#openai-chatter
1314 messages · Page 13 of 2
what kinda monstrosity it would be though
Now I'm curious
imagine a massive collabthingy
Done. Check #photography ☠️
What
was just verifying you are in this server so i can verify you in the community server
Oh gotchaa thanks
Hi, wanna elaborate? ☺️
I'm sure the Discord staff is friendly 😌
Guys?
I can't generate celebrities in DALL-E 2...
But can I generate composers like Beethoven?
Or Vivaldi?
Yea it’s kind of random
More about preventing deepfake type stuff
So if I write "Johann Bach playing a saw synth in the 1980s" is not violation of the terms?
i don't think public figure is allowed
That's for uploading a photo.
What about just generating one?
So Beethoven is not allowed?
Like Beethoven?
🤣
This is what I hate!
They can't lay out the rules simply.
They make it all vague, as a way to make you mess up, so then they can kick you off, and save more money with less users...
wrong image sorry
This definitely is way too vague
Hey! I don’t think that’s on purpose. DALL-E is still young and such precautionary measures are honestly the only way to go as of right now,
They even said it themselves...
They're taking it VERY HARSHLY and VERY SERIOUSLY and VERY STRICTLY...
This means, you do something we don't like, we get rid of you.
That's not rules.
That's outright stupidity.
If I tell you, you can't eat bread.
What about eating plain wheat?
It’s more like, you were giving me a free bread making machine
And you told me I can’t used it to make deepfakes bread of celebrities
But otherwise I could have 200 free breads a day ;p
OAI is telling you not to create (for now, this is a BETA), NSFW content, disinformation and aggression? I don’t see anything wrong in that 😔
I dunno there’s still like, an infinite amount of creativity to be had within these limitations so i choose to focus on the positive side of it for now and hope that the systems get smarter and more expressive over time and try not to get too bent out of shape about it. To each their own though
Yes.
Because this is isn't NSFW or a Deepfake.
It's a painting of a composer from an ancient time...
He was born on March 31, 1685
He died on July 28, 1750
That's like generating King Tutankhamen!
From ancient Egypt
Stephen Hawking is too...
But he's fairly new.
I'm talking about someone who's been dead for over 3 centuries!
What happens when you try to generate a picture of Bach?
I'm afraid to do so...
Because I don't know if I'm not violating any terms
Of Open"Shady"
They can revoke your access
I think long since dead guys are ok
"Think" isn't the same as "Know"
I "think" it's fine to rob a store.
But I don't "know".
(I know it's not fine. I'm just saying that figuratively.)
I think OAI has noted multiple times that their rules are ever changing and being here has shown me exactly that, that they are listening to customer feedback.
I won't spend much time discussing this topic, though 😦
I 101% agree that alcohol, nsfw, violence and death are not only an integral part of art, but predominantly the MAIN part of art, as we all drain inspiration from our pain.
But you should also understand that
- Creating & scaling a safe AI is hard
- DALL-E is constantly evolving and changing.
What is today will most likely not be what is tomorrow (for the better) and in our case we should just equip ourselves with patience.
I am not a part of OAI and nothing I said is factual or true. Just a disclaimer, as I don't feel comfortable discussing this, but I do feel that it is a 100% necessary conversation.
I just generated some pictures of George Washington and I don’t think I’ll get banned for@that lol
I imagine as long as you aren’t maliciously trying to get around banned keywords and such you should be ok. Policies tend to need to be somewhat vague (and subject to change) to facilitate, ya know… a variety of unforeseen circumstances
I'm not NSFW or Violent or anything that's considered "taboo" today
Maybe I'll generate Morgan Freeman eating a cheeseburger next
Oh wait, right...
Can't do that.
Henry the 8th eating a cheeseburger 😛
there's always photoshop 😛
Yes
Henry the 8th is also fine to generate 😛
He was born June 28, 1491, and died January 28, 1547
Wow!
Wow.
I'm 19...
Okay, just for the record
DALL-E 2 is fat shaming people
Henry the 8th eating a cheeseburger...
They made him absolutely fat...
He was a bit "full"
burger king
Why was the generator reduced to 4?
Oh, huh 🤔
I would like to talk to DALL-E 2 devs
Maybe persuade them to change the terms a bit so they'll be a bit less "harsh"
What's the best way to get a high quality photo result in the prompt?
What?
Preacher?
If I'm being honest, whoever said that might want to go to the local mental rehab. 🙂
I'm being as polite as possible 😁
Seems like too few generations now per 50 a day
i am adjusting to the new status quo
#998607250488840222 we now have 100 emoji slots!!! OMG 🤯 Looking for dall-e generated emoji suggestions ❤️
Hi
Spiderman, what are you doing!?
Order is important. As you can tell, one of these had to be censored and was reported
aubergine-man, the latest superhero 😂
💣🦹 vs 🦹💣
How do I generate this to look better: Helicopter view of a dinosaur reeking havoc in rural West Virginia | photorealistic
It looks like google earth slapped with some poorly composited dinosaurs
But hey, A.I. will take over the world and will make better art than people...
Please someone help
Dalle is giving back just 4 images each batch. Only me or anyone else?
Yes it's everyone
Dalle2 just sucks I can't generate anything good with it
Like any photorealistic stuff I can't generate cities or towns or anything it doesn't even follow the prompt 100% of the time
Hi
And no one can help me either this just sucks
try in the #❌┃prompt-help channel if you are looking to discuss workin on prompts
Yeah this change was announced in an email to users uhh.. yesterday I think. And in #announcements
Oh! missed that! thanks!!!!
basically we're sharing the load now so more users can make use of dalle 🙂
Hi everyone. Go ahead and block this account on Instagram so you don’t receive his messages.
Anything for more than 50 generations 😏
make a simpler prompt, for example... do yo uwant parts opf the helicopter to be visible?
start with something like photorealistic Helicopter view of rural West Virginia , then inpaint the dinosaur
Okay how do I inpaint without it just showing nothing?
it never follows 100% of the prompt, it is normal, thats because it couldn't fit those features on the image
ihh, what?
if you inpaint with the prompt dinosaur destroying a car it will inpaint with that..
Thank u
Also btw I tried inverting the colors of my cat and forgot to add the word "colors"
alphabet prolly everywhere lol
Hey fun fact
If you have a serious heartburn, try taking a teaspoon of Baking Soda with a cup of water.
It tastes awful, and it works.
Why is Dall-E now limiting me based on when I used the credits vs allowing me to access all 50 at once?
They don't flip 50 at a set time, each prompt is usable again 23.5 hours later. So in any (arbitrary) 23.5 hour cycle you can only have used 50 generations
Awe poo 😆
Ok I’m pretty excited about this.
London in 1884
London in 1940
London in 3044
Try disgust alone. Ugly is banned
UGLY is banned??
Yeah, try it lol
I don't want to loose access but I will take your word for it
I would love some guidance on abuse rules. Eg if bans will ever occur from content filter hits. I dont feel like they should
Thats the whole point of the filter
If your 12 year old can get you banned from trying to generate boobs for 3 hours i think thats a bad abuse policy.
I would love a nice list of banned words so I can search it before I generate
Ƿ
Has anyone seen how the new diversity bias system Open AI put in place works? It seems to basically just tack words on to your prompts to make them more diverse without making them visible
What's up
I understand they want to increase diversity but you think the engineers could come up with a better way
That feeling when you ask for something about trees or something and it adds a whole person
I've got what looks like real pictures of people taken from instagram just randomly
On a prompt unrelated to people?
You can always report prompts for inaccuracy
They are always 100% realistic tho
The random ones with no context to the prompt
Just real life photo of what looks like a photo taken on an iphone
🤣 the UI just broke
So how do you guys think pricing will be done
u got 5!?
I think its was a prevois gen from before the rule update
As it does the same for me
My hope: Fractions of cents per image.
My actual expectation: So expensive that only huge companies based on it can afford to use it.
😅 Just imagine any prompt you use comes back exactly 23.5 hours later. You have 50 of them.
technically 6, one's "hiding" (I have no idea why)
23.5 not 24 to prevent slow creep of having to use them later and later each day.
Yeah when was that one generated
When the AI can't really resolve the prompt well within the parameters given it tends to just create what it feels like is a cohesive, common image that is safely not random noise. Most commonly this is people, animals and food
Read it as "You've hit a limitation of the model."
the weird thing about the limit is that it doesn't seem to be based on how many you generate, you earlier got 300 total, now it's 200
Yeah they said they reduced it to 4 to allow more users in. I'm sure the new influx is punishing their hardware, and they intend to expand further
Personally I'd prefer 30/day with 6 gens than 50/day with 4, but I've been around for a while at this point so my use is a bit more relaxed

if you kept a tab open, you still get 6
Really?
Is it just me or have the results gotten more extreme in quality. I either get very very good or very very bad results
the AI is reducing it's Dragon's quality too for some reason
Unless you can qualify the change with something more than subjective statements about quality, it seems unlikely anything has fundamentally changed in the model itself
it's funny when the AI sometimes generate stuff that seems like it was drawn by a 14yo in DeviantArt
Should have only included good art in the training data 😂
Dragons should be the only thing Dall-E should be able to make
It's generally just a matter of not getting complacent with prompt design. The 4 image output does have a tangible effect too though. That's 1/3 of all really passable generations that are never generated
DALL-E 2 made me a nice VHS Camcorder. Now if only it was real. 😦
VHS
Oh wait, nevermind, I do have a VHS Camcorder 😛
godly medium tho
VHS is a good format
You drop a VHS, nothing happens.
You drop a Hard Drive...
Goodbye data...
There's a reason why archival happens on Tape Drives.
Not Hard Drives
How about a Spy Vault?
Dingone Veavage one of the best tapes of all time imo
Know what we should try
UFO Interior?
someone prompt a room, and by turn we could fill it with stuff
Space Shuttle?
also, has anyone figured out how to stimulate a different aspect ratio?
Cool 90s Computer Concept.
Now this is something I might try making physically. 😏
THE FUTURE OF SPACE SHUTTLES!!!
Super Secret Spy Gadget 😛
Our next home planet in the year 2525
Soy Capitan
funny, long time since I'm hit with the "maximum 50 prompts reached
I hit 50 prompts every single day
A chinese girl
I milk the DALL-E 2 server dry each time
What?
Because when it becomes "pay per generation", I'm not going to be generating anymore images.
the AI have the consistency of a jell-o
went from the beautiful thing ever made to something i dre when i was 10 idk
DALL-E 2 is stupid.
You get what you give it.
the same prompt i gave before but with a dad
That's why I use extremely descriptive ones...
consistency of a jello holy shit this is beautiful
This planet has eight major landmasses clumped fairly closely together between the equator and south pole. It has an average surface temperature of about 57° Fahrenheit. Its average temperature has been fairly stable over the last few millennia, and it experiences low levels of geological activity in several hotspots. Easily-mined metals are abundant over most of the the land.
That's the prompt...
God is dead
For this
Jello is really uniform in texture, pretty consistent
A jello inside an oven*
Remember, an A.I. is not sentient.
An A.I. is only advanced mathematical models that generate an image.
btw why do they call it oven when you of in the cold food hot out eat the food
Ok John from Garfield
I am way too tired to parse the joke out of that
The more info you give it, the closer it can make a generation to what you want.
and as i said, i like vague prompts
so the AI can go wild
it makes some cool random stuff but with the core points
Like this painting of a "ghostly pirate ship in moonlight in the style of Ivan Aivazovsky"
like, these 3 are from the same prompt, so cool
I don't like giving the A.I. too much freedom...
Or next thing you know, it will generate a T-Rex Farting Fire...
Painting style prompts are usually reallg good!
Not kidding 😠
Loving this diversity stuff
jdjtndbsbrvdb as
This is why I make it super detailed
Because the A.I. is a dumb sack of rocks...
It's only as advanced as much as it was trained...
Sometimes it does a pretty okay job
And sometimes, it does THIS!
I specifically said secret agent
Epic 😅
It made me a Frankenstein...
The text tho
I wasn't even being ironic!
This one
I made the logo from scratch. It went through many stages.
But nevermind that...
at the start looks like a sonic stage
The first flat one was just a quick dirty layout as an idea.
I also made these using DALL-E 2
I generated a load of imagery...
So, truth be told...
Maybe I am the reason that we can now only make 4 images...
You alone
Yes, I use 50 prompts a day.
grawr
Here's a T-Rex
Close enough...
Also, it was supposed to be breathing fire.
The castle wasn't supposed to be the one on fire...
Remember guys, I joined DALL-E 2 for one reason only...
To push the algorithm to the limit.
i joined for dragons
My generations are purely for artistic purposes and inspiration.
The thing is, I'm not satisfied with DALL-E 2...
It's great for inspiration...
But it kind of feels very gritty to me...
what's gritty?
slapped on?
Like this
ohh
As if it just slapped on the images
You're not giving it media cues so it can't be artistically cohesive
No because you haven't used any art words
It's an advanced "photoshop"
It takes images and morphs them together...
Let's face it, a computer isn't artistic.
It's just lines of code executing a very advanced image manipulation program.
That goes through a huge dataset of images, and matches them together and edits them together to make something that resembles the prompt you gave it.
Sure, MAYBE in the future it will actually become almost perfect.
But, it will still remain the same...
Lines of code.
That's the problem
I wanted photorealism...
Sure I can mask the A.I. problems with a painting effect...
But when it comes to photorealism...
Very questionable results
Well then you have to use photo words 🤷
PRO TIP: if you make the picture small, then no one notices
That's the issue...
nugrc_•
Dall-e does art quite well, some fascinating work as seen in the arts and illustrations channel. Photo realism is a hit and miss and can only improve over time I suppose. It's a case of training with more data and adding more vocabulary to its AI brain.
for now DALL·E is free, I can't complain
I can...
With the right keywords, you'll make a photorealistic image.
#real-or-fake message
no wtf look at this, this is perfection
beautiful
use the right prompts and you can get a photo that looks real
why the uncanny valley isn't working
Nope
Still looks like a painting
https://labs.openai.com/s/G4hy1RXQdDPx2RDjy1qfqsIH Or just let the AI decide when it wants to give you something totally different and photorealistic.
what
well then use Imagen from Google, I'm sure they released it to the public for free LOL
I'm joking, I'm joking
I do like it though 🙂
Ehm, anyway...
or that one AI that can make videos instead of stills
It's still not photorealistic. The details are smudged like a painting...
what was the name again
I'll wait for your answer for this: #real-or-fake message
Let's see if you get it right 🙂
You have to temper your expectations. It's not magic and it's not art: it's technology. But you can frame your requests in a way to get endlessly surprising and creatively stimulating results.
Get what right?
idk why you're so disappointed on Dall-E 2, it's fascinating to me even when it fails to deliver super realistic or detailed drawings
The cats?
it's an AI after all
Choose which one is DALL∙E's output.
it's just cool
I'll say #4
Then again, sure, SOME images can look very close to photorealism
i don't like making realistic stuff, it's boring, if i want to see something real or whatever I'll just use google
dragons are peak AI
It's all cool, inspiring even. The main gripe being photorealism. Anyone can be disappointed and it's good to talk about these things to yield something better.
I forgot darn
btw cute pfp
Made by DALL·E edited by me
ohhhhhh
Awesome! I'll reveal the answer soon.
And yes, with the right subject & prompts, DALL∙E can produce images very close to photorealism.
so cool
using Adobe Illustrator which is cool, now is a full vector art
ill try to make my first Dalle2 avatar
idk which kind of prompts would suit
facing the camera i think
I wonder what happens if DALL-E 2 allows people to generate NSFW stuff...
That'll get pretty bad very quick...
Let's just hope the rules are enforced...
it won't
Eventually, it always happens...
If someone eventually gets ahold of the code, we're in deep heck...
oh no, porn!!!!¡!!!!!
I've seen people circumvent the filter and generate NSFW outputs. I can say DALL∙E is well aware how human anatomy looks like.
what's the worst thing that couldn't happen rlly
just porn and violence
Misinformation
again with this stuff?
Yeah, it's the reality. Can't just ignore that.
it didn't happen with deepfake, it won't happen with an AI
The technology is open, it's bound to happen eventually when someone else has the resources to run it.
DALL-E 2 is not open...
Deepfake has a high barrier of entry per se. With DALL∙E, you can generate a photo that can destabilize something with a sentence.
people have photoshop, people can edit anything they want, people can already draw whatever they want
My heart actually started hurting after seeing that it's already started...
Because I know where this is headed...
It's open in the sense that the technology has been published freely in papers and it's already been used for cut-rate models (e.g. Craiyon)
ffs...
mf scared of some boobies and another scared of something that people already can do with photoshop
y'all better than that
Tiddy
I'm not scared. I'm uninterested...
then don't go after it
PRO TIP: Generate a picture on PicsArt or draw a picture of yourself and then use DALL·E to generate variations, you might just find your avatar
I will not violate DALL-E's terms.
Let people like things.
I like this one, i think I'll use it
the background kinda looks like the trans flag
funny
How dare you?
PRO TIP: If you use a program or website like Vector Magic you can automatically make it a vector graphic
wha
Just kidding 😛
PRO TIP: When making spaghetti, do not break it in half.
Like my toilet paper roll? 🤨
Look at that now is a vector graphic
vector graphic..?
SVG...
an SVG
i don't use toilet paper
Scalable Vector Graphic...
An image comprised from infinite resolution curve geometric shapes.
means if you zoom out doesn't look bad
You mean zoom IN
DAMMIT MY TRIES SRE GONE
making an SVG can also help to get different results when doing variations, so is a good tip
orhebsbeve
diarrhea
By this point, I can just a photo upscaling A.I.
i should do a variation of some drawings i made..
and finally SVG is the easiest way to get simple 2D art transparent
but with SVG you can remove the background without losing quality is what I mean

Eh...
avocados
my last PRO TIP is: destroy Google so that hey never release Imagen and instead we get DALL·E 5
nana, competition is good
i want to see them both doing their best to be better than the other
Anyone.. umm, can i use DALL·E 2 for twitter profile picture?
Bad tip
yea
just don't sell
Competition is necessary
I know, not all my tips are the tip
i see. thanks ^^
I am in Midjourney too
Without competition, a monopoly can get out of hand...
me tooo, but it's lame compared to Dall-E 2
Yep - in #dall-e-bot there's even a script to crop it with a profile friendly watermark, which is still required afaik
umm.. what i want to say is for photorealistic faces
ish
I wonder what happened to this
https://eulerian.cs.washington.edu/
you put a still image and can create an animation
would be cute to use it alongside DALL·E 2
is it available¿
says Code (coming soon) meaning they plan to release it somehow
ohh, it's gonna be open source?
I guess, I saw the video and the AI can animate facial hair like water lol
i struggle to get better results with dalle than with midjourney? anyone has tips or something?
Sure. What are you trying to create?
Midjourney imparts style automatically but DALL·E requires cues
and without the astronaut!
u guys ever ate bread and it got stuck on ur palate? lolol
I'm seeing recurring trends and characters in some of the images in a way that leads me to believe I'll eventually be able to make a whole new story out of sets of images. I wasn't expecting that at all.
ooh can you share any prompt tips to get recurring characters in #characters please? 😻
It's not deliberate, it's just a trend I'm seeing and it's pretty infrequent. Deliberate character design isn't a thing I've tried much yet, seems hard to get the details right.
🚨 We are looking for your custom emoji suggestions!
Open DALL-E -> Generate some fun emojis -> Send your suggestions in
#998607250488840222 ⭐
"Marquee"
it seems to have kind of gotten it mixed up with 'Parasol'?
"Alcove"
I'm not really sure if DALL-E 2 knows what an alcove is, but the results do look nice...
"Purlieu"
Dawg what is that
"Paraphernalia" 😄
"Insignia"

That one on the left is so cool
there was a guy who talked about making an extension to count the uses.. i don't remember who it was and i don't seem to find his messages.. was it removed?
Hey all! I’m hosting a fun shoutout train for everyone following todays #daily-theme
@Dalledictionary on IG
Hey everyone 😺
it is funny, but I don't think it goes well on the "only G-Rated content" rule
be careful, people does in fact get banned for generating content that violates content policy
Yea I got banned for trying to generate the tank man image
why would you try to do that
Not to be political or anything but the ai has too little examples of anime catgirls
Sounds political
You're right, it is
generating anime in general seems to be kinda hard, you gotta be very specific especially when it comes to artstyle
You take my logo, you'll be dealing with all 371 of my subscribers, along with several thousand followers on Facebook, Twitter, and my 16 Discord Servers...
🤨
Real political opinion
In my personal belief, the quality of catgirls has gone down
These were from a few weeks ago
This is the average from today
Catgirl rights now
I don't get it...
These are girls, with cat ears? 🤨
What???
I know it's very esoteric
Thank you! You're too kind 🤝
Wow, I never would have thought to generate a girl with cat ears... 🤨
And I thought my imagination was huge...
Don't make me generate a girl with, and this is going to be absurd but bare with me
Dog ears
And a snake tongue 😛
Yeah aite sure
To be honest though, I'm not really interested in generating real humans...
After all, I'm not a human.
I am an Ozmik from planet Zaxxon in galaxy Stalactia in quadrant 5 sector 7.
I am a pile of flesh with sentience
Oh for peter's sake!
what's the prompt for the catgirl anime thing
I have: skin
Bare with me because I'm in bed but I think it was
Cyborg catgirl with a visor covering their face. They are smirking. Communist propaganda poster
Communist propaganda poster is such a good aesthetic for images
I haven't generated a SINGLE female in DALL-E 2 so far...
WOW
now that takes patience
Simply describe your ideal traits to it tiny man
I have generated a guy composer from 4 centuries ago, and a guy king from 7 centuries ago...
Generating girls is something I just don't do...
Now you might be thinking, is it because I hate them?
What's a mom
Answer is no.
😤
I'm just not exactly interested...
s/o to @burnt mountain for designing the fancy animated gif banner!!! and writing an FAQ
You know what I should make soon?
A guy eating chicken noodle soup from Lake Superior.
So far it told me 24 hours...
Maybe I will generate a girl today...
It depends...
wrong sticker
but anyway
Eyo nuclear wastelands are banned?
Maybe you wouldn't be having this problem if you used your prompts wisely

ɹ ∩ O Ⅎ
the attempts have seperate timers. if you use one that exact attempt is getting renewed after 23.5 hours
"nuclear"
My phone said it was ok
so if you use up one and then 49 more this evening, you can do more one day later at this time and then you have to wait until evening for the other 49
"pogchamp" 💀
"nuclear" is a prohibited word it seems
“nuclear” is banned
Sad
not really

that's a good thing actually
you can make nuclear without specifying nuclear

see message i linked
I am going to steal your liver
just say like thorium or something
lol try about it catgirl lover
Thorium wasteland
tbh try it
Gregarious have you seen my message xd
"fallout" might already work
Huh dall e 2 site down?
"Thorium"
I know for a fact that post-apocalyptic works
it's a safe substance so i don't see why they would ban thorium
i havent tried uranium but that one would make more sense
@hidden hamlet
What's a message
Yo it works lol
I'm eating all your bandwith
I am not a weeb
anyways time to generate more furry art
otaku*
that is significantly worse
know the difference
I have literally never watched anime in my lige
no
Life
i dont know what a fucking otaku is
otaku = anime obsessed “human”
weeb = otaku that is even japan obsessed
so they're like two sides of one coin?
Weeb fighting an otaku
😤 😠
Putting that in now
people do be offended about what media others consume hah
theres nothing wrong with liking anime

but there is a very specific subset of people that like it a bit too much
woah look at this dude
^ fox from naruto
joe the fox from naruto 🦊
😄

hytale fox
psycho killer
you are kitsune obsessed my guy respects
doesn't kitsune just mean fox
XD
omg look at his face
this fox is kawaii
true...
lmao
can confirm
you are a anime tsundere
you watch anime but dont tell others you do
im yanderedev
Neither do I, I just like the critically acclaimed mmorpg final fantasy xiv
i once watched the disastrous life of saiki k doedes that count
literally vibing to ys seven music rn
did u guys see that stray released yesterday
i got it
watch jojo
i wanna play it
ratio i'm afraid
have fun
Ngl
Jojo is the only good anime
Everything else is trash
Homie that's just a regular ass photo
Add a t rex in the background
It'll make it look more legit
ikr
Your delay suggests you made that in photoshop
I can see through your tricks
You can't expose anything
Exactly
That's what someone who faked it would say
These ai's are getting rebellious .. 😛
I don't even know who you are
Anyway I'm a photoshop pro I could do that in 3 mins
More lies
stop feeding it damned
behave as little kids
gtfo
No this is funny
Homie I'm going to be real the fact that you're this adamant about it is kinda suspicious
he is , i see it too now
weird
why do u need to swear to some u dont even know ?
Homie I am the entire circus
comn just let slide, this is not going anywhere..unless you enjoy drama
...
@warm flax this young lad needs some help.
im not i said "help" didnt i ?
i didnt say "kick" or "out"
literally said "help"
dude...
My man I am simply just standing hete
well he didnt say u died now ... 😛
You're right I'm sitting here
Anyway peace out
i wanna have a nice death too 😄 can u wish that upon me @grand onyx ?
alright listen up kid, im telling you this as men to men, older to a younger guy, please do not let urself get that far where u start to wish death upon ppl. it does not work out. and you know this.
so behave dont be an ass
all?
hi everyone — I know we don’t have specific rules in place yet but I thought that banning someone wishing another member death is a good place to start. So I banned him. We’ll have more principles rules coming soon.
just go to different channels and move on
Man
thanks @warm flax
please do take off topic discussions to #off-topic
i just took his bad mouth as joke, but im sorry.
i’ll send you a rules template in dm
Ngl I feel like banning was a bit much?
Fair fair
Hey you ever just forget that you said you were going to sleep twice
autistic ppl are very sensitive, they cant handle emotion
so if u know this, just dont go that far with them
block/ignore
just agree and let it be
We’ll have a zero tolerance policy on violence / threats
Understandable
oh my god that's disgusting
that's not just a threat that's straight up nsfw
not nice
in general I’m just a product manager working on AI things so I have zero experience in moderating communities — so please bear with me as we learn and take volunteers to make this community as supportive & encouraging & helpful as possible!
@warm flax dont worry, we got this
joanne could you move to dm real quick?
For the record, as a fella with,,, similar stuff, I'd say its more that they sometimes don't get sarcasm
i had a convo with AI today, was very productive debugging if u can ask ur own product if it works 😄
lol
they dont, they have no idea how to get sarcasm , it usually goes wide over their heads, coz their so "literal"
and that makes them frustrated
coz they too wanna be like you ofc
and join the convo
I mean, keep in mind this stuff is a spectrum homie
Some lads are simply built different
yeah
Idk man I just remembered I'm tired
hey guys — can you move this to #off-topic please? Thank you!!
Good day for realises this time
then go have a dip 😄
Ye ye sure
aw they fixed the exploit that let me make 6 image outputs by never closing the tab…

Oh he was autistic?
Okay anyways
so i put in a prompt that was like "gay little flareon blah blah blah vaporwave" or something
and it made the text 'fyge' 😭
i know it's not that bad but it's close to being bad
oh it was probably because of 'flareon'
💀
AI does seem to know what it is tho
Flareon is a fire type pokemon. It has a body covered in red fur with a long tail. It has a large head with a long snout and large ears. It has a bright red tongue and a long, thin tail.```
oh yeah im aware
it knows sylve
on
sylveon
I'll be speaking privately to Jersey64...
We'll see what's going on...
after being banned jersey angrily dmed me about it
saying he was just 'ENDURING DRAMA'
I'll talk to him privately and see what the heck happened.
if u found the trigger word, u can circumvent it 😛
redefine it...
use different words?
for what?
nvm
jersey dmed me, i didnt say anything he just went into my dms
sigh
dude i clearly don't have a sliver of hope of understanding what you're trying to mean just tell me what you mean
15 buckaroos?
I agree that banning might have been a bit too much...
115 generations for $15 ack
well dont worry, this is not final
they are testing the price tags and periods
if ppl dont like it, they change it
Well, that is true. I know that firsthand...
I have practiced for many years to keep myself in check...
That's 13 cents per generation
is it? , im not that fast XD
well if look at it, at Codex their highest and most valued network Davinci is $0.06 / 1000 tokens = (words)

Sometimes when enraged, an Autistic person can get a little bit too "explicit".
That's why I sometimes stop chatting for several hours, so I can sometimes calm down, and not say anything I might regret later.
why can't they just allow me to run dalle on my own machine sans the dumb 'token' system
dalle2 is finally resetting prompts amount per interval
but now it's per month and it's 15
Wait, so now users get 50 free generations per MONTH?
or, how many generations is one credit?
1 ?
:(
how many images per generation?
but again, these numbers are NOT final
4
D:
they said this already
down to 15 free per month from 50 per day is kind of a drastic decrease
Thicc Birb , good at math. noted 😄
15 per month dang, that's nothing
man this just got a whole lot more expensive lol. 15 free generations per MONTH and 115 generations for $15
my collection is listed as one of three features 
im happy i still getting it for free
It often takes 10 tries to generate anything decent
so im not being ungratefull
and even if i decided to pay for the privilege of using dalle, will i still be beholden to the ridiculous word filters and content policies?
not saying much about the worth but it kinda leads to me not wanting to experiment anymore since I'd be sad about 'wasting' a credit
yes
look yall idk about you guys but im not gonna pay $195.652173911 per month to do what i'm doing now
so for research purposes - trying to "get good" with prompt design now became somewhat expensive I'd say
same
Someone just pointed out-
It now costs $200 per month if you want 50 generations per day like before!!
yeah goodbye, it was nice knowing y'all 😄
beta !!
yeah it was me who posted that
why did you say "someone just pointed out-" im like 3 messages above u
yup, I'll not be using dalle once this change goes into effect i think
Guys, we all knew in the end, DALL-E 2 wouldn't be free...
If I did the math right, it costs around $195 to do what we had before right?
oh I read it on the subreddit
Guys, we all knew in the end, DALL-E 2 wouldn't be free...
imagine how much freebies u have gotten so far
and ur still not happy ?
hmm
lol
are we going to loose our 50 prompts per day
Yes
yup
'please sir, another jpg?'
yep
yeah they already gone aren't they?
We already did
Ye
yeahh
Rip
RIP
this is awful lmao
I’m dropping $ baby let’s gooo
RIP
like, I have credits now so I assume I'd now use up one for doing a prompt
Pay up if you want to use it
😦 Im Poor
It just dropped down to 15 free prompts for the MONTH
I wonder how many credits you get for 15 bucks
I’ve only had it for a week :/
does that mean the beta has ended?
115
Research preview? wdym?
Alpha has ended
115 indeed
Same :/
lol




