|
Post by starskyhutch76 on Feb 5, 2023 22:51:37 GMT
It was the OpenAI chatbot. I gave it a few keywords like Vigilante, superhero, street level, and 1988/
It was very distressing in the story to learn that Bruce Springsteen had turned to crime.
|
|
|
Post by starskyhutch76 on Feb 5, 2023 23:05:29 GMT
I fed the whole story in and it gave me more
However, the victory over The Boss did not go unnoticed by the city's criminal underworld. They began to fear Vigilante and the threat he posed to their illegal operations. In response, they formed a coalition, determined to take down the masked vigilante once and for all.
One night, as Vigilante patrolled the city, he was ambushed by a group of heavily armed criminals. Despite his best efforts, he was outnumbered and overpowered, and he was captured. The criminals revealed their plan to publicly execute Vigilante as a warning to any would-be heroes who dared to challenge their authority.
But Vigilante was not so easily defeated. With his exceptional skills and quick thinking, he managed to break free of his bonds and fight his way to freedom. He made his way back to the streets, where he was greeted as a hero once again.
From that day on, Vigilante was even more determined to rid the city of its criminal elements. He redoubled his efforts, striking fear into the hearts of the city's underworld and bringing hope to its citizens. And so, the legend of Vigilante lived on, a mysterious hero who fought for justice and kept the city safe, no matter the cost.
Years passed, and the city changed. But the legend of Vigilante remained, a symbol of hope and justice for all who needed it. And wherever there was a need for a hero, wherever there was a call for justice, Vigilante was there, ready to answer the call.
|
|
|
Post by dans on Feb 5, 2023 23:40:56 GMT
well, it IS a story. Not much plot and lacking in detail. But I'll bet if you asked for a few different stories you might be able to extract pieces from several of them and shoehorn them together, and then do some heavy editing...
|
|
|
Post by dans on Feb 6, 2023 2:09:52 GMT
I do feel bad for you, learning about Bruce Springsteen turning to crime in such an abrupt way!
|
|
|
Post by dans on Feb 14, 2023 2:50:27 GMT
This is Trudy Carter. She is the steady date of Duke DuQuesne, the Volunteer. She and Duke have been attending a carnival, and Duke has won the teddy bears as prizes from the air rifle target shooting range and the 'Ring the Bell' high striker game.
This image is totally AI created based on a text prompt from me. This Trudy is not a real person, but how could one possibly know? The technology is actually somewhat scary.
|
|
|
Post by dans on May 9, 2023 16:53:11 GMT
Lib created Isis on Earth 2, and then we decided to change her name to Zephyr, so I updated her name in one of my stories. Based on Harvey's original introductory story, here is an image of what Zephyr might look like...
|
|
|
Post by dans on Jul 27, 2023 11:46:45 GMT
I was trying to enhance an AI-generated image of a superhero and the male image below came out. I like it - I think this might be Nightbird, from this story: Whiz Kid and Savant: The City of Truth.
So I figured I would try to create Flamewing as well...
If you haven't read the story, Henrietta King, code named Savante, has created a miniature city named Candor, populated by virtually microscopic artificial human-like beings, and temporarily swapped the minds of herself and John Garrick (Whiz Kid) into two of these artificial beings. They create the costumed identities of the heroic Nightbird and Flamewing before they withdraw their minds to their own bodies, and Nightbird and Flamewing become the heroic champions of the Bottle City of Candor.
|
|
|
Post by dans on Jul 27, 2023 12:11:02 GMT
This one popped up totally unexpectedly as well:
|
|
|
Post by reichsmark on Aug 19, 2023 17:38:53 GMT
A judge just ruled that AI is not copywriteable. Things just got a lot more interesting in the AI battle.
|
|
|
Post by dans on Aug 19, 2023 21:05:52 GMT
Well, I am not interested in copyrighting anything I have produced so far - but if I were, I think I would download it, delete the original on the AI site, and then claim I created it independently. I'm not sure how anyone can tell an AI image...
|
|
|
Post by starskyhutch76 on Aug 22, 2023 13:19:05 GMT
Are there any prompts you can use to keep the images from having extra limbs? I tried creating a cover with Superman fighting Tex Arcana and Superman either had multiple arms or a badly distorted torso in the images it gave me.
|
|
|
Post by dans on Aug 22, 2023 17:00:14 GMT
SH, what AIs are you using? AIs are not really trained to count, so they do it poorly. However, if you can use a negative prompt on whatever robot you are dealing with, adding this might help...
[ugly, tiling, poorly drawn hands, poorly drawn feet, poorly drawn face, out of frame, extra limbs, disfigured, deformed, body out of frame, blurry, bad anatomy, blurred, watermark, grainy, signature, cut off, draft]
on some AIs, putting the prompt text in brackets makes it a negative prompt. On others, there is a separate line for the negative prompt and if there is, you leave the brackets off. On Night Cafe, there is a slider and you leave off the brackets but move the slider into the negative region.
|
|
|
Post by starskyhutch76 on Aug 23, 2023 3:04:29 GMT
I used stable diffusion. I typed in a negaative on extra fingers. I forgot to say no extra arms!
|
|
|
Post by dans on Aug 23, 2023 12:07:06 GMT
here's a site that uses stable diffusion to make changes to an image you upload for free. You upload an image and make your description as accurate as possible, then run it and it will produce 4 new images based on your image and prompt. You can then make changes to your prompt and run the original image again with the new prompt, or you can choose one of the 4 new images and start again. If your prompt matches the image closely enough, you only get minor changes between iterations; if your prompt is different than the image, the results are unpredictable. I sometimes like to put in an image and a prompt, make a run, pick the new image I like best, and run again and again, to get images that are closer to what I want. You can also download an image that you like, edit it locally, upload and keep changing the edited image.
I use these other sites to edit images I've created...
|
|
|
Post by DocQuantum on Sept 11, 2023 1:39:53 GMT
I haven't been happy with the kinds of results I've tried using various online ai art generators. I've got a character in mind that looks a certain way, but I cannot generate the right prompts to make it look right. I think I'll have to find a similar character and modify it with Photoshop using a more Old School method.
|
|