//=time() ?>
@CthulhuToad A mouse swordsman in the style of William Blake
[I’ve noticed Dall*E doesn’t really get Blake]
@ScottMcHale Dall-e explaining the rules to the humans.(first two)
Dall-e explaining the rules to the humans, artstation. (Last two)
#dalle2 #dalle
WILLIAM TURNER
Venezia
La Punta della Dogana e Santa Maria della Salute al crepuscolo dall'Hotel Europa, 1840
grafite, acquerello e penna e inchiostro su carta
@graceisforyou @booksandthebook @turrible_tao <Strongbad answers emails>
Sadly I don’t think Dall*E is aware who Strongbad is.
@NekroXIII Hey, my name is Dall! I do mostly furry work, but I was trained in college in more traditional works. I do both traditional work and digital work!
Rispondere ai missili con una risata. @Charlie_Hebdo_ rende omaggio ai fumettisti ucraini pubblicando le vignette di 40 disegnatori tra cui Yuriy Zhuravel, Oleg Gutsol e Kusto tratte da una mostra organizzata a Odessa dall'associazione dei fumettisti. https://t.co/XqL2O299fW
@chrisevansbrcom E i complimenti per Dodger😽🐶dove sono🤷🏻♀️😤?
Questo è quello che succede quando a parlare sono altre parti del corpo e non il cervello🧠🔨
Dodger amore per me tu sei il particolare di tutta la stanza,una dolcezza infinita,un sorriso capace di tirarmi fuori dall’abisso,io “TI AMO”
@HorniXen actually i can show you the exact ones they're based on lol. they're from dall-e mini, it's a really impressive model. I reffed really hard but i edited out the more nightmarish parts
Using Sudowrite to brainstorm ideas and then Dall-e to visualize them. When writing is more fun than watching a movie or playing video games
Google has beaten DALL·E 2.
"Imagen: Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding"
https://t.co/3H3P7c8SMb
Two friends in the VR Argentina discord server (@Elecast_, Agusfn) gave me the idea of using DALL·E 2 (#dalle) from OpenAI to generate images based on Beat Saber ranked song titles.
Each input will be: "[Song Name]", digital art
Starting with Quasar Dreams.
(🧵)
OpenAIの報告によると、文章から画像を生成できるAI「DALL-E2」はすでに300万枚(3,000,000枚)を超える画像を生成した。現在では毎週1000人のユーザーを新たに追加できるとのこと。わずか1ヶ月半の間に、この質の画像がこの速度で生成された。これがAIの強みでもある
https://t.co/RdJ9fw8kOs
האיור הזה נוצר על-ידי מחשב. לבד. מאפס. על סמך פקודת טקסט פשוטה:
״אסטרונאוט-דביבון חולם על הכוכבים כשהיקום משתקף על הזכוכית של הקסדה שלו״
כבר כמה ימים שאני בוהה באלפי תמונות ואיורים שנוצרו על-ידי מודל הבינה המלאכותית Dall-E 2 והמוח שלי מתפוצץ
>>
"The Mandalorian in the Style of Van Gogh" by DALL-E mini / @huggingface https://t.co/pNKrHs7w9L
Sometimes I feel like this is even better than dall-e 2 with art, because even if Disco Diffusion is worse with humans, it knows a lot more artists for scenarios and styles.