The Creative Process

Recently, some parents have had questions about the creative process I use to develop the graphics for my books, in particular, the Art Fundamentals Workbook. I was going to make a series of videos on our YouTube channel to educate the public on how artists have integrated AI into their workflow; however, I will share my response here first:

"Regarding AI, I’m happy to answer any questions. I think the conversation around it is often reduced to extremes when the reality is much more nuanced - and I love to teach people about it. I’ve shared this with a lot of other parents already, and I pre-apologize for the length. It's the teacher in me, lol. I probably should make some YouTube videos on the topic now that I have my channel up and running.

The cover to the Dragon Rage preview book. The pencils were by Mark Goldfine. I did the digital painting.

Before writing curriculum, I worked professionally as a digital colorist and character painter in the comic book industry. I am completely self-taught and have worked my way up from making terrible little pieces of art to being a colorist for the amazing Jose Delbo, working on a project with him and Val Kilmer before both of their deaths, among other great opportunities.

Unfortunately, I was actually one of the very first artists to lose my job to AI.

When DeepDream, early DALL·E systems, and similar generators were made available to the public, I lost my clients overnight - literally - no exaggeration. My whole career, which I have spent 20+ years developing, was gone in just a few hours.

That sudden collapse of work forced me to pivot careers very, very quickly, and it directly led to me focusing on writing and developing the art workbook.

I’m also not one to take things at face value, so, after a few weeks of feeling very sorry for myself, I dove into understanding this new tool that had taken my job from me. As a professional digital artist, it’s part of my job to explore modern digital tools, and that now includes AI-assisted workflows.

The workflows that have developed over the last year or two, being used by a lot of us artists now, are actually very similar to techniques used across concept art, film production, matte painting, publishing, and digital illustration industries. They have actually been using AI and other tools like it for years (long before it popped up into public awareness).

Artists have always adapted new tools into their pipelines, and now, we can use AI-generated

images as part of our creative workflow, particularly when working with complex scenes or visual elements. However, many times these are done with significant modification. I often keep certain elements from the original but rework the image extensively, through repainting, layering, editing, color correction, and refinement. So the final artwork or project is cohesive, expressive, and truly addresses what I am trying to illustrate. In some situations involving complex scenes, reference generation, textures, composition ideation, or visual problem-solving, AI is an especially essential tool.

An older project I worked on when AI was first developed. Three different images were generated and then digitally painted to bring them together to create this illustration.

So, basically, for a lot of artists, me included, when we use AI-sourced material, much of the time the final artwork goes through extensive repainting, editing, layering, compositing, correction, refinement, and redesign work, so the end result is cohesive and aligned with the educational and artistic goals of the project.

For this book illustration, I actually learned to train my own AI model locally on my computer. It was trained on a collection of images I created. This model was able to place a reasonably accurate version of the main character in different situations, so I could use them as reference points for the final images. Photoshop's new Generative Fill tool was used to adjust minor things like his hair in front of his face.

This is the final illustration. The hand and knife were fully digitally painted by me, as was his body and clothing, glasses, and background. Various tools and adjustments were used in Photoshop to make the character's face accurate. This also involved a lot of digital painting in Rebelle (which mimics traditional mediums) and Photoshop.

There are also some wonderful bonuses to artists and companies who are using AI. In fact, it's actually given many artists, especially concept artists in the movie and video game industries, an ethical way to source material when they work. Before, many just simply took pieces and parts of images they found on a Google image search and collaged them together. Those images and photographs were created by other artists and used without permission. But now, these artists and big businesses develop unique pieces of art to provide the foundations they need and still meet deadlines without infringing on other people’s work.

On a personal note, my penciler (a person in the comic industry who draws the pencil images before sending them to an inker and/or a colorist) and dear friend has a medical condition that has caused his arm to seize up. He can no longer draw, and that's been just abysmal for him. He has put some of his original drawings through AI image generators to create new versions of his work, and at times, that has given him a creative outlet and encouragement.

Also, along those lines, tools are not the same thing as authorship. The lessons, teaching philosophy, exercises, educational structure, pieces of art, creative direction, storytelling, and overall vision of the workbooks, in their entirety, are built from years of experience and skills.

Hope that helps you understand the bigger picture of how AI has been and is being used in the creation process. It's really interesting."