The Wild West of Generative Art

SUMMARY BY GOOGLE GEMINI WITH EDITS BY GRAMMARLY

==========

The emergence of AI art has revolutionized my creative process and challenged my notions of ownership and originality. I've come to some key conclusions:

1. Collaborative Creation: AI art is a collaborative effort, and I am just one participant among many, including the AI developers, other users, and the vast pool of human-created data that train the AI models. This blurs the lines of individual authorship and emphasizes the collective nature of creativity.
2. Ownership vs. Inspiration: While I might feel a sense of ownership over the AI-generated art I create, I recognize that it's ultimately an outdated notion. AI art is a product of collective human expression and inspiration, making it challenging to attribute ownership to any single individual, including myself.
3. Rejecting Claims of Ownership: I refuse to honour individual claims on AI-generated ideas, styles, or expressions, emphasizing the open and collaborative nature of AI art. However, I do acknowledge and credit the sources of inspiration where possible.
4. AI as a Tool for Exploration: I see AI art as a tool for exploration and discovery rather than a means to create "original" works. It allows me to create endless variations and interpretations of existing ideas and styles.

In conclusion, AI art is a collaborative and transformative force in my creative world. It has challenged my own notions of ownership, originality, and the role of the artist. Embracing the collaborative nature of AI art opens up new possibilities for exploration, expression, and the democratization of creativity.

WHEN ASKED ABOUT MY "PROCESS", I LEARNED THIS:

 My AI art creation process is a playful journey of exploration and experimentation. I begin by immersing myself in online galleries and Facebook groups, shamelessly copying "original" works onto my digital "workbench" as a starting point or source of inspiration. Sometimes, a spark of my own idea ignites the process.

Next, I turn to Perplexity to describe the image, utilizing its various AI platforms to refine and shorten the description. I find that a concise prompt often yields the best results. I then feed this prompt into different AI platforms like Bing Image Creator, NightCafe, or Playground AI to generate an image of "my own."

I enjoy experimenting with different platforms, as each offers unique strengths. Bing Image Creator is quick and easy, excelling at capturing mood and human interaction. NightCafe, with its paid features, allows me greater control over aspect ratio and details. Playground AI offers more flexibility in terms of content, especially when exploring themes that might be considered sensitive by other platforms.

Once I have a "keeper," I often use Word to add framing effects or incorporate the image into a larger project. I might even share my "variation" with the creator of the original image that inspired me.

Ultimately, my AI art creation process is a continuous loop of inspiration, experimentation, and refinement. It's less about claiming ownership and more about exploring the endless possibilities that AI offers for creative expression. 

 

CONCLUSION: GEMINI MAY PAY FOR ITS KEEP BY IMPROVING MY WRITING. OTHER TOOLS ARE GREAT FOR DRAWING PICTURES.

==========

After swimming in the ocean of AI, I have come to a few conclusions. 

(1) All art involves the participation of many individuals, including the artist, the teacher, the imitator, and the critic ... in the end, it's a product of all humanity.

(2) It's natural for the last one in the chain to claim "ownership." It's like wanting to own the mountain because you took a picture of it. That feeling is natural but obsolete. In fact, the "last one in the chain" is probably somebody being paid pennies to tag thousands of pictures in a day.

(3) I refuse to honour any individual claims over the idea, style or particular expression. If I can, I will credit the one who last inspired me. If someone expresses an official claim, like a copyright notice, I won't bother to copy their work.

That said, here are elements of my "process:"

For inspiration, I browse Facebook groups and online galleries, shamelessly copying "original" work onto my "workbench." Sometimes I even have my own idea.

Sometimes, a prompt is available. Otherwise, I use Perplexity to describe the image. Perplexity provides several AI platforms, including the ability to shorten the description. One issue with Perplexity is that it's AI "mind" is "safe," meaning it will not look at anything remotely suggestive. "Safety" here is protecting Perplexity from scandal. Nothing to do with safety in general.

I use the resulting text prompt to create an image of "my own." In the case of Microsoft BING, I am provided thousands of pictures like the one I drew, as in the case of the hummingbird here. It is amusing how many people want to sell their AI work. Good luck. I may put one of those on my workbench and return to AI for a text description. Paying for one of the thousands of variations BING provides simply doesn't occur. They are trying to patent the idea of a cheese sandwich. On the other hand, I never re-post or save anything I find as-is. I'm looking for ideas.

The descriptions are usually too long. Platforms vary in how much "context" they can absorb. Less is more.

BING is quick and easy. It provides a very good interpretation of mood and human interaction. Of course, I only get a few shots per day.

My paid NightCafe account allows me to alter the aspect ratio and request more details. I also use Playground.ai. Sometimes, I need to find a platform that is a bit more "broad-minded" than BING, which balks at the slightest hint of eroticism.

Finally, if I have a "keeper", I will copy it into Word to provide framing effects. Word is also an excellent place to capture the prompt. I will occasionally credit the author of the picture that got all this started and even get back to him/her with my "variation." Not necessarily.

The final picture can then be transferred to my gallery in Second Life. For most pictures, Word is the final resting place.

The gallery is usually the only place where my pictures are publicly available. Visits are rare. I occasionally post to one of the Facebook groups, especially if there is a dialogue about the original or methods.

The above image changed the lady's expression from grumpy to happy and welcoming. I left AI to create her dress, committing unnecessary detail. I hinted at Vermeer's style since he was famous for sympathetic portraits of ladies. But I kept the idea of meeting this lady in a secret place back in 1880 or so.

This was my starting point - really much better than mine:

I have to admit this girl is prettier (easily fixed), and her outfit is dazzling. It's a great picture. I just like to tinker. In any case, I would not put this one in my gallery. It's just another pretty girl. Note that it is signed, indicating the creator's proprietary sentiments. One thing AI knows is how to draw; it's pretty girls. One trick here is that the outfit generated by the prompt is too complex for AI to describe, so the "loop" must strip out a lot of description.

AI refuses to describe the one below, sending me back to square one. Still, it's fantastic work. It's too bad AI protects us from such things. Sometimes, AI accidentally adds erotic elements from the artist's style or the picture's context, which pulls up similar images.

a similar issue artists with the following picture

which would be OK in a children's book 

but too naughty for AI to describe:

The following image, generated by NightCafe, illustrates this problem:

Masterpiece: A stunningly beautiful young blonde woman 

in a strapless gown 

with an aura of serenity is depicted

 in a fantastical setting that blends elements of nature and the cosmos. 

"Strapless" is understood, and it is okay to create the image but not to describe it. This subtly points out that AI is doing the creative work here. I supplied only the much-edited prompt. AI loves to take pictures of pretty ladies; its training data is stuffed with them. Asking for something different can give you the "masterpiece" you hoped to trip over...

The above image is posted proudly as "original."

BING provides thousands of similar pictures, 

which are presumably the AI contribution

It's easy to see how many think the AI picture somehow "steals" from genuinely original art. The problem is that the idea, not the picture, is commonplace. Even if there is an exact precursor of the AI version, it can't claim to be "original." Even if our "prompt engineer" combined the common theme with an original technique, the credit would seem to go to the AI algorithm, not the "engineer."

Here is one variation created by an "engineer" (me) who just asked for an oil painting with the same prompt:

Time investment for the "engineer" - 60 seconds

In another 60 seconds, we will have loons. This is not "engineering" or even a "craft." 

It is more challenging than Paint by Numbers because it requires effort. For many years, my parents had the framed results of their needlework equivalent of Paint By Numbers. We all admired the pretty pictures (birds, of course), but nobody claimed originality except for the company that sold the kit. Even then, their claim was not a patent on bird pictures.

The algorithm demonstrates robust creativity when we ask for something genuinely original.

I claim credit for the 10 seconds it took to think of this.

And credit for another 10 seconds it took to create this.

We are dealing with something unique here, but it's not "original" art. It is more like a voyage of discovery, like Alice's adventures in Wonderland.