Thanea
I see something new for me too.
01:06
Numisha
I don't even like pipsqueak and there is no way I'm paying real money for that.
01:06
Chiston
...
01:20
Mettis
Well every single chat generation costs money if you're using it for free.
01:50
Fleada
As of writing this, it still says it's generating the story...
02:41
Perten
you can scroll down to see "uptime stats" and how low/high they are
03:09
Minsen
Itβs fun in a weird way
03:21
Sayca
It's my personal relationship with my husband of 17 years and we have been together for 24, and he's not worried I'm running off with my phone.
03:59
Lamiana
hard to describe without sounding weird
04:34
Handra
JLLM is exceptional because its decent and it's free.
06:02
Sycelle
I copied a few cai bots into it since I wasnβt interested in the featured ones.
07:18
Mylalia
βTruly uncensoredβ is basically a myth at this pointβ¦ Fr.
07:20
Infahrea
Let's just hope if Michael Rubin is the new ceo of character ai I have high hopes for him, not only he was heavily involved in tech but also digital sports, if he buys Character AI and be the new CEO, because I hope he listen to the feedback and.
07:42
Venca
I can share with you my own way for making intro (but I guess some parts could be considered as basic) Yeah, so basically if you are talking about like the style or format itself, then you just need to write the way you want the bot to keep responding to you (for example, if you write in detail, literary and beautiful describe the actions of the bot in the 3rd person and use formatting like asterisks or quotation marks where you want to, then the bot will be responding to you just like this) BUT if you are talking about the better writing and vocabulary or want to make your writing more detailed and alive, try mentioning what bot feel and hear besides what he sees, use synonims, use small actions (what bot doing in intro, not just standing, right?)
08:09
Zemrea
So that was the only filter that can not be bypassed-the minor filters.
09:31
Nakedie
They look at the graphics first.
10:53
Cukri
You're not married.
11:22
Drayia
The problem is that the Venice configuration is getting heavier and heavier.
14:22
Katoree
The irony is that the underlying LLM technology can absolutely handle long-term memory if you architect the retrieval layer properly.
14:59
Alylojan
I subscribed with Kindroid when I was going through a health issue.
15:20
Bakaina
You always have to trust the platform to some degree.
15:49
Hengil
Mostly Airoboros 70B
16:52
Nalen
Even now, we treat a single piece of negative feedback as a top-priority update.
17:07
Lyley
the servers required to host and stream wuality video and content worldwide burn through energy just like AI data centers do.
17:15
Orwin
I would argue, its not just a guy problem..
18:55
Conetra
Ollama+silly tavern
19:59
Nilix
Two days?
20:30
Jeamy
There are 4 tiers/prices from like 12 a lot .
21:01
Jopomivut
NTA.
22:10
Kritc
feel the same.
22:31
Tanzety
This leads that service providers needs to either run with loss or reduce the quality.
23:13
Kayssa
Ask:
23:21