AI Images Generated from Text

[StrawMan]

This bunch is purely experimental. I created the text prompts and DALL-E 2 generated the images.

[StrawMan]

Still experimental. Changed the text prompts:





Definitely fun to play with DALL-E 2.

There's a shit ton of other AI image sites out there if anyone's interested in trying Text-to-AI Image apps.

[HighPlainsGrifter]

Very interesting images. Were you using straw man as the primary prompt?

[StrawMan]

The 1st two used this text: “A 3D realistic strawman on fire in a dark forest background.”

The 2nd two images, the text prompt changed to, “A 3D realistic scarecrow on fire in a dark forest background.”

For the 3rd pair the text prompt changed again to, “A 3D rendered scarecrow on the moon looking at earth rise.”

I tried generating “A 3D rendered straw man on the moon looking at earth rise.”, but they all looked like shit [see ex. below]. I'm still reading & learning how to better format the text prompts for more accurate and aesthetic results. I'm no artist. Just need a bigger buzz next time and a little more creativity...

“A 3D rendered straw man on the moon looking at earth rise.” WTF!?!

[HighPlainsGrifter]

The first one in your OP is terrific. I'd hang something like that in my house. The last one in your most recent post is simply awful.

I have not played with AI at all. No time.

[HighPlainsGrifter]

This video reminded me of your AI art project

[StrawMan]

Looks like I missed that video - it's been taken down.

Sorry I missed it... traveled this past weekend.

[StrawMan]

When good prompts go bad. Samples of an imperfect science

The text prompts for these two images were the same used to create a couple of the images in the AI OEB post. Sometimes you gotta ask yourself, "WTF was the AI thinking when it came up with THIS!!?"

Prompt: “A 3D realistic baby, dressed in white and blue, with a black patch over her left eye.”

Result 1: WTF?? Instead of a black patch over her left eye, I got a friggin' WHITE patch UNDER her RIGHT eye. It's fucking opposite day.


Result 2: WTF doesn't do this one justice. I mean, I love the Hannibal Lechter-ish feel to it, but it's NOT even close to what was asked for.

[Roach]

Very cool. My daughter was messing with text to images a bit ago. Think I'll look into it.

Cool

[StrawMan]

More experiments!

Text Prompt: A 3D rendered, short, ugly, angry lawyer, mouth wide open, in a striped business suit.








Text Prompt: A short, ugly, angry, male lawyer in a pink suit with his mouth wide open.

[HighPlainsGrifter]

More experiments!

Text Prompt: A 3D rendered, short, ugly, angry lawyer, mouth wide open, in a striped business suit.

Racketh

I see our midget pretend lawyer self-identifying as the nappy headed ho. Missing teeth and all.

[StrawMan]

Some seasonal experiments:

"A realistic, giant Santa Claus stomping on a gray, dystopian New York City."

[StrawMan]

"A realistic, giant Santa Claus stomping on a gray, dystopian New York City like Godzilla stomping Tokyo."

[StrawMan]

A couple links in case anyone wants info on getting started with AI art from text prompts:

https://www.makeuseof.com/how-to-make-c ... ed-ai-art/

https://www.makeuseof.com/how-to-create ... -art-tips/