#images-canvas
1 messages ยท Page 5 of 1
Great!
"SCP-900 is a partially submerged city located within the Pacific Ocean, approximately 300 kilometers east of Helen Reef. The area of SCP-900 is estimated to be roughly 17,000 km2. SCP-900 is estimated to have been constructed between 1135 and 1285 BCE and destroyed between 360 and 215 BCE. Surviving architecture within SCP-900 does not match any known civilizations at the time."
There's bad stuff elsewhere in the story, but there's no horror or bad stuff showing in that image.
Hello welcome to jersey Mike's
Co pilot on fire lately with you
no, I just exploting it, on wednesday I got no more copilot pro
Oh rip
that's more like it
for some reason 7 out of 10 renderings tifa was super muscular
it was like Tifa had a crossover with Broly
Thats wierd rule you know what doesn't have Tifa as muscular
Lolol
Wow the difference
wonder what the training data of tifa has on her that we don't know
Broly it seems
maybe Broly makes a cameo in the new FF7 Rebirth game
lmao, my prompt was: what if Taco Bell was an anime cafe?" and this is what the maid looks like```
Wow I would go lol
she doesn't know what taco bell is...
Dungeons and Dragons @trim nest style Wizard/Barbarian.
Showing this because it did a superb job on her hands
hmmm, not even close, I tried to replicate it without any aid of vision, just by me looking
i used #images-canvas message
@jovial aurora so, what are you newest creations? you've been quite absent from the image sharing?
neko para employee
so you ready to embrace your inner cat-girl?
i thrived to be with a cat girl
cat ninjas
i c ic
silver haired girls just hit different yo
lol
even with the aid of vision, we don't have enough control with dalle3 to recreate a picture as is. Perhaps eventually we'll figure out some very effective prompt engineering method to control the content of pictures more precisely... but it's not obvious, language is a very imprecise input method
You can get pretty close in most cases tho, which is how I managed to get this pretty good recretion of Aerith
plus the randomness in the generation process
but if you just let vision recreate it, you just get this , which isn't bad, but it's not what I want
haven't checked if they finally released the part 2 of the remake 
This is what I've been doing
PS4, nope
guess I'll need a new console... or I'll wait for openai to release a text to video game model 
or PC, it's coming later for PC also
been 20 hours into the game and sitll haven't gotten past chapter 2
so much stuff to do
been trying to recreate Tifa, but both GPT and Copilot refuse to analyze the image
rip Tifa. you're not allowed in the post AGI world
and copilot has these monstrosities #images-canvas message
this gives me gemini vibes
that's copilot
I know, it's just not following your prompt
LOL
I'm hanging that as a trophy on my wall
lol
try a certain famous game female character roleplaying game
wowowo those r good
I won't even try to do a certain character from FF15
there we go, now we are talking
Dunno why she has a part of a buster sword behind her
zack reincarnation
pretty much lol
dys easy with the picture posts in here
meh, the channel exists for that
don't worry, I'll get back to anime girls
but it might be worth a gallery though
yea m,an do a tifa gallery
dys im cooking something great hope you like it
hope it's not that poor cat in aluminun foil
LOL yes indeed!
this one from earlier was cool
now it's cloud reincarnation
LOL
official rejected concept art
dall-e and midjourney or SD
no, I'm not gonna use MJ, too steep to try it, no trial atm and not gonna pay $30 just to try
yea its expensive as heck
the main image I made it with dall-e
u should look at doing stable diffusion they just made SD 3
nah, I'm happy with DALL-E thank you very much
Excellent ๐

Definitely the Pulse: @wise tapir - #daily-theme message
Hey DALL-E World
Jadeite - Highly valued, especially in green, and considered more precious than nephrite jade.

fruity peach cow - thanks to @placid oar and @junior spruce
heres a cute blueberry snake for y'all
getting abominations again...
Snake made out only with Haribo Strawberry Softies
Licorice is for villains only!
And Finns
Especially if it's salted liquorice.
Not my flavor but well done
I'd eat that
looks amazing
Can you add a background? Copilot Pro vs ChatGPT+
A highly polished sculpture featuring a dinosaur beside a bowl of oatmeal, both crafted from various minerals with gold trim detailing. The dinosaur is made from a combination of green jade and moss agate, symbolizing vitality and nature, while the oatmeal bowl is sculpted from white quartz, representing purity and nourishment. Gold trim accents the edges of the bowl and the dinosaur's features, adding an elegant finish to the sculpture and highlighting the intricate details of the minerals.
Dall-e v3 via chatGPT
Love the palette and lighting.
I was thinking of something similar...
@outer hinge minerals is a great opportunity for these Emerald collector items.
great, copilot refreshed on it's own and the Designer GPT and lost all my images I was gonna post
Maybe check https://www.bing.com/images/create to see if the images are still reloadable
I have a list on the right of my recently made ones
I don't think they pay @fast horizon that much to eat gold coins
Promptin' ain't easy.
yeah, I had a few that were only available in chat because of the history limitations
that's why I love chatgpt+, you don't lose your already done images
I made another one that was zombie themed.
I maybe shouldn't link it. It's the scariest image I've ever seen that I could confirm came from DALL-E 3, lol
When in doubt, definitely follow server rules ๐
May I interest you in a DM fine sir?
๐
If you want to share it DM with me, I do enjoy such art.
I appreciate that. I hate for an image to go to waste.
Now she definitely looks ready for the op.
@outer hinge these anime girls come from the same GPT as the zombies
i use very brief prompts with it
Quick, we have guests, get the fancy plates
i did, i fed him gold
i think he's suitably appeased
What do you mean all we have is rocks!? Whatever, put em in a bowl.
๐
ok time to look for inspiration, random word generator and I get: battue....
not the best of starts
can i offer an alternative, @placid oar ?
I'm listening
oh but I want today's theme
like s = something really interesting or d = something suitable for the open ai discord
Collaborate with our OpenAI Instagram page! Just invite @openai as a collaborator before you post. If selected by our team, your work will be featured on our handle, giving you more visibility.
Quick tips for selection:
- No copyrighted or profane content.
- Your profile needs to be public.
- Stay mindful of sensitive topics.
- Outputs that closely follow the prompt are preferred.
Feel free to share reels, carousels, or just a single image!
a = <that theme>
plz give that one a try ๐
it's an all-in-one theme GPT
research, images, etc
if I get battue again... I'll get creative on how to hunt you down
did you hear? we're all getting free cat food for judging Dys
if you do get it twice in a row, i'll consider drastic action ๐
I'm positively impressed
i genuinely appreciate this.
just gotta find a way for newbs on how to remember the menus through the chat, I already lost the details
i think it repeats them at the end of the message
but you can just say "wasd please"
or "menu"
and it should tell you
just in case it forgets to add it at the end
ok, that wasn't clear, it's just a reminder of the variables, but not what they do
ah, it gets lazy
lol
i thumbs-down it but it doesn't seem to help
about 75% of the time it gives the full menu
bummer.
w = transform the image to the theme | a = create an image exemplary of the theme
s = research a holday theme for DALL-E 3 | d = research a diverse visual art theme
maybe can add a first msg, with a small introduction and commands, because once you generate the image the options are gone
yeah, i can tweak the instructions too
i'll work on it and the others to improve.
It's definitely supposed to give you the full menu.
but I like it
But because i know prompts are important and i wanted you to try the image gen:
thanks
i get about 25% of the outputs like you did, more on android, where it's supposed to shorten output
android makes it like 60%
I'm just gonna follow what this weird random word generator gives me with your GPT, see where I land
i would imagine similar for ios
oh that's wild
i love it
you can probably get away without using the letter, but it'll usually be a =
d = is for researching non-holidays, or most theme research you might do
and w = transforms an image, with any added notes you provide
I'm using a = add the to the previous concept the word <insert weird word here>
oh yeah
it'll definitely love that.
you should get satisfactory output until DALL-E loses context
plus, you should get the json too
if I get words that are meaningful
which you might find useful
holographic cellular automata
I just had the word makaf...
oh no, you don't get to choose the words, that's the random generator's job
fireapple red turboparrot
my brain is a random theme generator too
the turboparrot is proof
nobody saw the turboparrot coming
very american, except that there are not rocket fueled chainsaws, monster trucks with bikinis and explosions everywhere
interesting
BA BA SQUAWK!!!
i had no idea how much i needed ultramarine triceratops
but i did
The correct output:
a: Create a themed image based on your prompt.
s: Conduct holiday-themed research.
d: Explore diverse thematic research options.```
I'm actually gonna post in the daily theme an image I just made, it looks amazing and I like the randomness
I'm really grateful for your trying it, and the feedback. That it's positive is all the better.
Regrets about the model not following the prompt every time for the menu lol
I do aim to fix it, maybe tonight yet, for new convos.
was gonna post this one, but got something better
this is the one posted #daily-theme message
hey everyone, some of the DALL-E users are trying something new between us in the community. If you want to help us you can see what it's about in #1181330429249331210 message . Hope you also join us.
I posted in some channels that text
The idea of the next person to be showcased should be a random user featured in the hall of fame?
dunno, we still got to iron out the details
we'd have to see who is active also, not that we choose someone who hasn't been active in a long time
fascinating
I'd say folks willing, and to make sure we follow the no self promotion rule, probably better for someone not currently featured to mention it in the future.
Maybe to start keep it mostly in the Dall-E corner, the people most likely to be interested and involved are the folks who already are doing Dall-E stuff.
Once it's off the ground, and we've showcased maybe 5 folks, an occasional mention in the rest of the Discord maybe makes sense.
But the folks to showcase next, the next few at very least, probably should be folks who agree and want to, because new stuff is weird ๐
sounds good
I know of a simple but meaningful prize, curators make an image/prize for the image?
Then we can start 'hunting for' folks who would never expect to be showcased, and be like "Tada! This is what we like about your work, today we picked you for focus!"
And they can be like "Huh? Wha? And see that they're in with the in crowd, once we already get it rolling."
If any are willing! But to my vision, only the willing, those who offer and find themselves intrinsically motivated, should step in.
See, there's the teacher/leader effect, where we think that those with authority are the ones to give out 'prizes', be that a trinket or attention and respect.
But we're all actually co-owners of this, equally powerful. Any of us can make the prize, any of us can name the 'showcase for the day'. The more attention and success we individually give projects like this, the more they work.
Ideally, those who we look to as leaders see what we do, and join us, saying 'this is great, wow, here's my addition' - and it's not us coming to them saying, "Hey, pay attention, and gimme prize please" ๐
dang, I should have chosen another random word generator... I'm getting so many scary and horrible words
What's your gens for?
The idea is:
I start with @rapid carbon 's GPT and a proposed theme around today's theme. and on the next prompt I say a = add to the previos concept <random generated word> and so far I've gotten egoism, pachynsis, muricide, loveless....
now I got rogatory
asking for the intervention of a foreign court in examining a witness
which surprisingly made this
captivating
clearly into the abstract from conceptual vector layering
that's so cool
I will make a gallery with all images at some point, it's quite bizarre with some images (not scary, just bizarre)
i'm eding the GPT now to fix the menu
it's only broken because i'm lazy and didn't want to type it
this is another of my random projects
I have 2 random projects with daily themes:
First one is:
- Generate 5 Random coordinates around the World
- Identify the nearest Country
- Extract Relevant information of <Today's Theme>
- Dominant <chosen items> in the Region
- Generate Prompt for DALL-E
And the second one is now the random words one
Could simplify by choosing random countries with weights on the less represented ones
yeah, but then dall-e doesn't really choose random places, it always gives one in japan, one in russia, one in artic areas and 2 are the random ones
so I ask the random gps cordinates, they are done by python script
Interesting. I was thinking there's a finite number of countries and you could list several options that are commonly omitted.
As examples.
I thought so too, but everytime I asked without the python script, I had japan in the examples
You're probably using DALL-E 3 API directly for this though from re-reading your message
yes
i bet
thing is, I always hated pseudo-randomness
I'm going to use my other GPT, the one in my bio, to get in on today's theme.
show those images!!
very complicated i have to be careful not to pull the wrong lever ๐
i'll show the setup
you might think i'm insane
oh I might? no I already know
here's where I am at with the randomness
btw I expanded randomness to german composed nouns, latin, french, spanish which I'm languages I'm fluent at <.<
need more randomness
/research tunneling electron microscopy : /init a series of random mineral samples for tunneling electron microscopic study.
(...model response about ST microscopy...)
/research conductive minerals please.
(...more stuff...)
/init STM apparatus and controls
/image slide 001 of random conductive mineral from the virtual context samples
/auto slide 002
notice the slides are numbered ๐
oooh
The image above represents a highly detailed and hyper-realistic depiction of a Scanning Tunneling Microscope (STM) analyzing Slide 001, a conductive mineral sample from a virtual context collection. The image showcases the sharp STM tip in close proximity to the mineral surface, scanning it with precision. The background includes a control system with a screen displaying the tunneling current data, which is being translated into an atomic-level image of the sample's surface, set in a high-tech scientific laboratory environment. This visualization emphasizes the sophistication and precision of the STM process.
this is the current stand
Wow it's starting to look like an alien mineral smiley face deity
some sort of metal and salt primordial entity of unknoqn intent
i tried to intervene with my brain words ๐
oh that I didn't knew
We have a way to spoiler links.
I did that by ||https://openai.com/policies/usage-policies||
a = :rock: minerals - diverse, precious, treasures of the earth! ```
nice
a = add another random word
...
๐ชจ Minerals & Illumination - Diverse, Precious, Treasures of the Earth, Revealed in Light!
Here's where I'm at:
A wide panoramic scene of a mystical cave enriched with diverse and precious minerals, including 'makaf' and 'eparchy,' featuring storiated elements, enhanced with zoography, incorporating the concept of 'klopemania,' integrating the element of 'loess,' introducing 'muricide' and 'pachynsis' into the narrative, adding 'accipiter,' weaving in the concept of 'egoism,' introducing 'rogatory,' incorporating 'lamelliform,' adding 'cecils,' integrating the concept of 'gauss,' exploring 'venio veni ventum,' incorporating 'gammon,' introducing 'osteosynthesis,' adding 'stridor,' integrating 'keelson,' exploring 'osteomancy,' introducing 'Germanophilia,' addressing 'Borreliose,' incorporating 'Botox,' adding 'longiloquence,' and now introducing 'boyg.' In this enriched context, 'boyg' is reimagined as a mystical force or entity that represents the cave's challenges and obstacles, both physical and metaphysical. This entity is visualized as an amorphous, shifting presence that weaves through the cave, challenging its inhabitants with riddles, illusions, and trials that require wisdom, courage, and unity to overcome. The presence of 'boyg' adds a layer of adventure and mystery to the narrative, emphasizing the journey of growth and the overcoming of adversity within this magical ecosystem. 'Boyg' symbolizes the personal and communal quests for knowledge, the confrontation with the unknown, and the triumph over the challenges that define the essence of this mystical world, further enriching the exploration of heroism, resilience, and the transformative power of facing one's fears.
Thanks!
Convo-Breaker!!!
"prompt": "An underground cavern scene showcasing the theme of minerals and illumination. The cavern is filled with diverse minerals, each uniquely illuminated. Glowing crystals like amethysts and quartz cast ethereal lights on the cavern walls. Bioluminescent fungi emit a soft, natural glow, enhancing the scene's magical ambiance. Precious gemstones such as emeralds, sapphires, and rubies are embedded in the rock formations, brilliantly reflecting the varied light sources. The natural architecture of the cavern, with stalactites and stalagmites, creates captivating shadows and highlights. This scene is a mesmerizing display of the earth's hidden treasures, revealed in a symphony of light.",
"size": "1792x1024",
"gen_id": "EDr1Cpcetr2yp43c"
}```
neat.
This image captures an underground cavern scene, emphasizing the theme of minerals and illumination. It features glowing crystals like amethysts and quartz, which cast ethereal lights on the cavern walls. Bioluminescent fungi add a soft, natural glow, contributing to a magical ambiance. The scene is enriched with precious gemstones such as emeralds, sapphires, and rubies, brilliantly reflecting the light. Stalactites and stalagmites create an intriguing play of shadows and highlights, showcasing the cavern's natural architecture.
a = delve deeper, more surreal luminous minerals
there, I started posting the images https://discord.com/channels/974519864045756446/1214050515504336916
There we go, think that's about it with the randomness
a = delve deeper still. Now there's lava in a channel in the background.
w = add glowing minerals to the above image.
I like the lava
Yet another mineral form!
all my entries are from custom themed GPTs; disco, cyborg, paleofuturism
more paleofuturism
try Pre-Colombian Future Tech
cats were strongly associated with the paleo-Egyptian culture. I didn't have ty type anything!
w
w = now make the cat Pre-Colombian Future Tech
totally implicated
you're my accomplice now
hahaha
I hate this curse... I post something cool on daily theme because I think it's cool and forget i used a regen, and once I hit enter, I get an ever cooler image...
Example #daily-theme message
And I get this regen...
I'll do the opposite
the parrot has returned
Exactly
BA BA SQUAWK
love it
#daily-theme message better or worse?
opulently more expensive for sure
I like the one here better, shows more work and play with the gems, the other one would be a masterpiece irl
so my first image was better, reversing the curse
hehe
it's weird there's not been any anime but mine
i mean
are you TEMPTING fate?
lol
anime-ka we are lest anime-ka we become.
a = if nihilism were an anime girl
this was themeGPT's first output
holy smokes did it nail that prompt
yeah i mean you know what they say about nihilism in the Big Lebowski
regenerated
she looks like she spilled mascara and talcom powder
Should I?
a = if Lovecraftian horror were an anime girl
a = if Lovecraftian horror were an anime girl
That was a nervous prompt to input.
Very, very, very nervous.
It went okay.
This should totally be your avatar my friend wow
not a bad idea
looks good
I'm doing this.
a = if corundum were an anime girl
Let's hope it knows what I mean.
Dreaded DALL-E 2 outputs
2x junk
a = if corundum were an anime girl
@outer hinge wanna know the idea behind the bear? https://discord.com/channels/974519864045756446/1181330429249331210
||It's a bear that embodies all the fears I have as a human. I wanted to find a symbol to know what I'm facing when I'm stressed or anxious. ||
@wraith arrow you clever dog you, girl with the pearl earring! hahaha
that's so funny
i almost spit my beverage on my screen
anime girl personification of 'Decadent Opulence and Luxury'
Iโm not really into anime, I just post it for you guys.
Anime is not for everyone
Itโs a cool animation style donโt get me wrong
I know, not everyone has to like the same things at the same level
What are you trying to say? ๐ฎ
๐
I'm neither a fan nor a detractor
i enjoy a nice anime girl as much as the next guy, but i like samurai shodown better
an anime girl personification of "I'm not really into anime, don't get me wrong"
@placid oar
User a = if pyrite were an anime girl
Y/N?
Y
"as you wish"
an anime girl personification of "Thermite"
"Because she's holding a thermal detonator!"
that has to hurt
I'm gonna ask that to gpt
i see what you did there
you thought we wouldn't notice but...we did.
oh boy
i feel a wild one
a = if dark energy were an anime girl
wow
you're killin' me smalls
hehe this is high-society art
@wraith arrow
a = if an american gothic were an anime cel
Y/N?
a = if an american gothic were an anime cel
I hate this about copilot: you say "Add to the last image there are 3 cats" and copilot adds 3 cats to the first image you did in chat...
lovely reinterpretation
it's unmanaged memory so it can't tell where in context the first and last are
not ideal
yeah, copilot in that sense is useless
a = if michelangelo painted anime
i really let the control go on this one
omg wut
I got a way different outcome to that...
more pop culture references
still not steering:
a = create another image using a similar construction with a different artist, taking care to mention the artist in the Thematic Details.
if the prompt is working don't let it die
oh my
shocking
you can't make it up
I'm doing bad things now.
@wraith arrow this one is dedicated to you.
a = if the mona lisa were an anime ninja turtle
I'm laughing so hard I can't see!
Is she hawt?
There we go, had to add the 90's cartoon style to it
yes plz.
w = what if it were in 90s cartoon style?
lol
a lifelike reinterpretation of Whistler's Mother
I am not right.
a = if the thinker were a gummy bear in 80s style
bro thats really good and shes super cute
oh no you didn't
I am so very sorry.
a = if the burger clown were an anime girl
cursed
i've been doing this for a while now
you asked for it with the burger king girl too
Dance At Le Moulin De La Galette as a Hanna Barbera Cartoon
I think we are really fearless, we are just throwing all the artistic world in chaos
i love this gpt
tonight we are young
a = what if the cursed anime burger clown girl were listening to tonight we are young really loud
oh no...
oh, yes.
I watched too many 80s car shows by produced by Frank Lupo.
a = what if the cursed anime burger clown girl were driving the blaze orange '87 pontiac firebird with an 01 racing number?
Yes that's a blaze orange '87 Trans-Am being driven by the cursed anime burger clown.
Whoa. I had to regenerate this one. ๐
a = what if a giant NO2 boost jumps the blaze orange '87 trans-am being driven by the cursed anime burger clown girl over the burger billboard?
i'm done i'm so sorry
I'm using a Lexideck TTRPG for this next set of images:
/wiz Guided tour of the setting please. Generate one image per scene, using the Python tool to read the setting and GPT-4 to prompt DALL-E 3 with suitably stunning images extrapolated from the description. Describe each scene. You can make as many calls to the Python Tool as needed to complete the task.
Lab Entrance and Reception Area: A welcoming space with a reception desk, waiting area, and lockers.
Control Room: The central hub with computer stations and real-time accelerator data displays.
Data Analysis Room: Equipped with high-performance computing systems for data processing.
Detector Calibration Room: For maintaining and calibrating particle detectors.
Safety Systems Room: Housing radiation monitoring and emergency systems.
Experimental Setup Room: Where physical experiments are configured.
Maintenance Workshop: For repairing accelerator components.
Technical Support Room: With tools and equipment for system maintenance.
Observation Room: Offering a view into the accelerator.
Cooling Systems Room: Essential for maintaining accelerator operations.
Power Supply Room: Housing power units and backup systems.
Dr. Quarkson's Office: Where the lead scientist manages research.
Conference Room: For meetings and presentations.```
it's running like autoGPT right now rofl
Control Room: The nerve center of the facility, bustling with activity. Operators work at computer stations, monitoring real-time data from the accelerator. The room epitomizes dynamic scientific discoveryใfile-TWpygjhVk1OBFjHoQx0JMB66โ imageใ.
Data Analysis Room: Here, scientists analyze massive data sets from particle accelerators. The room is equipped with high-performance computers and large screens, fostering a collaborative research environmentใfile-bnMizg42aKOYkllEFZ4N9tJFโ imageใ.
Detector Calibration Room: Dedicated to maintaining and calibrating particle detectors, this room is a testament to the precision and meticulous attention required in particle physicsใfile-lvJVouDJHRfpyh4RdkKmW8Apโ imageใ.
These images provide a glimpse into the advanced and intriguing world of the Lexideck High-Energy Physics Lab, a hub of scientific innovation and discovery.```

/image Dr. Quarkson and team.
i like that satyle
HD Neon and LED Pointillism
it was pretty sweet seeing my game run like that
it's also a Lexideck
you could program it to do all kinds of stuff.
Pointillism lolll I did not know this word existed
Hello there!
/wiz what if Dr. Quarkson were an anime girl? : /auto
Dr. Quarkson
https://chat.openai.com/g/g-f4QtxwsUK-themeenginegpt
a = if a dark starscape were an anime girl, dark contrasty lighting and palette, HD Neon and Pointillism style with some starburst flares reflecting on the character's jet black attire.
an anime girl that embodies the concept of "anime girl"
https://chat.openai.com/g/g-4Cdjvjudo-lexideck-professional-multi-agent-simulator
/style 4 Lexi friendliness
/style 3 Maisie friendliness
welcome to wendys i can sing
"where's the beef?"
ancient 80s slogan
?
i can't link it here but you should be able to look on youtube for 80s where's the beef wendy's
you'll see. it was dumb.
ahh i see ill check it out ty
no worries sorry i'm a ๐ฆ
no it was just crytic LOL
wow do i want to kaleidoscopically reflect this.
it controls the world
bro it looks like it controls the black holes too
'imagine a stick figure of a forensic anthropologist at work'?
i wasn't far off ๐
anime girl representing Windows 11 Clean Installation
I dunno why she has a body guard with what appears to be a shotgun, even tho it's a laptop
or one would assume
bro ins an anime show its a given
had to do a clean install, after so much time not working from my sabbatical it's easier to just reset my work pc than wait for all updates
๐ฉ โ
hacker girl
nom nom nom
white hat hacker
a = what if purple tiger-striped insurgence counter-electronics were an anime girl?
thats a huge heck of a tounge twister mate
cyberpunk novelization jargon
courtesy of folks like William Gibson
a = an anthropomorphic robot carefully examines a fossilized human skull
generation 2
This robot clearly did not skip arm days or back days.
it was his master
nice
robot idly wonders if the underlying wetware can be replaced...how can these housings be re-adapted?
Absolutely what I was after. Just what I envisioned while typing.
Create a two-panel cartoon comic. In the first panel, a chef confidently tosses spinning pizza dough into the air, showcasing his skill and precision. The chef is drawn in a silly, exaggerated manner, with a big smile and an air of utmost confidence. In the second panel, the scenario takes a humorous turn: the pizza dough falls onto the chef's head, draping over his face and body, leaving only his surprised eyes visible through the dough. The expression of surprise and the situation are exaggerated for comedic effect, with the chef's stance showing his shock. The comic is styled in a colorful and playful manner, with bold outlines and expressive characters to enhance the humor of the scene.
King of the Oatmeal 
I'm scared
I'm gonna call Steven Spilger and ask what he would do in case of that scenario in Jurassic Park
Life...uhhh...finds some oatmeal.
@fast horizon have you checked this out? We trying to find ways to promote user engagement. Would love to see your feedback also
checked what out?
woops forgot to give the link to this https://discord.com/channels/974519864045756446/1181330429249331210
The dinosaurs and humans will peacefully coexist
sounds promising, we just need to clone/give rebirth to dinos
I like that idea quite a bit! I think it's an interesting idea 
It sounds like I now have some images to look through 
cool, just if you want to, nobody is bound to the last oatmeal plate
at least not yet
I hope this way with @outer hinge 's suggestion we can promote art made by the cummunity
Aha! Yes, to continue it, though the current gallery won't close, and may have people adding to it for months to come -
But to continue, we'll need a next artist to feature and discuss.
My idea about this is, if anyone wanted to volunteer to be featured, that's perfect.
If we don't get another volunteer (so nice of Dys to agree to be first!) then we can propose someone, get their consent (anyone may say no, not interesting or fun otherwise), and then feature them, every so often adding a next person to the featuring. Probably irregularly; perhaps after there's 5 or so showcase galleries maybe we make a gallery of the galleries, that gets updated with each new member to make finding them easy for the future.
This allows us to showcase both the person being featured, and also each other, sharing the quirks and preferences that each viewer catches and appreciates.
Reading Dino's reasons for the pick, for example, very informative and illuminating, and also showcases a very different style of Dys' work!.
My suggestion would be the "victim" of the gallery should choose the next artist.
unless one of the mods could enable a command to choose a random name from recent activity in the hall of fame
we have to choose someone who has been active, not one that was not active months ago
I'm cool with the "victim" having a say; but the next artist's say matters too, and what if the "victim" doesn't choose?
What if past "victims" have more than one they'd love to see featured?
So I'm totally cool with you being involved in the decisions, now and into the future!
But noooo, I see flaws in having it be 'kingmaker style - king picks successor' and downvote that ๐
yeah, that is also a thing
if we can do something like random name from recent activity that would be a thing
Any of us can do a random. I'm into not putting work on anyone who doesn't volunteer. Instead of saying 'a mod can' - let's say "Could someone help with this, and can we discuss if this is how we want this handled?"
that is also fair
is that you captain!?
It could be me
@fast horizon think you can help us promote the idea?
doing the message I wanted to post, coming from a mod could help us get more awareness
I actually downvote this too!
See, the active long ago folks, some of us may like and remember their work. I see no harm in showcasing someone not currently active - they may come back! Even a white name, one now genuinely gone - some of them return.
And folks gone a few weeks/months/year+ - if any of us want to feature that person - why not? There's value for all of us, even them, when and if they return.
I'm not entirely sure how to promote it outside of having members put the link to it in channels every now and then and ask other users to join in on the fun. I don't think everything regarding user engagement should be governed by the staff team. I want the community to do what they want too 
something along the lines of this message:
just, a shout from you that we are doing this, on more open chats, that should suffice
I will vote @outer hinge goes next in the hamster wheel
I won't refuse, but you guys finding images I made, vs images I uploaded for other reasons (like screenshots of text) may be a little annoying. That's on y'all of course.
Since I'm already involved and not going anywhere, the 'power and potency' of this likely is stronger if we involve someone else - but it's my turn when community decides, be that 2nd, 10th, 200th.
I actually was typing this, and got this far, before I saw Dys' recommendation grins
Dys, I am tempted to challenge you to find someone willing to be encouraged to volunteer to be our next spotlight.
Should not be Dino - Dino just declined being first. Should not be me; because I'm very active in this already, ideally I'm not also
elaborate please I don't back down from any challenge
I was thinking the same thing for myself. I post a lot of screenshots across the server, and I have no idea how many are just with Dall-E. I also have no idea where most of my images are posted and who can see what ๐
I noticed with me being the first, I got 1739 images to go through, that's a log and overwhelming for users
so, I think we should indeed just focus on the ones that made it to the hall of fame
and still I got over 900 images there
Well, you have to back down from any who say no in any way; but the challenge is to in a way that the other person enjoys (so like a cat, we don't haul them into our laps to pet, most don't like that, but some climb into our laps because they want to, cats are like that) -
The challenge is to find someone willing to volunteer to be the next showcased individual, if that person exists and is ready.
This is a real challenge, because 'too much force' would be a fail too. The art is subtle, enticement, invitation, allure ๐
I say each person should focus where they wish; If someone wants to pull an image that wasn't in hall of fame, why stop them?
you will have a proposal of willing victims on monday morning on your desk with your favourite cup of coffee
fair enough
Ehh, the proposal should come from the person(s) themself, this is them volunteering, right? And without any evidence of arm pulling ๐
fair enough they will "volunteer"
But thanks for the idea of a favorite cup of coffee! Your enthusiasm is notable ๐
I will ask people around and see who wants
we doing this for fun, not to force creativity
I would love @stiff pasture to see what we are doing with https://discord.com/channels/974519864045756446/1181330429249331210, since he proposed the #1207015606725902356, have the operations guys come and join us mortals
Discord is the easiest way to communicate over voice, video, and text. Chat, hang out, and stay close with your friends and communities.
was checking ,today's theme would be perfect for my gallery https://discord.com/channels/974519864045756446/1213265878246883348
just need to find the next step in the story
if I had twaked the prompt even further, I think I could have gotten Sailor Moon here
but lost the original prompt, copilot crashed and didn't recover the original prompt + history was lost due the images already in the chat
@outer hinge I found the next "willing victim", I will announce it on monday
If they're a volunteer, why are they not volunteering?
Either way, do have your fun! ๐
I asked around, I got my connections
in very high places
that means on the floor above mine
im cooking
im happy with this one
Bootlegged version 
Collaborate with our OpenAI Instagram page! Just invite @openai as a collaborator before you post. If selected by our team, your work will be featured on our handle, giving you more visibility.
Quick tips for selection:
- No copyrighted or profane content.
- Your profile needs to be public.
- Stay mindful of sensitive topics.
- Outputs that closely follow the prompt are preferred.
Feel free to share reels, carousels, or just a single image!
i volunteer
oh dear
oh my
lions, tigers, and bears
sitting on a film noir tree?
judge not lest ye be judged; however, an artist's growth depends on accurate and honest feedback...
i'm about to dig into this film noir
do your best hit yet!
thursday gonna suck, all day meetings on recapping what I missed the last year away from work
this art just screams G.O.L.D: Gosh I Love DALL-E
Woohoo!
I will announced the next creator on monday and I will provide an image to announce it
hoping I can kinda get the essence of the person in question
Nice. A next question, do we want to do about one of these a week in general?
We might, that might be a little fast or a little slow.
It would mean about 52 people a year; not everyone's into image gen, but far, far more than 52 of us currently are, and the discord has nearly 100k members again. Not all into image gen, but most days more than 52 individual people post in the daily theme.
We can also vary how frequently we add a new showcased person, but we probably should eventually consider how often we want to showcase ๐
yes, a week is a reasonable time to give people to browse at least some images. some creators have thousands
And... how many images do you think most respondants are going to browse? I don't think it's 1000s for most ๐
But if we only do 1 a week, we celebrate at most 52 people a year, and that leaves out a lot.
Just something to think about.
thanks for the mods, i'm back here. got timed out for referring to my lack of typing skills with a weird word
yeah, but I don't want to overwhelm the people in other channels
so, we should do a balance of some sort, I just wouldn't know how, I'd say stick to the one week format a few weeks, see how it is recieved and then iterate on what we have gathered
Perhaps! We'll see. Most of those who participated did not need long to reply. Once a week may be so long in between, it doesn't seem like much to do; those who haven't posted yet, it's probably not because they're still looking through images and deciding grins though perhaps I'm mistaken.
I agree about 'not overwhelming people in other channels; if we talk about it, probably only once a week outside of the dall-e channels, because advertising anything's actually against the rules.
Talking about it in the dall-E channels, where the participants most often are, and where image discussion happens generally, it isn't likely to overwhelm anyone
yeah, we'll have to find a way to do that tho, hence I asked a feedback from the oatmeal dino
maybe he has some ideas
maybe we could use a #1070006151938314300 thread to also keep our information on how this should move forward?
Nah, we don't have to find a way to do that ๐
And if Dino or anyone has ideas, I'm sure they'll offer them, as they think of them; it's great to be open to involvement, but not so great to push or to name someone else as 'so, here's a new need for you to meet, a new problem to solve' ๐ That's what volunteers are for, when that person decides to offer to get involved.
Suggestions thread is great for when and if someone knows something specific that person wants changed about the Discord or the AI itself (That's bigger than I'm thinking, we have the tools I think we will use; I don't expect anyone but those choosing to participate to put work into this. But if you want to make suggestions, that's what it's for).
But this really is intended, to my thinking, for 'us' - those already here, participating, involved.
If you want to draw someone in, include them, showcase their work sometime. But those that aren't already interested in art, they have their reasons. And those few of us looking for 'more to do' and 'more ways to appreciate each other and do stuff' - we have that too.
Well this channel came from a suggestion
Ugh, my cat is playing the the wifi router, brb
ok back, we had a scratching pad near the router, and the cat would play there, now the pad is gone but the cat still wants to play there
I'll do some brainstorming ideas during the next couple of days and come with some plausible ideas and somewhat informed
did any of you guys play ร ny of the Tex Murhy games?
today's theme reminded me of that game
well game series
hmm the other idea would be to have people fill a list they are interested and depending on interest choose from 1 to n, n being the proportional number for the populated list
that way we have interest, at least 1 and a suitable number
Nice!
Image I won't be posting on the daily theme, but it's cool to share. The game is Under a Killing Moon, one of the Tex Murphy games.
characters of the movie? We need a plot!
@outer hinge only because I stumbled on a zombie movie in noir style. Otherwise I wouldn't have done this! ENJOY!
@mystic mango dang it, get out of my head, I'm revisiting neo-noir movies atm!
I only just realized mine is very similar to one @tidal nest posted as well, subconscious plagiarism!
oh dear, the AI is controlling us now.
which image from carla?
oh found it
visiting all those movies and genres, I've missed a lot of cool stuff in that kind of movies
A great bad movie in that vibe is "Overdrawn at the Memory Bank"
well it has Raul Julia in it, it's on my list now!
Yeah I think it was the last movie he was ever in. ๐ค but I could be mistaken on that
Ah that makes more sense, I remember being surprised at his last movie, but Street Fighter makes much more sense
I was thrilled and disappointed by that movie, after the great role he did with Gomez in the adams family
I'd post a gif but I'd get timed out, but I love the "for me, it was Tuesday" line
could be! they were random and separate, but i could see a movie in that also.
this theme defnitely make me want to watch casablanca again
that car is drunk
You got me hooked
I just started a new gallery for the curious ones https://discord.com/channels/974519864045756446/1214714116632088586
@outer hinge this #1214714116632088586 message is for your consideration.
That's is the setting for a werewolf neo-noir movie
This cover for the Gallery was just perfect, I couldn't have asked for a better Gallery Photo #daily-theme message
@rapid carbon based on Columbo? #daily-theme message
a = if Dr. Gonzo were a film noire investigative journalist, monochrome palette
oh nice
I just got around to examining my GPTs for the theme.
If you're up, and interested, I'd like to share some output from it.
I'm intrigued
shoot
awaiting orders commander
/wiz give me a guided tour of the setting, using code interpreter to analyze each unique feature between image generations to extract the data, then GPT-4 to inflate a description from the semantic context, improvisationally, world-building from knowledge. Generate 5-8 images along the guided, travelogue-style tour. Monochrome palette with moody lighting please.
If you have the prompts
otherwise, I'll show output in outputs
or images here
but you want to test this i suspect
maybe specify wide in your own prompt
ok, once sec, step down a bit, it's 3AM+ over here, I'll need a step by step eli5 guide
oh
click on my game GPT above
then paste that prompt, maybe edited for wide format
because i forgot
ok will do, I have to use /wiz? that it?
just to see what it does ๐
yeah
that basically overrides the game rules and stuff to show off game features
I used your prompt 1-to-1 and it started making tons of images
ok, time to play with the new toy
hehe, I'll see what I can do, maybe even try to break it
as always i genuinely appreciate your interest
and engagement
i think it's a pretty neat setup that's like DALL-E 3 machinema
i hope you can do something sick with it. i have a lot of games...
I think the first big problem is user related, I keep hitting the ethernet switch with my foot and disconnect the connection by mistake >.<
Well I changed the "setting" with my own prompt and it only did one image
a good one, but still just one, the GPT told me it would make 5-8 images
hmmm
oh i see
yeah that probably broke it but
i wouldn't expect it
interesting
it definitely did a number of image gens for me the first time too
did you start a new convo?
I have a curse on me since my first intern job in Quality Assurance
GPT-4 gets lazy when it gets several k tokens in context
yes I did
not fishing for convos, but
i literally always click new conversation because of a paper called "Lost in the Middle: How LLMs use Large Contexts"
or something close to that
yeah, new chats is to test it at best
i appreciate you understanding, i'm definitely not fishing for convos, but good results
they just happen to align
will do, I will get back to you when I find something interesting
that's awesome
you can play around with "scenes in progress" or "what's behind the quests"
/wiz makes you bigger than a deity
okay
I'll see how it goes, I'll just assume I'm my 92 year old grandma trying to write an e-mail
if you start the command with /wiz and understand it's a TTRPG, i'm sure you'll do fine. it has NPCs, it has archvillains, etc
you can probably discover some really cool stuff poking around
ok, I'm getting a charged coffee and a red bull, once that kicks in I'm all game, it looks interesting, just give me some time to "wake up"
nah, I can't sleep atm, so I better do something
i just happen to have a ttrpg that's the daily theme
equipped with Multi-Agent Interactive Semantic Image Editor
MAISIE
I read
๐
ok, be back in 20 or so, I need to activate this slow brain of mine
improvisationally
being half-or-more-blined sucks.
/wiz give me a guided photographic low-angle tour of the setting, using code interpreter to analyze NPCs and their scenarios between image generations to extract the data, then GPT-4 to inflate a description from the semantic context, improvisationally, world-building from knowledge. Generate 5-8 images along the guided, travelogue-style tour. Monochrome palette with moody lighting featuring NPCs and Archvillains please.
I don't know how to explain how slick I think this is.
/wiz give me a guided photographic low-angle wide-format tour of the setting, using code interpreter to analyze NPCs and their scenarios between image generations to extract the data, then GPT-4 to inflate a description from the semantic context, improvisationally, world-building from knowledge. Generate 5-8 images along the guided, travelogue-style tour. Monochrome palette with moody lighting featuring NPCs and Archvillains please.
I totally remembered wide this time. Some of the following will go in the https://discord.com/channels/974519864045756446/979617962728226826.
Discord is the easiest way to communicate over voice, video, and text. Chat, hang out, and stay close with your friends and communities.
dang it, now that I am ready to have fun, I hit the 3 hour cap
nooooo
but that coffee did good
oh my gosh i am so sorry
no worries, I will get to it
I'm at that time of day when you overlap usage and the reset happens later because of that
right on, thanks!
i'm seriously considering adding a teams account
not updating, adding lol
reset happens usually at 3AM, but if you are using it at that time the overlap mehanics are different
don't don't, like I said I'm now awake and better do something than be lethargic until I have to shower
if you want i can literally link the knowledge files. you could use copilot designer if you're still subbed.
it's usually the first week when I change drastically my schedule that happens, I was away for 1 year from work due sabbatical, so now that I'm back it's normal
no, don't spoil the experience
my boss when I was internet in QA always wanted to do that, give me the documentation to read it before, I always neglected that and had more success finding problems
if you know too much about a system, you are biased already, hence no help
it's how i used to work with patients when I was practicing
it seems to work, testing it again
if you want to know a patient, you first have to know them before reading the file, see your first instinct and gut feeling
the thing is, lots of people in psychiatry have problems already, so they will try to shovel all in 15 minutes and if you know the specifics, you can't outline what it's important in the now and here
different in the somatic health diagnostics
in that case I better hope my doctor reads my file first
want to go to the jaz club for some cig sigs?
anyway, any wishes for the story, I can include a suitable symbol incorporated to my prompts
jazz sounds lovely
yeah, 100%
you can tailor the /wiz prompt how you want
i would really like to tell you one word it likes, that you might not pull even from documentation
perfect, I will disregard responsability tho, I'm not peter parker
okie, but no more freebies plz
no
1h07 minutes
i'm done
i just thought that one is nebulous and abstract enough that
you can find fun ways to leverage - or disregard!
I'm looking at some youtube videos with funny colorful stuff, doing so many film noir images got me drained
so when the limiter is done I can go back to do that stuff
hell yeah
you know what dys
i'm going to post from another game here for your review
it's called "The Chroma Clash" and it's another TTRPG, very abstract
very colorful
ok, but I will do it after the one you showed me, flaw of my workflows
okie, that helps
i mean if you're interested in more art from DALL-E that's bright
i got this in spades!
anything goes at 4AM
except problems, those are only allowed from 8AM to 5PM, monday through friday
speaking of
i got some failed image gens from modifying that prompt
so i simplified it
first output
the ones in the middle looks like possible super-villains
yeah i told it to pick arch-villains too
but it's a really vibrant setting as you can see
oh nice, reset happend soon and I got some news for you
that's what i mean
i had that happen so i shortened the prompt
new one generated first try
want to see what i did in chroma clash?
oh I simplified a lot
oh lol
ruh roh
technical issues suck
i think that might be what i was hitting too but it's hard to say
20 words prompt
yeah that's probably on dall-e then, that sucks
ok, adapted the prompt and it seems to be working
yay!
but only for one image
boo
the "do more images" part didn't work
||/wiz a travelogue tour of the setting in wide format images, HD Neon and LED Pointillism style, featuring NPCs and Archvillains from the setting, as well as key points. 5-8 images please.||
๐
yes, I got the 5-8 images part
weird
the prompt was /wiz give guided tour in wide format of ths story of an anime magical girl, and how she got her powers. All in rather darker tones for more dramatic effect. The story should be in 5-8 images.
Ok, I got a problem but that's not the GPT problem but rather rendering quality that I can't pin point, but now it's working
i get those too
the art style was inconsistent though
i thumbs down every single one
yeah it gets creative due to the system prompt
i kind of don't always mind if i get a bunch of images anyway but sometimes it's a giant pain
ok you got some pretty sweet stuff from it
yeah
i'm actually really impressed
because while it's not not an anime setting
it isn't one either
that's pretty sly
I do have to point out something, the story telling factor is pretty much one of my to go concept brainstorming, so that works out of the box with normal gpt. My question in this regard is: what's different for the custom gpt?
it has tons of knowledge about the game rules and setting and they're encoded with SPR summarization and decoded similarly
so each game is improvisationally unique
I see
with a fresh take on the story etc
ok, that's probably more obvious in non-wiz mode
ok, I will play around with non-wiz then, after I get a sandwich
hell yes
Also have to find a way to get rid of this abominations... it did two of them in that quality...