View allAll Photos Tagged Prompting

Artwork created by Midjourney from a sequence of text.

 

Artwork created by Midjourney from a sequence of text.

The Prompt gallery and nightclub site is at promptgallery.space where there are event listings for shows in AltspaceVR - also see account.altvr.com/channels/abstraction. The world code for the Prompt is YQT013

 

There are videos of the Prompt Events here www.youtube.com/playlist?list=PLfzKgAP3NhxBe0FW2bXY5Xvk8h...

The Prompt gallery and nightclub site is at promptgallery.space where there are event listings for shows in AltspaceVR - also see account.altvr.com/channels/abstraction. The world code for the Prompt is YQT013

 

There are videos of the Prompt Events here www.youtube.com/playlist?list=PLfzKgAP3NhxBe0FW2bXY5Xvk8h...

created with prompts using recraftai

Artwork created by Midjourney from a sequence of text.

 

Artwork created by Midjourney from a sequence of text.

 

I colored a template using acrylic and glitter. Once I started, it was amazing how many different treatments presented themselves. So many patterns in the simplest of designs.

Untitled (disjointed and unrelated)

2016

Single pole switches, LED lightbulbs, wood

Artwork created by Midjourney from a sequence of text.

spent the day at Horseneck beach in Westport Ma. so I'm sending some cool blue waves to those of you in a heat wave!

Artwork created by Midjourney from a sequence of text.

 

A joint effort for the photo prompt today.

After spending yesterday morning, then most of the late afternoon/early evening cutting down the front hedge, we were both tired.

This morning, we went out to get photo evidence of our labours....

It's a big pile of dead hedge, Olivia woke up immediately, "It's a trampoline!!" She doesn't often get tired!

Prompted by Sarah Kershaw to be like "angry monkeys eating each other's faces".

thread sketch on canvas painted with gouache

Le prompt, qui a été le même pour toutes les IA.

 

Génère une image au format portrait. Collage de mode automnale contemporaine : une femme à la silhouette élancée, habillée de tissus fluides aux tons chauds, mêlant textures velours, cuir patiné et fibres végétales. Composition dynamique en contre-plongée extrême, effet œil de poisson sans vignettage. L’éclairage dramatique, inspiré du cinéma, révèle des contrastes puissants et amplifie les matières, les plis du tissu et les reflets métalliques des accessoires. Atmosphère vibrante, énergique : l’ensemble fusionne mode, abstraction et art conceptuel. Décor automnal onirique profondément irréel, brume colorée, feuilles suspendues et horizons déformés. Ultra-haute résolution, détails d’une netteté exceptionnelle, textures biomimétiques irisées, précision microscopique. Luminosité constante, contraste élevé, couleurs éclatantes. Style hyperréaliste abstrait, 16K, aucune pollution visuelle, perspective fluide et immersive.

 

Explications de chatGPT sur les différences obtenues.

 

1. Le hasard (contrôlé) fait partie du jeu

Les modèles de génération d’images utilisent un ingrédient essentiel : l’aléatoire. Même avec le même prompt, une “graine” (seed) différente – soit un point de départ différent – produit une variation naturelle. Sans cette notion de hasard, toutes les images seraient des copies carbone, et ça deviendrait vite lassant.

 

2. L’IA improvise autour de ton idée

Le prompt donne des règles, mais les modèles ont une énorme liberté artistique pour interpréter les détails manquants. C’est comme dire à trois peintres “peins-moi un chat sérieux” :

l’un te donne un chat en costume cravate,

l’autre un chat avocat en pleine plaidoirie,

le troisième un chat présidentiel devant un pupitre.

Même prompt → multiples visions.

 

3. Chaque génération effectue un nouveau calcul

Chaque fois que tu renvoies le prompt, le modèle repart de zéro : il reconstruit l’image, couche après couche. Pas de recyclage d’une version précédente. On repart à la case départ, comme une pâte à crêpes fraîche.

 

4. Pas la même architecture, pas le même style

Gemini, ChatGPT, Sora (et d’autres) utilisent chacun leur propre technologie, banque d’images, entraînement, goûts, lubies et crises existentielles créatives. Résultat : trois IA, trois sensibilités artistiques.

 

️ 5. Cohérence dans la diversité

Le point clé, c’est que tu constates que les images changent mais respectent toujours le prompt. C’est le signe d’un modèle sain : créativité + fidélité au cahier des charges.

 

En résumé : c’est comme demander à un chef étoilé de refaire le même plat plusieurs jours d’affilée. Il suit la recette, mais il y aura toujours une petite variation subtile dans la sauce, la cuisson, ou l’humeur du cuisinier.

  

Prompt

Full-length cute girl, asian girl, pretty, large brown glowing eyes, black hair, long hair, black stockings::5 photo studio, happy, short sleeveless dress, legs::5 heels::6 --aspect 3:2 --version 5.2 --stylize 300 --chaos 50 --stop 100 --repeat 2

 

4x enlargement in Midjourney.

 

Went through several enlargements and enhancements in Adobe Raw, PhotoShop, and Portrait Professional.

 

Detailed in Adobe Camera Raw, Photoshop, and Portrait Professional especially eyes and lips

 

Converted to aspect 16:9 in Photoshop

 

I highly appreciate your viewing and commenting on my work.

 

Prompt three: as soon as I saw this prompt I knew exactly who my letter would be written to.......

I never felt like I did prompt 2 properly, so I revisted it. First I wrote on the page with gold marker, which was later covered with paint, collage, and the writing in black.

 

Made for Dawn Sokol's Art Journal Stimulus Project. (www.dblogala.com/)

Riley Darchicort, Sparrows Point High School, Grade 12

WU69 ENN

2019 DAF CF 480 FTG

Prompt Transport, North Common, Bristol

Buckingham, 7 October 2020

Taliban Takeover Prompts Relief, Women’s Rights Fears in Afghan Cities

 

After 20 years of devastating war, Afghans in cities far from the capital Kabul are feeling a mix of relief and dread about what awaits them under the Taliban.

 

www.breakingasia.com/360/taliban-takeover-prompts-relief-...

Day 09 - Paper // So today's prompt was to take a picture of something paper. I thought a plain white sheet of paper would be just boring as could be, so I decided to take one of my favorite Bella Sara cards off my desk and take a picture of it. They are the most beautifully illustrated trading cards I have ever seen in my life and though I'm more of a cat person, I was really drawn to this series of cards. I always wanted a horse as a little girl and toy version were as close as I ever got.

Everett Lawrence, Harford Hills Elementary School, Grade 3

Kept in Adam's old Treasure Box

Documented Life Project week 10 prompt: What makes you you? Blogged drawdaily.blogspot.com/2014/03/documented-life-prompt-wha...

 

Artwork created by Midjourney from a sequence of text.

 

Today is my birthday!

music paper, acrylics, hand cut stamp, acrylic, gesso, white gelly roll, and black sharpie

Well, the prompt is sadness. I was "stuck" on it because I couldn't find anything with the essence of sadness. I didn't want to continue with the other prompts until I created this one. I waited patiently, knowing something would be placed for the prompt.

 

Today, I saw this truck by my son's class. It's completely burnt. But obviously someone then went and constructed boxes on it for some reason, so I couldn't tell if someone is attempting to live there, or was. It wasn't there a few days ago, so it's fresh. This to me is the epitome of sadness - from whatever reason it was that the truck was burnt, to the boxes, to the garbage and fire debris strewn all over.

 

This entire thing reeks of pain and that makes my heart sad, indeed.

The project is developed using various pre-set text-to-image models, processing text prompt inputs. On the borderline between sensitive content and an easy slip into topics of violence, this project visualizes the depths of the subconscious of these models, excavating the influences of media and online information exchange. The quantified traces of reality and collective histories allow algorithms to generate content that recycles the past - building the spine of quasi-historical narratives - often obfuscated with prejudice and misinformation, along with the author’s personal bias. Generated outputs are presented inside a hypertext object – a tent that the audience can enter, as it were.

 

This exhibition is part of the category "Interface Cultures".

 

Photo: Kristina Tica

Artwork created by Midjourney from a sequence of text.

 

Artwork created by Midjourney from a sequence of text.

 

1 2 ••• 30 31 33 35 36 ••• 79 80