Yesterday, I made a post that started off with my fascination with Loab, the woman that is “haunting” the internet after being created by generative AI, and discussed how I wanted to figure out a way to use it.
This is my first dabble into generative AI.
Generative AI is a relatively new technology that uses artificial intelligence to create brand-new original images based on a series of prompts you feed it. As I understand it, it’s own understanding of images is based on being fed the wealth of images and contextual information that the internet has. So when you tell it to create an image of, say, a fluffy pink elephant juggling five balls in outer space, it has the entire internet to learn and understand the concepts of “pink”, “elephant”, “fluffy”, “juggling” etc., and how these things relate to each other; how “fluffy” could be applied to its understanding of “elephant”, how an elephant could conceivably juggle and therefore, which limbs of a juggler correspond to which of an elephant’s limbs, and so on, and can then, create a logical image based on those prompts. It’s similar to how, when I wrote “a fluffy pink elephant juggling five balls in outer space”, you drew on your knowledge of those concepts to imagine that… despite having never actually seen it in real life.
Actually – I might try that. Are you imagining the Elephant in a spacesuit? I’m imagining it in a helmet at least. I wonder if the AI’s contextual references would tell it to put the elephant in a spacesuit if I don’t specifically tell it to?…
The first thought I had was that, as someone who has never and can never visit the southern states of America in the 1860s, I am very reliant on reference images to get a feel for the place, the mood, the clothes, the buildings and architecture, the time and so on to be able to draw my Robert Smalls graphic novel. Searching out these references is very time-consuming and actual photographs from back then are sparse. Finding just the right reference picture for what I need is even more time-consuming.
But if an AI, with instant access to the internet’s wealth of images can now combine all of that and “imagine” these reference pictures… that could be a game changer!
So here, in Midjourney generative AI, I entered the prompt: “children playing on a busy Charleston Street in South Carolina in 1860, photorealistic –v 4“(the “–v 4” tells it to use the most recent, Beta testing version of the AI.
These are the 4 images it returned immediately. I’m impressed to say the least. Ignoring the odd person with 2 heads or whatever, for giving a sense of place and time it is fantastic! I feel this could be incredibly useful! Understandably, these all look like paintings of the time despite me specifying photorealism – a tell, I’m guessing, to the fact that these are the references the AI itself has to draw on.
Do I feel it is in some way cheating?
No, I don’t. Initially, yes – I had some weird thoughts about whether I am shortcutting something that I should be doing. But it isn’t as if I haven’t already scoured the internet for hours finding reference pictures for my project. I have many. I am simply using more of the tools available to me. These AI images are not enough alone. They still don’t create exactly what I need. It’s not as if I can simply copy or even trace them into my novel. I will still be comparing them to actual reference pictures from the time for accuracy. However, ultimately, my aim is to tell a visually interesting story. I would like it to be somewhat accurate in terms of how it looks, but that is not my sole aim. It is more to create an accurate feeling representation of time and place. I want the setting to serve the story, not the other way round. I feel using AI images may actually increase how well I can draw these things as I will have access to a limitless form of reference.
Those are my thoughts on it at the moment. They may change, and I will be sure to let you know if they do.
Unsung Superheroes is an ongoing project, creating graphic novels based on the stories of real-life historical figures whose stories may not be as well-known as they could be.
New pages are added to the current story on the Homepage.
If you would like to help support this project financially, please consider signing up to my Patreon. For less than the price of a cup of coffee per month, you can become a Patron of the Arts and help keep this project going. There’s no minimum sign-up period. You can stop at any time.
Also available is the opportunity for you to appear as a character in the novel.
I’m donating 10% of all pledges to Anti-Slavery International to help them in their efforts to stop modern-day slavery and human trafficking.