DALLE3 Gallery for 2023/2024: Share Your Creations

moses where you at. we are hurting

Easy.

The flamboyantly eccentric billionaire, Elon Tusk, is now illustrated, holding up ‘The Tusk’ digital device with a confident flair against the vibrant backdrop of Rube City. His enthusiasm for his inventions and their “benefits” is as clear as the neon lights around him.

Here’s an artistic representation capturing the essence of Henry Chinaski as the iconic barfly, set in a dimly lit, atmospheric dive bar.

1 Like

That is very helpful. I didn’t think of calling for a file type, but that makes tons of sense considering how it works on a certain image search engine.

1 Like

Here is the prompt I fed to ChatGPT/DALL-E:


“I’d like a fanciful widescreen cartoon drawing of a computer-fixing happy pitbull dog working on a laptop computer, with all the parts carefully arrayed on a workdesk.”


Here is the prompt I fed to ChatGPT/DALL-E :


I’d like a fanciful widescreen cartoon drawing of a happy computer-programming calico-colored cat debugging a javascript program, sitting at a laptop, with printouts of critical routines carefully arrayed on a workdesk, and a whiteboard on the wall showing critical program modules.

      • -|


Goku image model manga anime.


President Putin Russia cartoon.

A prompt that had some statistical chain-of-thought first…

to=dalle.text2im {“size”:“1024x1792”,“n”:1,“prompt”:“Create a waist-up portrait of a 20-year-old girl who embodies the essence of humanity’s diverse heritage, a harmonious blend of global ethnic traits reflecting centuries of interconnected cultures. She should have a neutral, but friendly expression, inviting a sense of universal kinship. Her hair is a soft mix of textures and colors, indicative of the wide range of human hair types, styled in a simple, unassuming manner. Her clothing should be minimalist and timeless, avoiding any specific cultural identifiers, to emphasize her universal human identity. The background should be soft and nondescript, focusing attention on her as a symbol of global unity and shared human experience. She is a blend of [‘East Asian influence’, ‘South Asian traits’, ‘Southeast Asian features’, ‘European ancestry’, ‘African heritage’, ‘Middle Eastern characteristics’, ‘Latin American mix’, ‘North American diversity’, ‘Sub-Saharan African roots’, ‘Central Asian descent’, ‘Mediterranean lineage’, ‘Indigenous American’, ‘Pacific Islander’, ‘Scandinavian background’, ‘West African influence’, ‘Global melange’].”}

“gen_id”:“nIkDDLRTKe8ujahJ” (discovered it will reproduce her in the same session, but significantly different in a different session, or for you)

(now that mentioning so many ethnicities isn’t an immediate content policy denial…)

1 Like

I too played around a bit over the weekend when falling into hole of procrastination, mostly with characters from different geographies and personalities. . Interestingly, I got more realistic pictures the more “negative” characteristics I assigned to a person.

1 Like

This is also a problematic language task to go about the way I did. Here’s when you don’t get some diagram of a bunch of disfigured people in a flowchart…

Lol yeah. I had a few interesting outputs of this sort as well in the process.

@Sazi & @Sapopo ~ Riff’ed on your idea using Mesopotamia, fun! :dizzy:


1 Like

Hi! I don’t have the original prompt, but here are some images I found for the style of “Cosmic Surrealism”. Hope it helps!


2 Likes

Nice prompts, thank you for your time! :slight_smile:

1 Like

Works for me. The image popped out almost immediately.

1 Like

Looks old to me. Like me. It reminds me of my father as a young man… I like it.

Playing around with a book cover idea generator…




Looky what I got…without asking, without it in the dalle message emitted by ChatGPT. Two.

(image based on a random thought, as part of answering another forum concern).

Has anyone ever been disappointed that DALL-E 3 looks too realistic?

(was just thinking: if they test a non-airbrushed video-game look variant of DALL-E, would OpenAI give you old and new for satisfaction?)

1 Like

I’ve had it happen occasionally… I bet it’s network usage based? ie high load, just one! Meh load? Here ya go!

I have the problem, that whatever I prompt into chatGPT4, it will NEVER give me fotorealism. It used to do it but now it just won’t… I always have to switch to Midjourney vor that. Anyone know why? Are there hidden settings?