#🏙│ai-images
1 messages · Page 32 of 1
didn't try. First time I actually saw that tbh, love the OP song btw
I think model is unfamiliar with it, so probably i'll need a lora for that
or damn i2i magic, but I am unfamiliar with it yet
You should totally watch the show and read the comics its your thing
how could you know
I still didn't play in Nier
xD
it's magic
I just keep waiting for a perfect moment for it
but it never come
Me neither and you posted nier artwork so you probably would love ex arm artwork
it may be
I've saved it for later
have no time for anything for now
Have no time for anythingggg nnrrrrrrrr
Joking
Take ur time
⁉️
no way
do u know da wae
To the person who randomly DM'd me... Yeah don't do that without talking to me first.
Mr. Ramthrust can I draw her??
That’s uh… certainly an image on my display!
is this not . Softcore porn lol
A figure of my maid character sitting in front of a Dell monitor that displays 3D model of that figure, generated in Google Gemini with given prompt words.
Yes, this is the maid character I created long ago, still can't think of her name. I trained the image model with my drawings, but Stable Diffusion sometimes messes up supposed details on the character lol.
ai for the day
dude that looks like an 8 String, did you do that on purpose?
AI is perfect in every way and never makes mistakes
well AI image generation has changed a lot since i messed around with it a while ago
it is so much better now
so do you use text only prompts?
wow you must make those pretty detailed prompts then
meh sometimes
depends on the task
I make models and they do most of the heavy lifting
Interesting, ok but last question, do you use Chatgpt or something else?
Chatgpt helps, but booru tags are king
Oh wait, to gen images?
lol.
I use Stable Diffusion -> ComfyUI
okay i was just wondering cause ive heard many people saying chatgpt cause i guess its image generation is really good
It is but it's not trippy / uncensored
Can't get this with gpt
:O
yeah chatgpt has a lot of limits/filters on it so i could see that
thanks though
not bad?
gpt's big strength is test etc; but google's "nano banana" is better
google / gemini
i used stable diffusion for this one
what modellll
the standard one, why?
oh is it? i couldn't tell
don't know what you never tested
toss me a prompt and I'll test it on my model and you run it on sdxl
ok give me a second
ok here "anime girl sitting on a park bench under cherry blossom trees, petals falling around her, soft spring sunlight, peaceful atmosphere, highly detailed, vibrant colors"
do you need a negative prompt?
I got my own; I have to add masterpiece, best quality, newest, because of model architecture
okay thats fine
You win
k
honestly besides that, i think they are perfect.
so did you train your models?
I have before, I took one that was trained and merged it
oh cool, ive never heard of anyone doing that though
clearly, well keep going at it, i might take a little time and mess around with some image generation
i'll also look forward to the stuff you generate
HMU if you need halp but the best place to get new models is civitai .com
thx i will. have a good rest of your day, i got some stuff to do so im going now
I'm trying to find people that are into image generation, hit me up please, just to exchange ideas or places to experiment on discord
show me what u gotttttttttttt
the point was to talk about it like chill friends
bros a member of the workforce 🫡
I need friends,add me plz
dying
I like this server ngl
prolly the most active of all ai, I would most likely promote it on my server too
Yayyyy
HURRY I HAVE TO POOP
any idea how to search a good lora for these faces / or there is a tag for that?
That looks like a lora
I am rofling SO much when I see that
every time
that is Grand Blue dude
but ok, gotta inspect civit then, ty
You can try asking MlGnom
oh shit, right
I won't ask them for help anymore 
I won't sleep today
my ATOT lora?
lmao
yeah
Yeah it's super flawed but KEK
oh was it urs?
yes lol
I saw a few images with this lora from you
didn't notice that it's urs
lora itself is good
but not what I looked for(
I guess I should just wait for me buying new card
for now I'll just stick with my shitty quality
or go back to forge for a while
aight
aight what? 
wanna sing songs?
I'm not very funny person actually xD
just an internet troll
oh and I am professional vodka sucker
(was. after 25y it's not that easy anymore)
lol
da
lmao
vonna sum russian lessonz?
I need 3 words
sure dude
beer, bathroom and girls
ehh, i need them too
I can get by anywhere with those 3
[pivo], [too uh let], [de voch ki]
ugh no gifs!!!
Someone in here DM'd me over something and I hit the ignore butt
Ruh roh
began with a J and referred to an image I made
they should reach out to me here again
yurp
army of your funs don't let you to chill again? xD
Doooood idk!
ah dude, can't believe that one day I'll get new card and it'll be even better...
Wait why is their car right hand drive??
Are they…
B- b- British??
You never seem to dissapoint
Only my parents
idk how to respond to that
have you seen what my drink this afternoon looks like?
well it has water in it
OH NO I PROMISE IT"S NOT PEE
ITS TORANI FLAVOR SYRUPS
Yeah he never disappoints(well except his parents supposedly)
this
Did the ai polaroid trend
Gonna start generating more images now i think
What y'all think
I think polaroids are... what they are
<@&1420600254041952317>
Didn't mean that
Meant what y'all think about the ai pics that i generated with the ai polaroid trend prompt
I think the arms are missing but I like the subjects
dude ur nick is sick
damn now i have this green name i liked my special sd creator blue
:(
kinda wish the contributor role was higher ranked though since it actually took effort to obtain it
DONT EAT THE MUSHROOM
but what if I want...
do iiiiit
I have no idea how or when I got that xD
daaaamn I don't wanna werk(((
first day after 2 weeks of chill
it will pass, it's never more than a week until the next weekend ^^
I'm on shifts
ugh :(
I let Google Gemini to make an image of toy figure of AI Hub-chan. It gave me the human variant of the character. 


This character is named Miocony, the character I also created long ago.
The generated image also borrowed the text I draw on the original artwork and made it stay on the box, which is cool. Also, there's a Dell monitor behind as usual.
🔥
I generated one too
How do I post image
spam in #✦│chat .
Spam in #✦│chat
I'll make one for you; kek.
but the email address will be impdickscambot@crabpenis.net
Give me $500
per month for a year in one day.
so $6000 please.
In bitcoin
...Do you need a verified one?
Because that costs extra.
You won't get anywhere without one
kek
Dummy.
sure just give me your IP
<@&1159293140440723499> Meh, more scambozos
Btw have y'all ever did the ai polaroid camera trend
This result is life inside of a lava lamp given to Weights Voyages enhanced and copied to MonAi enhanced
@MonAi_Art #MonAiArt
is that a reference or a promptemation?
oh coz I thought it was u re imagining it as a prompt
nah, it's basically all from scratch
gtk
i need some prompt ideas for tmrw
that's what you get when you go to cheap traumatologist
which ai u using
SD, ComfyUI
self-made model
http s://civitai.com/models/1465491/ramthrusts-nsfw-pink-alchemy-mix
I will check this out
not my cup of tea
👍
🍌
i see you are going into AI videos now, looking forward to seeing it
Uh
I've been into AI vids bruh
half the stuff I post are vids lol
i know but you've been posting more, that ive ssen
Ah gotcha
sorry i didnt know
so you made this model too?
All the same model, just img -> vid
the vid processor is called wan 2.2
and now I have a music style and lyric generator for Suno... It's really fucking good.
Just made them
hii, where can i write my prompt to generate an image ? is my first time trying to use this
bro bananabot kinda going crazy, with ZERGS base🍌 [hope remix is fine] @dapper linden ping me everytime you post, i think we can make a good combo
oh, it's cute
OKAYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYYyy GAD DAYYYYYMMMMM, yo @dapper linden come check this shish out
emmm
dude, sorry, mb I just don't get it
what should I check? banana bot?
I'm not interested in banana remixes of my stuff that much
I'd like to improve my generations instead
no, the image dude xd
wont remix your stuff more then, got it
nono mate
feel free to use it, remix or whatsoever, no problems
I was just confused by ur tag, thought u forcing me to get into it)
I respect your craft, I'll just remix as a way of respect
fr nah wasnt my intention
when you post just ping me man, you do cool art
absolutely no problem with ur remixing, keep going if it touches
let's get it then!
ty ^-^
will let you know if I do something SFW what I may post here xD
yeah @scenic oxide, and yeah for sure I check that shit give it a lil like
here we go
It's snowing today on AI LAND, cold edit man
face here got fucked haha
details are crazy good
maybe more texture here
eye issues unfortunately
it's 1216x832, after that I'm hiresing it to fullHD
idk how to make it better on this reso, my guess it's too low
I can do tiled upscale to 4k, but this will lead to artifacts
and yeah, every try to upscale is 5-6 mins for my poor 3050(
I think you should prescind upscaling from so low
if you have a good baseline and the upscaling is actually AI upscaling from a standard AI-tool that actually is AI-centered then it works
emm
from which resolution I should upscale then?
Literally anyone I know are generating from 1mp
change upscaling tool
thats lit all
or double upscale, two step upscale, you would rather two time stretch something than one time, it's like physics
I'm scaling to fullHD with latent scaling. Some guys tell that it is shit, but anyway, I like results most of the time
to 4k I use RESRGAN, which is ok as far as I know
what do u suggest?
Right now i have nothing but the tips I gave are just based on logics, tool is not giving results and miss-resolution problems happen, so just 2step upscale or change tool
yk
idk which tool tho
kek
For now I'm stuck with Forge
when I get new vcard I'll just switch to comfy, I have a flow with multistep scaling (one cool guy provided me with that)
but for now... can't do a good multistep. Well, I can - it's just too many moves for that, I'm too lazy to do that for any picture
you are just missing eye damage
your picture is insanely good jsut missing eye position
lets do a test
now imma upscale it, and yeah all perfect except for eyes
they suffer
it's okay that's not upscaled sometimes, can you send one that hasn't been upscaled?
although these shines really look good
wait
here you can see how I'm doing
left is fullHD, right is 4k
it messes with colors, I kinda hate it
and it's tiled, so it corrupts sometimes
for some god damn reason 4K corrected eYES
goooood
this shit is good
you used upcale.media?
meh
no, Forge
not fun for me, I'm afraid
understandable
true haha but the eyes thing we were dicussing, that one does it best from the two main sites
Zerg you're like a magnet
Legit
lol, I suppose ;D
That one just made front page on civ
well, this one did.
cool birb
can I bananafy it? 🍌
I don't think @pure dagger would mind
Nano Banana pretty damn good.
Great for making loras
yeah "GoodEnough" Rank Adaptation Models yeah GeRAS
welcome to "XOHK" and "DЮLUN BIAL"
We have beautiful drunk 6fingered girls and good drinks
😂 😂 welcome to XHONbeubfhjwefu funny af
I just sent a bunch of laugh emojis then reedited
nah dude there were 4 photos of monitor and phone screen
scam
ah, happens
oh so before mine i guess
cool cheesy
fingaaaas
Insane stuff ZERG
I was a little upset lately about my stuff
because I'm surrounded by people that are really nailing it
ty for support, made me much happier
Wdym
about?
The sentence
...
people. together. strong.
there is some exact local discord server. there are some people in it.
and these people do generate things. pictures.
Very beautiful pictures. loras. models. concepts. methods.
for my perception they are like gods
and I'm suppressed by that
cause I'm not even close to their level.
so far that sometimes I think I'm not going forward. More like backward.
and even when they praise me all of this make me think like they are saying that to me because of pity (I hope they don't, but you know... human imagination is wild)
more understandable?
sry for crying, didn't mean to do that
it just... somehow ended like this
yo why did you have to ask
its obvious he is just happier you made him say a whole ass paragraph ahhaha @dapper linden
hahaha
I know
but sheesh this guy needs some line reading
hahaha it was pretty obvious
he is thankful . period ahhaha
well
actually kinda was
cause I need to damn think, cause england is haaard(
well
ofc it's much simpler than russian
but hey, it's my native, so it's easier for me
ty also for ai stuff
don't mention it, I'm just doing what I like
I'm happy that my stuff can make someone happy
Stylish ✨
why does she have 3 BALLS
Experimental style!
you legit just read my mind i was about to say make a mouse girl
MOOSE GIRL?
hmm maybe
In process
Today's daily theme is "cheese"
So I'll have to go a little off topic
:O
But hai
I wonder if your model could do like nachos or something
$10 yes on nachos
I owe you $10
didn't know that's how you're supposed to eat nachos
It's also how you eat popcorn
Not my fault
it's crazy the ai flex it is to just output two different drawings styles big powermove shit
@compact pivot how much do you rate Nana then?
huh?
lol

bro what is ts
obv
ah, forgot.
some stuff above, if you want to mess a little @scenic oxide
meh
too much banana for me
I know hahaha
sometimes too much is cool
like you gotta gooo harddd
BANANA THEN BANANA
using nano banana no less.
Respect 💯 . yeah bananabot really hits+
where?
heh where it hits?
I'm full hp
secrett centerrrr ahahaha
its just a screenshot okay? ahaha
dont worry about quality
hahahha I think I wipped to hard I dont want to overshow now
dont kill me plz
🍌
elon musk man crazyyy
hands broken as always, but idc this time
yesssirrskiiiiii
can we just ignore her ear? 
looks familiar
not mine ofc i was just poiting it out
i know but doesn't it?
i guess the face structure
ai man
AI is the average anime face
so it looks familiar
bro HAHAHAHAH
i mean this
ZERG VS BODYPARTS
?
well thats what came to mind
well gg
kinda wierd the AI got it so good huh
you should see if you could get one with a visor over one eye
Blinky Bill in Hell's Kitchen.
What could go wrong?

dem lewds they sneak up on ya
Its okay, i'll do what i can to get you unbanned
ramthrust you need to be put in a lab for scientific examination
uh why
dude this mommy is so damn good 
Dakimakura featuring AI Hub-chan. 


@dapper linden
yeah, you know me too well 
For some reason, AUTOMATIC1111 is hella slow compared to ComfyUI in terms of generation time, even with the same model and settings. This was run on my RTX 3050 6GB by the way.
A111 is outdated
Are there any modern SD GUIs or should I stick to Comfy?
try Forge
it's much faster than a1111, interface is almost 1to1
my gpu is 3050 as well (but 8Gb)
difference in time between a1111 and forge was like 40-50%
stick to comfy for a myriad of reasons unless you're just looking to make pickydoos
comfy is the defacto standard.
Kitty!
I want to design a surreal painting that looks like it was photographed with a professional camera from a medium-high perspective
It appears as a sad person holding a tape of his life in his hand
A long tape in each part of a period of time in his life
The perspective shows the person's right back from above, showing the hand holding the tape
With cinematic lighting that carries as much emotion
The bar is smaller
Show both hands
Reverse the shooting angle and make it a little higher
Realistic colors (gradation from light (yellow) to dark (blue). Make the person appear in his entire body sitting on a remorse session, wearing a formal dress, and make the room full of elements that add meaning to the painting.
Make the photography angle behind the person from top to right, making it look like an Arab in formal wear
mixed up this channel with GPT?
ok have fun with that

fuck it I'll do it
durp?
You're welcome
kek
lol
In short, the painting represents a man at a crossroads of existence, looking back and realizing that the vibrant, warm, fast-moving river of his life has slowed to a cool, still pond of memory. He is holding the evidence of a happy life, yet his present is defined by a sense of loss for those very moments. It's a brutal depiction of how the beautiful past can become a heavy burden in the quiet of the present.
0_0
Alchemy can read these prompts?
divided by dots, and with 'extra' comas in tags?
DEAR GOD SHE NEEDS TO GET SOME CABLE MANAGEMENT
oh, right
there is no '1girl' in the picture
bingo
Total angel, nothing lewd at all
What’s the name of the repo btw if you can? I might try it out.
ht tps://github.com/lllyasviel/stable-diffusion-webui-forge
@cold lintel
A surreal painting entitled: “The Weight of Decisions”
It represents the internal conflict between the past and the future, and between what a person has lost and what he seeks to achieve.
⸻
Perspective
• A 3/4 angle top perspective (meaning as if the camera is tilted from above to the right).
• It shows a person with his entire body sitting on an old chair in the middle of a large room.
⸻
Personality
• An Iraqi man in formal dress (dark suit, white shirt, tie slightly loosened).
• A session of brokenness and contemplation, looking ahead but with a tired expression.
• He holds an old brass scale in his hands (a double-pan scale).
• On the scales: his life history (clippings, pictures, symbols of memories).
• On the second cuff: a large golden key symbolizing the future and opportunities.
⸻
Background and general atmosphere
• The room's walls are slightly cracked, but it has a window through which hot yellow light enters, fading toward a dark blue in the corners.
• On the back wall: blurred shadows representing people and events (symbolic of past memories).
• Floor: old Iraqi tiles (alternating black and white square).
⸻
Artistic character
• Realistic colors with a surreal touch (the scale and the bar appear brighter than the rest of the elements).
• Color contrast: yellow (hope) versus blue (sadness).
• The lighting is cinematic, focusing on the scale and the person.
؟
kek
I am sorry, I am unable to complete your request. I am a human being, please try again later
MALFUNCTION MALFUNCTION MALFUNCTION
MALFUNCTION MALFUNCTION MALFUNCTION
bro writing a lengthy essay here, go draw it by yourself then
Ok
By dreamseed 4.0
out of Google AI Studio
@lapis flare wat
Nothing i just found funny that she's using a t-shirt with that logo mane 
...Yes. You're here.
Welcome to the promised land.
@steel haven 🤗
Ayy, who knew the AI could get it so right?
...by day...
and foot worker by night
anarchy infighting, in melee, in the streets, in action depiction, watercolor, darkend
that's cool
I like the style
Thanks
I've made a bunch of pics with St Louis and that's so funny
I can't even imagine a picture of her that would be SFW
I just literally can't post her
she is too lewd
Alchemie Labor mit getrockneten Wurzeln
Gotta be kidding 🥀
Party pooper is back.
Hi party pooper
What's up party pooper?
You still poopin?
Parties?
:{
Dude needs to stop being a wet blanket
Do you even know who I am?
Lego Batman?
alr calm down batboy

Why did one leg do 2 steps before the other 🥀
Maybe you have suggestions on how to fix it
?
I probably do
and down she goes like a sack of potatoes
That shit gonna HURT if she fires off
that shit no gonna fire
it's wire is focking wooden, and it's stuck into it
lol
look at the placement of her left hand
I see
that thing goes off it's taking a finger
look at it's bowstring
Classic AI kek
that thing not going anywhere
It could just be a lil dirty 🤣

UR DIRTY
🤣
True lmao didnt even notice it myself 😂
ChatGPT and Gemini are playing Chess
Whoever replies I will post an image of tatsumaki flying…
(Not weird)
Ah, english not a first language?
Im literally from America
That's a strong assumption that you speak English automatically.
Looks like someone might be in another country
...Have you ever watched the news?
In america?
Like... Hello?
TBH I think you live in the batcave.
or a cave with bats
True, but Gotham is in America
The manor
You must have a really nice computer then.
kouyou 
What does that mean?
ask google
Luckily I’m learning Japanese on Duolingo
And what does kouyou have something to do with gojo in chiyoda Tokyo station?
Is he holding the mini Tori
ig it just hallucinates because you might haven't specified proper prompt for that
calling someone a nihonjin living in tokyo or whatever because they know hiragana or such doesn't seem a right way to do
Your bio is one sign
And cool pfp btw
Part 2
I give up
Your not even that wrong. Those were early pictures i generated with openai sora many months ago
I just randomly posted them here because i thought it would be kinda funny
how to work with hitler photo in chatgpt, it says that tos policy etc just it wont do anything
of course contents with dictatorship are not allowed
I'd ask why do anybody even want to do that...
lol wonder why
Wat prampt
May I have permission to help you with images of that’s fine?
Tell me and my 7000 followers how I can improve plz
I'm curious what you have to say
While you're at it tell me how I can improve my model
I usually use Gemini and copilot and grok.
I do not.
Not enough flexibility and I can't make 4k images with them
Let me improve this please
Only if the output is 4k
It’s really easy tho
yes I'm aware
I can type in graphics like GTA or anime
so can I dude
I've been doing this for two years
Again, I made the model
I feel like you don't know what you're saying
Anyone can upload an Image and say "make eyes bigger"
or whatever
But I'm curious how you can objectively improve something with a limit on resolution
c'mon man I'm going to bed
cmonnnnn
@small edge cmonnnn
Sorry
That's not "improved"
Like this
that's just different
I tried
I mean it's pretty and all but it's just not what I aimed to make
like I can type in "realism" and get that
I just like anime
it's what my model and the architecture specializes in
The background is New York
The background is chicago 🤪
It keeps doing the same images man
I tried Microsoft ai
It’s like better
Wdym
It's censored as fuck
: )
Look
Copilot is good
And then Gemini
But gemeni I think in graphics
Dude again
been doing this for two years
I know what bing, gemini, chatgpt and grok can do, and they can't do what I do
Open the image in a browser and zoom in and you'll see noise
Mine doesn't make noise
I render it on my computer, not a datacenter
I built the workflow. bing holds your hand and says "no" if you wanna see a boobie
Bing can't do this
wowzers
again, open on browser and zoom in
cuteee
I started making stuff whenever the novelai leak happened back in like october 2022 i think
so about to be 3 years
very primitive back then, and very good now
holy shiiiiiit
original sd 1.4 was first released in august 2022, 1.5 in september, NAI leak in october then SDXL in 2023, so much stuff happened so fast
still been using the same rtx 3060 from when I started lol, it's gone through a lot, and it's hella resilient. Surprised it hasn't started hitting the shitter yet.
What’s bing?
Bing/copilot is microsoft
the old DALL-E model it is using, ig
the new one is probably ChatGPT's
my old zombie webtoon draft
Give me death
my art is getting better
yea
nano banana my fav model
i hate sora or chatgpt
not sora
chatgpt image generation
Why does it look like sab
heh ball knowledge




