AI Art Experiments in Regards to the "Barbarians of the Storm" Series, My Real Career, and Beyond
What I've learned, its utility, and where I see it going
*The images above were all created with StarryAI; character info at the bottom of the post.
Recently, I decided to dabble in AI art. There were a few reasons for this:
I am a professional artist and I am always interested in emerging technologies and their potential as tools, going into the future. Things always evolve and it is best to stay ahead of the curve and learn these new things, instead of outright rejecting them, as many artists are doing. I need to understand how best to communicate with AI in regards to art, as I think that it has a lot of potential utility going forward, especially in regards to a lot of artistic avenues that most of us aren’t even thinking about yet. Adapting and knowing AI well, will only serve me better in my real career.
I don’t have the time or ability to produce a lot of character portraits for all the characters in my books. With that, I wanted to see how well AI could solve that problem. They’re far from perfect, but they have provided me with a good point of reference to give to a human artist, at some point. With my personal projects, I am able to practice and fine tune my skills.
It allows me to create dynamic imagery for this blog, as seen with the character pieces above. When scrolling social media feeds, people tend to stop and click on links that catch their attention, visually.
On this strange but neat journey of AI experimentation, I’ve tried several different AI softwares. The two that attract me the most and have given me the best results are Midjourney and StarryAI. At least, these are the two best in regards to what I’m looking to create, right now.
I actually like the potential of Midjourney the best. However, its free trial is extremely limited and I ran out of credits before I even knew that I needed to pay attention to that. Additionally, I hate that I can only use it in a live, fast-moving, Discord chat. I also don’t like that your work is public. I get why they do that, but if I wanted to experiment with AI on more proprietary or trademarked material, I don’t want those experiments to be seen publicly. As I work as a creative director for a well-known global brand, I don’t want potential competition to see what I’m playing around with.
StarryAI, however, also appears to have a lot of potential, and as I get better at talking to it, I’m able to massage my creations into something more defined and closer to the idea in my head. Also, overall, the cost of credits is pretty cheap. I also didn’t run into issues with the software trying to censor my input, and for those of you over 18, you can get some interesting results by simply typing something like “sexy” or “brutal” into your prompts. Granted, I’m not making porn or blatantly gory images with it, but it’s nice to know that I don’t have to work within a creative framework that’s strictly PG.
At this point, I am having a lot of fun experimenting with AI, and honestly, the more I play with it, the more I can see its utility and potential going forward. However, a lot of artists are losing their shit over AI art and how its results are “soulless”, and how “it can’t do hands”, or ever replace a very talented human being. What they fail to see is that this will only get better and evolve. Eventually, it could be indistinguishable from a human and possibly, dare I say, surpass most artists in certain ways.
This is a sentiment built off of fear. It’s the same as when artists moaned about the emergence of digital art 30-40 years ago, whether it was Photoshop, CGI, or computer animation ala Pixar. None of these things killed traditional art, they just became tools and created new mediums for artists to use and explore their creativity with. They also provided creatives with more time, as many of these things significantly cut down the time it takes to do certain tasks the old fashioned way.
Ultimately, AI generated art is just a tool, a new medium to experiment and exploit for the benefit of the artist. I understand the apprehension and the fear, but I also understand that some of it also comes from a place of laziness in artists that don’t want to learn new things, expand their knowledge, and adapt to an everchanging world. That’s not me, though, I want to learn the new things and adapt, because I like paying my bills and since I still have a passion for creating, I only want to improve and become a better artist, which comes with understanding what’s on the horizon, as well as one possibly can.
Also, AI isn’t going to take your job, per se, if you’ve become a master at using it. Someone has to know how to talk to it and type in the prompts. In a way, that’s an art in itself. Become the master over your perceived demons and those demons can’t come for you.
Whether one likes it or not, this is something that could become much more common place in two-to-three years. In five years, it’s probably going to be reaching a semi-mainstream status. A decade from now, it could be incredibly common. It’s probably best to get to know it, even if you still don’t like it. Over time, you probably will come to appreciate it, just as most artists have come to appreciate the common digital art software that’s so commonplace today… even if Adobe sucks as a company.
*Characters featured in the images above are (from left-to-right, top-to-bottom): The Seawolf, Merith, Dan, Fenrik, Frank Murdock, Kildr the Coldclaw, Queen Valkyana, Xulgog, Erzulyn, Shakon-Kazoth, Ielic, King Bruxor
I find Midjourney is most useful for abstract landscapes and fantastical images. I've even gotten fairly-competent character portraits out of it. That said, even if it becomes indistinguishable from human artists or surpasses them entirely, I can't ever imagine relying on it or using it as a replacement for a human illustrator. Perhaps its sentimentality, but to think this technology won't negatively affect traditional artists in the future is being quite optimistic, from my perspective.