I’m sure we’ve long past the point of realizing that generative A.I. imagery is probably here to stay. It has proven itself excellent in many cases at generating stunning imagery, and it’s only going to get better at doing so. What it can’t do today, it’ll will probably be able to do soon. Sure, it has its kinks, flubs, and artifacts, but those problems are actively being solved. I remember people scoffing about A.I. because it couldn’t render hands realistically. Now some models can do so far better than most humans could hope to aspire to. The technology is improving at scary rates. It’s leaps and bounds better today than it was a few months ago, and it will most likely continue to be improved upon for a long time to come, regardless of how strongly anyone would like to will it out of existence. It will probably be coupled with advanced robotics in the future, and it will probably continue to be used to create some compelling work in the future – both digital and physical, both commercial and fine art.
In the very beginning of my own creative journey, I was making images because I wanted to make cool things that I hadn’t seen before. That simple motive is still at the heart of my creative impulse. If A.I. can help make this happen, that could be seen as a plus! But if my goal is to make a living making cool looking things, A.I. could be seen as a very threatening detriment to my existence. If I call myself an artist because I make cool looking things, but those cool looking things don’t mean much, if those cool looking things don’t provide more meaning and value to the world than an average person with A.I. is able to produce, then I’m probably going to see A.I. as existential nightmare fuel that will replace artists entirely and ruin art forever. Thankfully, from my perspective, great art is more than just well-rendered eye candy, and an artist (or at least one with staying power) is much more than the ability to make cool looking things.
I don’t see A.I. replacing all artists, but I can see it replacing many talented craftsmen or image renderers. What the camera phone in every pocket did to photography, A.I. will do to every artistic medium, by automating craftsmanship and making artistic expression more accessible to the masses. Now a person won’t need to learn too much about craftsmanship in order to become an artist. Instead, they can write a prompt that will bring their visions to life. Does this imply that every A.I. image will be a work of art? In the past, a handful have recognized and outlined a clear distinction between artist and craftsman. Most creators today have been okay without recognizing such a distinction, and their work has flourished by the simple excellence of their craftsmanship. Likewise, the general public has been just fine with a fuzzy and overly broadened understanding of art, conflating art and craftsmanship, and recognizing everything and anything we feel like calling an artwork as an artwork. It’s easier to assume that every painting or drawing that exists is snugly categorized as works of art, than it is to think about what art is, what it does for us, and how we can make better art. Now that A.I. can generate amazing imagery that is on par with the work of many artists, and people are arguing whether or not A.I. imagery should be considered art, it might not be unreasonable to think that such a broad definition of art could prove problematic. Meanwhile, the same artists whose careers have depended solely upon craftsmanship and style are left feeling like their livelihood is on the line, fearing that they could very well be replaced over the next few decades. There is, after all, going to be a lot of competition in the image-rendering market as A.I. progresses beyond its current stage of infancy.
When I say image-rendering, I refer to the act of producing images via physical or digital media, and all the excellent skills of craftsmanship involved in that. Great painters, illustrators, sculptors, animators, and so on – who have spent a great deal of time mastering their abilities – will be outpaced eventually in terms of their ability to render the artwork desired by their employers or their audience. On the other hand, great artists are more than their ability to craft technically excellent images, or to develop unique and appealing visual styles. Great artists typically have a well-developed creative vision, with which they can translate their inward experience of life into the language of their medium, to create a new experience which can be shared with others. If an artist has a strong and cohesive creative vision guiding the creation of their work, A.I. could simply be an excellent tool at their disposal, but it won’t replace their creative vision, nor the essential spark of human consciousness and experience that drives and forms their work. Those are the artists who I think will be fine, whether they use A.I. or not, and ultimately A.I. will just be a tool at their disposal to bring that creative vision to life.
How does A.I. work? From what I’ve come to understand (open invitation to correct or educate me in the comments below) it goes something like this: A.I. is trained on huge amounts of data. Referencing that huge library of data, A.I. will produce content (like text or an image) that it predicts the user expects from it, based on the user’s input or prompt. The more images and data that the A.I. has been fed, the more accurate its predictions will be, as it will have more to reference while generating such content. For example, Midjourney has (I’m guessing) been trained on hundreds of thousands of photographs, and I’m sure metadata has been included. It has references for what a man looks like, and it has references for how a man (or other objects) photographed with an 85mm lens at F/1.4 could look like. If someone types in a prompt ‘Portrait male f/1.4 85mm,’ referencing all of its data, it can predict what it assumes the user expects to see and generates an image that can arrive at a pretty convincing result. The more details the user provides in their prompt, the closer the A.I. generating will arrive at in proximity to what the user envisioned.
Image generated using Dall-E 2 with prompt “portrait male f/1.4 85mm”
Notice, though, that A.I. is only producing content that it predicts the user wants to see, and the accuracy of its predictions is only as great as the amount of data it has been trained on and can reference. It can combine concepts, attributes, and styles based on the descriptions provided by the user, and does a surprisingly decent job at imitating, mixing, and mashing different visual elements to create what the user is looking for, all derived from the massive amount of content that has been scraped from the Internet and fed to it. What it can’t do, however, is take your personal perspective, your beliefs about reality, your worldview, your life philosophy, your history, your narrative, your most precious memories, the depth of your personality, the culmination of your conscious and subconscious experience of life, the minutia of your aesthetic attractions and repulsions, your inward experience of consciousness, and translate all of that into an tangible artistic experience that substantially expresses who you are and how you see the world. All it can do is render images based on user input.
Again, what A.I. has essentially accomplished is mass accessibility to decent enough craftsmanship, decent enough image rendering and compositional automation. Unfortunately, craftsmanship is all that most artists have, and the result of their work is a product similar to what anyone with A.I. access can now achieve, (and what cannot be achieved with A.I. today might be achievable in the near future.) Though their craftsmanship is highly developed, their creative vision is severely underdeveloped. The extent of their creative vision is often no more than a mixture of influences, inspirations, subjects and styles fused together. They are like authors who have mastered a language and writing techniques, who know how to describe anything in excellent detail, but their works do not feel alive, they do not bring forth meaning nor value, they have no real compelling perspective to from nor any real message they feel compelled to write about. Anyone can take a moment to list different inspirations, styles, and subjects they enjoy, and now that A.I. has automated image rendering and made craftsmanship accessible, they can type those inspirations and styles into a prompt, leaving to the A.I. to generate dozens of new images, from which they can select the ones which most aligned with their tastes and desires. Now anyone can do with A.I. what most artists can do, because most artists have relied too heavily on what A.I. can now accomplish, whether that’s creating images based on a handful of visual preferences and subjects, or choosing to create what they predict others want to see or buy from them, providing eye candy instead of food for the soul.
Now I admit that I can be overly idealistic, and one of my ideals is the role of the artist merging with the role of the philosopher and storyteller. Whether or not that should be the case in reality, I still think creatives should begin to seek a higher calling in the position of the artist role. The days of celebrity artists making careers out of work with broad surface level appeal could soon be lost, because the broad audience they appeal now has the tools in their hand to create the same work with little effort. Much like the camera in the pocket has led to millions and millions of photographs, A.I. will lead to millions and millions of “artworks,” and what works to separate artists from the crowd today won’t work for very much longer. Instead, artists ought realize their job is not only that of the craftsman, but that their job is to form a creative vision and bring it forth into the world using their craftsmanship (or A.I.).
How does one develop their creative vision? Most artists today aren’t thinking about their creative mission and the meaning they are intending to imbue into their work, mostly because they haven’t yet clearly recognized the meaning and value they feel driven to express. In order to discover this mission, we could work to pull unconscious or subconscious material into conscious observance and thought. This might involve reflecting upon their framework of reality, their worldview, their philosophy of life. It might involve questioning their conclusions, assumptions and beliefs. It could involve recognizing and clarifying emotionally important imagery, or reflecting upon the implications of their core memories, or questioning why they gravitate towards this and not towards that. It could involve actively observing their own conscious experiencing of the world, actively observing themselves as a product of their society, working to recognize their biases and searching out from where those biases arrived. It could involve placing themselves in others’ shoes, experiencing more angles and perspectives, making a practice of empathy and understanding, which could be useful for obtaining a clearer and more holistic view of oneself and one’s place in reality. I suspect all such practices could be substantially helpful in developing a creative vision.
As artists take on this search for meaning in their life and all life, they ought to continue learning, wondering, exploring and discovering, learning to think critically, and taking their varying fractured perspectives on the numerous aspects and issues of reality and realigning them in a way that isn’t self-contradictory. They must also take into account their personal story line, their personal history, and how they uniquely experience life through their own body, mind, and soul, and recognize the unique expression of humanity that they embody. They ought to work to connect and bond all these fragments into a unified creative engine, to ultimately recognize and develop said unique perspective from which they can pour forth their particular message in an artistic vision, so that they have something to actually contribute to the world instead of merely mirroring the impulses and shallow reflexes of the world. Once they have this perspective, granted that they have also mastered the visual language of their art, and their have mastered the technical craftsmanship of their medium, the unification and integration of all three of these could comprise a potent creative vision.
So I think it’s true that many artists will be (rudely) woken up by a higher calling, held by the throat to a higher standard by the technical prowess of A.I. They will be called to form a compelling creative vision of the world, to discover their perspective and integrate it with their craftsmanship and visual language and artistic style, so that they can successfully translate their personal inward experience of life into a sharable experience with others (a play, a sculpture, a dance, a song, a film, a painting, a photograph.) Yes, an artist should still master their craft – whether that craft is painting on canvas or writing excellent prompts for A.I. generated digital art – but the aesthetic side of artwork is only half of the recipe of a powerful and potent image. If an artist doesn’t have their creative vision, the severe competition we’ll eventually face in rendering neat looking eye candy against A.I.’s prowess will wear us out, regardless of the medium we work in. Without the search for meaning and value in life, through introspection, self-reflection and observation, we won’t have much to say, to share, to teach, to tell, to bring about for others to wonder, we won’t have insights and angles and wisdom with which to indwell and imbue into our artwork, our artwork won’t be anymore terribly meaningful than the average image output by generative A.I. A.I. will continue to be improved and is already a powerful tool, but it makes a poor human being and a poor creative visionary. If, however, we have our creative vision, it won’t matter if we use A.I. or not – our work will be something technology cannot yet replace.