You know, it’s been some time since we’ve really worked with Squagon’s images. The OpenArt upgrades arrived, and I got totally sidetracked with all the new bells and whistles.

It’s time we went back and worked on our favorite flying squirrel. The upcoming summer events will be requiring a good many images of our intrepid squirrel/pirate, as he’ll be the focus of our crew’s adventurous expeditions.

While I can’t really tell you what those adventures are, as of yet, you can certainly help me pull up some images that we’ll all love seeing, once we set sail. Don’t worry. I won’t leave anyone at the docks. I give my solemn oath!

Anyway, let’s see what we can come up with for our fun loving, flying furball…

If you hadn’t noticed the problem yet, let me fill you in… his tail looks pitiful! Let me see if I can fix that…

Good grief! All I did was ask the ai to make his tail much fluffier than it was, but things got a bit out of hand. Let’s adjust our request to say make his tail somewhat fluffier…

The reason I’m showing you these images is to reflect the sensitivity of OpenArt’s new chat editing feature. Sometimes Norbert, as I call him, likes to take an idea and run with it, often ending in disastrous results. At other times, if I use the right wording or the image is simple enough to fix, this new feature works out rather well. I’m not saying that I want to use this image, but it gives me a bit more confidence about some of the finer adjustments needed, if I want to perfect certain scenes.

Do you notice how the image feels somewhat photorealistic, while still having a bit of that whimsical softness. I think it brings Squagon more depth of personality vs. an actual image of a squirrel. I’m able to give him an endearing quality that reaches deeper than just a plain old, backyard peanut thief! What do you think?

Honestly, I think it has more to do with the program understanding my query, than it does about my skills as an artist. I am a writer, not a painter, nor sculptor, nor full-time national geographic photojournalist, though I may have done well as a traveling writer. The only way I can communicate with any of you is through my literary donations, here on WordPress. For that, I need the help of an art platform that can assist me in portraying a thought, memory, or imaginative tale. I want you to see what I see, which can be tricky when using an artificial form of images.

The thing to remember is this… the program only as good as the one using it. It’s an artificial means of creating images from words, nothing more. The ai doesn’t think for itself, nor you, for that matter. It simply responds to a pre-written set of words, and it holds a memory bank of definitions for those words, phrases, and sentence structures.

Normally, I will start with a simple descriptive sentence. I offer the ai a character image, a brief visual descriptor, and an action to perform. On occasion, this first request provides a usable image. However, for most of what I do, the program offers an ai assisted expansion of the original request. I like to use these extra helps because many times its not what you ask, but how you ask it.

For instance, when I ask the ai for Squagon sitting on a bale of hay, this is the first response from the ai:

Now, I’ll ask for the ai to assist with the query. This was the result,

A photorealistic image of @MySquagon perched atop a golden bale of hay in a rustic barn setting, natural afternoon sunlight streaming through weathered wooden slats, casting warm amber highlights across the textured straw and creating soft dappled shadows on the worn plank floor, the hay bale’s compressed wheat stalks visible in detailed layers, dust motes floating lazily in the sunbeams, aged timber walls with peeling paint in the background, the character resting naturally with balanced posture against the organic texture of the dried grass.

These were the images offered from the enhanced prompt:

In this situation, I much prefer the first image over the assisted one. It captures much more of his personality, and looks more realistic. Sometimes the ai assist works out better, and sometimes not. It all depends on what I am asking the ai to do with the character. The action will always play heavily into the image’s final features.

For Squagon, I think we shouldn’t have too many obstacles along the way, but it was better to be safe than sorry. I wanted to do some work with the little guy before summer expeditions get underway. I like to be prepared for any and all perceived hang ups and/or roadblocks to a successful sailing season. You’ll thank me for it later, I can assure you.

Well, that’s all for today’s tinkering session. I think we got a lot of things figured out for now, and big or small, they all will help to get us where we were headed, right?!

May your writing hold true, and your words make a difference to the world…

Hugs