Steven is a good friend of mine, a remarkably gifted concept artist whose career in the gaming industry spans well over two decades. A significant portion of his work revolves around transforming the nebulous realm of people’s ideas into tangible, visual images.
When we talk, we frequently engage in an eclectic mix of conversations that delve into a myriad of nerdy subjects, from the vibrant world of Anime to the complex universes of sci-fi literature. Lately, however, our dialogues have often gravitated towards the burgeoning realm of AI-generated art.
These conversations initially started with me sharing examples of creations I’ve made using Midjourney. My efforts were typically mash-ups from various influences — Toriyama’s Dragon Ball Z with Star Wars. Steven, in his usual way, remained unimpressed. His disdain for my pop culture remixes ignited a desire to delve deeper into my own wellsprings of AI art originality.
This time, however, I had a surprise to share with him. I confessed to having used his name as an input on Midjourney, leading me to believe that the AI tool recognized his unique artistic style. Using a simple prompt like “in the style of Steven M—” I could get the AI to generate artwork that strikingly mirrored Steven’s unique aesthetic style.
I shared a couple of these images, art pieces that were clearly modeled after his past work, to support my argument. One example was a piece he’d created featuring cartoon penguins acting as tourists. I asked Midjourney to redraw that image with the prompt “in the style of Steve M—”, and the result was eerily accurate. Likewise, I had it recreate a Cthulhu piece from his portfolio, and it achieved the same eerie atmosphere and scale as the original. Midjourney had demonstrated its ability to translate text into visual art in a way that paid tribute to Steven’s style.
This revelation sparked profound contemplation about the implications this could have on his production techniques and how its creation of original artwork in his style could affect his work.
Steve mused over the benefits of having a tool that could generate initial drafts, thereby reducing his workload. This naturally steered our conversation towards the potential impact AI art might have on the entire industry.
As we talked I realized that the true merit of these tools doesn’t lie in creating uninspiring pop art mashups, but in harnessing their potential to craft solutions to real-world problems. “A lot of using AI effectively,” I remarked, “is about taking control.”
“The real skill of being a working artist has always been about control,” Steven echoed. “What really matters is whether you can consistently deliver what’s required to move things forward on any project you’re on.”
This perspective was a revelation, an “aha” moment, that helped me understand the true value of AI and the objectives those of us working with AI should be striving for.
In numerous articles discussing AI innovations, a common rebuttal points to the inferior quality of AI-generated work compared to human-created art. Critics argue that there is something ineffably pure and human about art, an essence that an artificial intelligence, no matter how sophisticated, will never replicate. This may hold some truth for fine art, particularly those created in traditional mediums, but Midjourney has already been used to fool art contests into awarding prizes to AI-generated submissions, oblivious to its artificial origins. And with AI developing at an astonishing pace, the day when it can regularly mimic human creativity isn’t far off.
AI’s potential to change the art industry’s landscape is clear. Companies adopting AI for content creation might need fewer artists but people capable of guiding AI to generate consistent, useful material will be invaluable. Additionally, as AI art becomes cheaper and easier to produce, we should anticipate more products incorporating it, generating more roles for those who can control AI art output effectively.
The real prowess of an “AI artist” will, in fact, be their ability to consistently deliver original, creative work. Not merely through traditional methods of pixel manipulation, but by mastering the art of guiding generative AI to produce content that consistently meets the demands of their projects. The real challenge lies in taking control of these tools to deliver the results that will solve real-world problems and earn them their keep.
For Steven, this may present an opportunity not just to boost his productivity, but to explore and inspire himself with work that already resonates with his style.
Of course, this doesn’t mean that human-made, custom artwork will entirely disappear from the commercial art landscape. But ever since we’ve embraced the digital sphere for creating and delivering artwork, the unique “human touch” has become a subjective matter, just like distinguishing between “handcrafted” and mass-produced items, even when both are often churned out from the same assembly line.
The frequent laments of “art is dead” or the fear of artists losing their livelihood to AI ignore the promising reality of artists already adapting to these new tools, using them in ways that not only fulfill existing commercial art requirements but also push the boundaries of art into territories we’ve barely begun to fathom.