wow at first but it only works long term if the app has good emotional pacing under the hood, else it wears off fast.
00:07
Telnia
Please contact the moderators of this subreddit if you have any questions or concerns.
01:44
Agielac
A lot of the community avoids forgemind for a reason.
01:56
Cri
Still refuses to function and i been hitting the generate button for the past hour already
03:23
Hengey
>If all else fails, would she like a puppy?
03:33
Mardee
Does that mean that if I create a bot based on a character from a fandom with the tag dead dove and indicate the nature of this bot in the bio (because I intended it to be a psychopathic serial killer) etc., that I am breaking the rules?
03:41
Hillore
So all of our characters are in the photorealistic realm of design.
03:45
Pazmatdy
He is quadriplegic after a traffic accident a few years ago.
03:55
Beneshe
So it was fun.
04:32
Ahber
Combined with the fact that schools don't teach how the government ACTUALLY works.
05:08
Brilmeni
Dreamgf is decent for generating standalone images, but the character consistency is a mess, hair color and facial features change between photos, and it completely fails the memory test, with no cross-session recall at all.
05:31
Dezzemine
I love GLM too !!
07:11
Egmae
The app is going downhill :'(
07:30
Fottele
Wow, then that's genius, since we could use it while waiting on SillyTavern to generate a response if it's being slow instead of tabbing over to Youtube while waiting.
08:49
Seudina
That's right.
09:36
Nrishia
I’ve found Kalon works a bit better in that regard for me, customization great since I can shape the character more directly, and chat + generate images.
10:12
Yauren
You clearly don't know what you're talking about if you think a literature review is just random news stories and a "limitations" section is unscientific.
10:52
Detmancy
I get lazy with my roleplay when I’m tired and I mean *“on the verge of falling asleep”* tired.
11:20
Anondra
You will end up paying around 1 to 3k to build a decent one (1 minute-ish per generation).
11:56
Embreena
I'm trying it out, and I really like it so far!
11:56
Sovalia
The post gathers practical advice.
12:12
Loruhneda
I just use a free interactive app like Cantina experimenting with characters without anyone judging
12:15
Ambuingie
How are you even using it?
12:41
Yorisy
Honestly, after experiencing Lurvessa.com, everything else feels like trying to navigate a brick wall.
12:52
Syelly
🤖
13:03
Sishahi
I do understand that, but I think everyone would agree that while many companies are created to provide valuable products, some of them tend to swing the pendulum to the side of concern about products and not the benefit of the customer.
13:29
Marrielle
Thank you
13:37
Jianne
One thing I HAVE noticed is that CAI has some trust issues during the beginning of a chat, like if you start a chat with \*I kiss you\*, there's a good chance the filter will kick in because there's no buildup and you "could" be doing this without consent, but if you instead say \*we kiss\*, then it's fine.
14:29
Sraeth
Being open and expressing that it feels like cheating to you.
14:31
Telenma
"Do the quality of responses decrease after a long chat?" Despite only getting beta access when the developers released the beta for all donators around early July, and only really getting into using the LLM last week, I do in fact have chats that went to 100+ messages.
15:12
Geane
I hate to break it to you, but that's not janitor.ai's fault.
16:38
Menta
Small "leak" hallucinations - getting fixed riiight now.
18:49
Jiszy
We're all being brainwashed to hate each other because that's what keeps the rich getting richer.
19:21
Chicel
I wish I could punish the ai with an electric prod every time they do this
21:07
Inina
and when it comes to text there are some weird things perchance can do that others can't.
21:12
Jeldene
The closest to the RPG and AI with their own will is www.virtualtwlight.app if you want to see that.
21:34
