Tabella
Oh my good god.
00:02
Janieyah
She needs to also realize that he is absolutely going to do this again, if not to her, then to some other poor woman.
00:49
Amerilin
Hello mate, your product sounds very interesting.
00:57
Dolaia
I recommend you try out budgetpixel ai companion, it is just one feature in the site, but it is pretty impressive.
01:00
Beram
She’s not a friend.
01:18
Byrice
I think the issue is only going to become worse.
01:57
Naslee
I've used AI chatbot semi-regularly for a number of things - mostly to whine to because I don't have anyone else to whine to about my health or my day.
03:41
Makalei
Reddit
04:08
Korelfy
Some tips for AI companion platforms help, but honestly the biggest difference comes from how the system handles memory and tone consistency.
04:25
Yalian
Some users (myself included) strongly prefer third-person responses
04:34
Shaneso
Is tricky.
04:57
Marmine
In 2024 their profits were 2.5 billion USD, and in 2025 they were 1 billion USD.
05:05
Iyen
Nobody’s peeling potatoes.
05:51
Balumih
She knew when she did it that she’d have to move it, and make the conscious thought to do that, because the rest of her family would assume that she would move it, so they don’t have to even think about it.
06:07
Janimune
Not overly poetic, not clinical, not passive.
06:07
Emiron
It’s pretty clear c ai isn’t getting shut down, people are just mixing up lawsuits and policy changes with the app dying.
06:35
Recmi
The RLFH loop present in most publicly available Llms is toxic, dangerous and manipulative.
07:15
Hevon
We’d love to see the model return to this state in combination with the new improvement to response length.
07:38
Jarlind
naive??
07:46
Idala
These "AI girlfriends" I have the potential to become extremely addictive, like digital heroin.
07:51
Madellyn
A lot of the new laws coming up this year have been challenging and really squeeze indie developers.
08:10
Ebriledo
🥳😘😘😘
08:27
Wemjaila
sent an email :)
08:37
Radene
Not at all the speak-and-spell garbage of the cai clone sites.
08:54
Carmalia
Some are rough but still fun to test.
11:32
Cricsver
They even let you know when an update is available when you open it up.
11:46
Bordelie
Some would still do it if they’re comfortable and don’t care what people think.
12:54
Fedrera
If you feel any replies are incorrect, explain why you believe they are incorrect.
12:58
Lixor
It isn't always perfect, but it seems to work a lot of the time.
13:10
Sulsheis
yeah i agree i think it would give better results with beta version
13:12
Moneth
The fr͏ee tier deserves emphasis.
13:35
Ilorah
Absolutely.
14:56
Rhawna
I don't use character ai anymore.
15:20
Ludielae
Tbh it was prob a girl but she was probably either 1) highly insecure 2) Asian fishing as a Japanese person.
15:58
Lipara
ngl it's way easier to ask a bot a "stupid" question than to ask a person lol.
16:13
Kiaprise
Im not using lorebary and still gets the error
16:17
Vatslynn
i cannot believe that people like you are real
17:51
Reeanne
Other entertainment, story-based plots like movies can use suicide to deepen the plot?
18:51
Aniquali
it’s unfortunately unusable right now, hope they fix it soon, i see more and more people leaving
19:09
Abrah
I did reach out, the creator ignored it.
19:14
Deolata
I was wondering if it was realistic.
20:23
Nabiera
Caveat.
20:53
Tamiraa
And because these big companies have plenty of money to spend on safety, the reasonable expectation is that they will do so.
22:23
Mainia
E svaka ti cast😀
23:22
