Thoughts Regarding The Current State & Our Use of AI
Our current, and possibly wrong, thinking about AI images
This is a long, in-depth article. I apologize for the length, but this is a new and complex topic and there was no way to address it in a shorter manner. I hope you’ll set aside the time to fully read and, if you wish, respond by leaving a comment.
An artist wrote to me to ask:
"Hi. In your latest newsletter email you illustrated it with an AI image using DALL-E. Ballsy move given all the artists angst over AI..I have no problem with AI, it’s just another artist's tool (with a learning curve). Just curious why you poked the bear?"
The question above represents a whole class of feedback we’ve received about the new AI image tools. This essay is my attempt to answer these questions and it represents my current thinking about AI, while fully acknowledging that I may change my position in the future when there is more clarity, especially legal clarity, about this new technology.
Short answer: I use the tools I need to use to make the art I need to make. I hope you do the same with your own art. Please read on for the longer and more nuanced answer.
Use the Tools You Need to Use
I don't censor myself in terms of what I create. As the saying goes, “write like a baby, edit like a bastard.” So, as I create (writing mostly in my case, but this applies to all art forms), I don’t censor myself in terms of the subject matter or in terms of how I create. When I am creating, I am respecting and reflecting the grandeur of the universe…of Creation itself, the universe gifted me with life, and my creations mimic that gift back to the universe. I should add here, as some readers may be wondering that “it’s just a newsletter”, I’m also talking here about my own personal art, my writing, which includes essays, reflections and fiction, and is published weekly on my personal site here:
But these thoughts also apply to the knowledge that I share with our artist customers and followers, as again, I want to use every tool possible, and every inspiration possible to share what we need to share.
So, back to the topic at hand: In my own work, the only way I know to properly respect the inspiration that Creation bestows upon me, is to utilize everything I have access to. It is the only way to realize and honor the small piece of Creation the Muse gifts to me.
That includes all tools available to me such as, my computer, spell check, my phone, photography, pens and paper, walks outside, reading other authors, inspiration from other sources, and, while it may be controversial to some, yes AI tools.
Text based AI tools may (and have) help me find inspiration. And image based AI tools may be used for inspiration or for appropriate illustration of my stories. I owe this to the Muse. And I will not shy away from the controversial, nor will I stick to only what the worldly order approves of, for if I need to color outside the lines to realize my art, I will do so unapologetically. For that is one noble purpose of artists: to express things that need to be expressed, even, sometimes, things the social order of the day does not wish to be expressed.
Here’s an example illustration that goes with a short piece honoring my (deceased) beloved cat Lily and how I sometimes feel reconnected to her when I meditate in the backyard.
Excerpt from: The peaceful, eternal now, where you still live
And thus, when you died, that little piece of me, that lived inside of you, died with you, leaving a gaping hole. But conversely, a little piece of you lives within me still, and when I meditate in nature, in our backyard…in your backyard, that little piece of you lives again, for just a moment refilling that hole as we step into the peaceful, eternal now. And you once again revel in the sheer delight of feeling the wind on your face and the sounds of birds frolicking.
And all rational thoughts I might have disappear for a few eternal minutes, and in that ethereal space, we are again together and we are again happy.
Would my short written work here have been better illustrated with the standard stock photo of a woman sitting in a meditative pose? No, the piece wouldn’t work with stock photography. The image and the words are connected and complete one another. Nowhere in the piece do I mention that I’m talking about a cat, although I hint at it, it’s not revealed until the end, when the image is reached. I’m not a visual artist, and I’m not going to hand-draw my own illustrations, and this has given me a tool to, in my opinion, improve my stories in a way that is accessible to me. It has expanded my abilities to share my visions.
Legal and Copyright Issues Around AI
Back to the topic of AI, and I realize this is a controversial topic in some circles and for some artists, there may very well be legal issues surrounding some of these tools. Resolving those issues is something for the courts to decide and not something for me, personally, to pass judgment upon. As with any new technology, there are important, confusing issues to resolve, and indeed, lawsuits have been filed by companies and by artists against some of these AI companies. These suits will be litigated by the courts and we will be given legal direction regarding regulations and legal protections surrounding AI usage. I applaud these suits and look forward to the clarity they will provide for AI companies, content creators and artists alike.
Here is one class-action lawsuit being pursued against StableDiffusion, MidJourney, and DeviantArt for copyright violations: Click here to learn about this lawsuit.
And here is an IP Attorney’s opinion as to why the above suit will likely fail: Click here to read a counterpoint to the above suit.
One thing I should point out - You cannot sue for copyright infringement of an unregistered copyright. This has been ruled by the US Supreme Court. Did you know that? Have you registered for copyright protection on all of your art works? If you ever want to register your copyrights, you can do so here.
By the way, there are some positive changes coming to StableDiffusion that I think are a step in the right direction to provide protections for artists who are concerned about this technology:
1. StableDiffusion has announced that future versions will not allow people to ask for an image “in the style of” a particular artist.
2. StableDiffusion has created a way for artists to Opt-out of having their images included in future training data sets for the AI. You can do that at this link: Opt-out of StableDiffusion
AI as a useful tool
However, in the meantime, if a tool is available to me, and it can be used to create, then I will use it, and I believe that is the proper attitude for any and all artists to take. If we went back in time, not much more than a dozen decades or so, there would be people demanding that we, “Don’t use cameras! They are stealing work from artists!” Yet cameras greatly expanded the scope and size of the art market, and provided a great new tool for artists of all kinds. I dare say that, of the artists reading this, nearly 100% of you use a camera in some fashion regarding your work. If nothing else, simply to capture inspiring images. And, personally (and I’m not alone in this belief), I believe AI has brought us to the cusp of an explosion across all forms and methods of creating. We will see an expansion and explosion of new innovative art and new types of artists.
AI does not replace experts or artists, but it does give us the equivalent of a million interns to assist us with our early experiments and explorations that eventually, may lead to finished artworks. Today, you can take 1,000 photographs and then choose the best ones and mix and match ideas to create your final painting. Tomorrow, you may do something similar with 1,000 AI generated images which are narrowed down to inspire your final piece. For while AI might give the appearance of intelligence, it can, like all computers, only follow instructions. And what real artists do to create real art consists of far more than following instructions.
Responding to Questions
Let’s explore some other questions that we’ve received:
“Why are you using DALL-E images when you should be using artwork by humans?”
Well, actually, the DALL-E images I used were created by a human: me. Sure, a machine executed my instructions to create the image, but is that so different from taking a photo with my iPhone? Both involve a human deciding WHAT AND HOW to compose something, and a machine doing the rendering work.
However, in all fairness to the spirit in which this person asked the question, I’ll address this assuming this person meant the following:
“Why didn’t you hire a human artist to create the illustration you envisioned?”
In our professional newsletters, we’ve always wanted something more unique than stock photography (although even stock photography sites are also now adding AI), something that represents what we are trying to present, something more unique than what you see on every blog across the internet.
We explored working with our graphic designer to create such images. But it is not practical, financially or time-wise. We typically are writing these newsletters a week or less in advance. There is, usually, simply no time to find a human artist, have a phone call, explain what we want, go back and forth several times all to illustrate an article that most of the artists reading are receiving for free.
Not only is time a limiting factor. So is money. How much would commissioning a human artist to illustrate one newsletter cost? Surely at minimum, a few hundred dollars per issue. And that is optimistic. That would be at least a thousand dollars a week, or $52,000 per year we’d have to pay to illustrate our newsletters. Now, if each of our readers would like to pay a subscription fee of tens of dollars a month to hundreds of dollars a month, then perhaps we could illustrate every article with a hand-created piece of art, custom made. But as long as we are providing newsletters at no cost (or included with other services), we will be stuck with affordable and fast ways to illustrate them, either stock photography, art we create ourselves, photos we take ourselves, or…..AI. (Unless anyone is interested in illustrating our newsletters from specific instructions, with less than a 24 hour turnaround, for a few dollars.)
But, joking aside, this example simply illustrates my point about AI EXPANDING the market. Previously, there really was zero market for unique art illustrations in newsletters (with the exception of a few bespoke and paid newsletters). But now, there is, and this trend will grow. Many Substack writers, both non-fiction and fiction that I read are increasingly illustrating their stories with AI. And by experimenting with these AI generated images, when some of these ideas will grow into full books, these AI illustrations will then be used as inspiration for a real, human artist to do more refined book illustrations. There will be millions of instances of AI images being used as inspiration for projects that grow up and need human artists. That would be a phenomenally good development.
It wasn’t a realistic choice to think it’s either “a human commissioned piece” or “stock photography.” So in this instance AI has not taken anyone’s former job, but it just might have provided a scenario that provides an additional job for a human artist in the future.
AI Image Quality Concerns
“These DALL-E illustrations are SO bad!!! The color makes my stomach turn and they are horribly rendered. I wish you would stop using them because it makes me want to delete the emails before reading them.”
We had several comments about how bad AI generated images are and how they lack story, depth and heart. It may be true they lack story, depth and heart (although I disagree when they are used as illustrations with an actual written story) but if that is what someone is looking for in a how-to-market article, may I ask - did you find story, depth and heart in our previous stock photography illustrations?
Regarding the quality of the images - making someone’s stomach turn. That is a fair criticism. We should do a better job of providing better quality images. I suspect, however, that is a failing of mine, and not the fact the images are AI, per se, because I’ve seen some very well done and compelling AI images by artists who are using it, and others who have more skill, patience and understanding about how to craft the prompts.
Running the services we run takes a lot of time and resources, and we are often pressed for time. So I admit, sometimes there is a rush to illustrate a given newsletter. But that is akin to an artist submitting the field sketch rather than using the sketch to refine the idea into a finished painting. I’ve seen AI generated images that are quite polished, and, in my opinion, not stomach turning and horribly rendered. And I know traditional painters who are embracing this technology and utilizing it to make their own digital art….and in doing so, providing themselves with a second medium and a second income stream.
As examples of better using of AI images than the ones I created, here are a few pieces that FASO customer Francien Kreig (an oil painter), created and sells online, I believe these were created with StableDiffusion:
Francien’s AI art is available as prints on her
FASO Website here, or as NFTs on Foundation here.
While certainly, an element of beauty is in the eye of the beholder, I don’t believe any of Francien’s AI images above are “soulless”, “heartless”, “storyless” or “stomach-churning-ly bad.”
Still,if those are valid criticisms, then they are criticisms of my poor mastery of the AI medium. And believe me, I’ve seen physical paintings done by people with poor mastery over the medium and I have seen some stomach-churning-ly-bad oil paintings in my day, a few even done by my own hand.
One can produce great art or poor art in any medium - it is the mastery of the medium that makes the difference, not the medium itself. So the proper criticism would be to say, “You are not skilled at using DALL-E to make appealing images.” And, as such, I will make every attempt to spend more time refining the images we use in the future.
Is AI an insult to artistic talent and abilities?
One reader asked the following:
“I enjoy your [FineArt]Views postings but I do wish you'd stop promoting AI in art such as DALL-E. Most artists consider it an insult to artistic talent and human abilities.”
I don’t know exactly how to answer this one. We aren’t exactly “promoting AI art”, we are simply using it, although I guess one could argue that using it equals promoting it. I’m not sure how anyone would know that “most artists” would feel this way. We have had some negative comments about our use of AI images, but it would be a tiny handful compared to the size of our subscriber base. Do those other artists not care? Are they supportive of it? Are they simply too busy to respond? I don’t know. And we’ve also had positive comments as well as pointed out above, traditional oil painters even moving to make AI art of their own.
So, I would ask an artist who feels it is an “insult to artistic talent and human abilities”, however, how do you feel about real, traditional artists (such as the one above) using this technology to explore new ideas and as a new source of income? Do you feel artists that are using it in this way are somehow betraying the artistic community? I’m serious about this question, I really want to know what people feel. Do you really feel this way? Or are some of us just having a negative knee-jerk reaction? Did artists feel this way about cameras? Did artists feel this way in the early days of digital art? Have artists considered, seriously considered, how they might use the medium creatively?
As an aside, I find two things interesting about Francien’s use of AI as an artist. First, and more mundanely, she’s embraced the technology to create another way to express herself and make additional income. Second, and more interesting to me, having followed Francien’s traditional paintings for years, I’m pleasantly surprised at how her AI images still “feel” like her style. I can’t put my finger on it exactly, but when I see one, I can instantly tell “Francien Krieg” made this. In other words, her human soul is reflecting through her art, even across different mediums.
FASO’s Responsibility to Explore New Tech
We Have a Responsibility to Our Customers to Learn and Use New Technology
More practically, as to why we use AI, we owe it to our customers to use and learn new technologies.
This has been a core FASO guiding principle from the very beginning. We watch for new developments in technology, and we adopt them early so that we may determine if any new technology will be useful for artists and if we can adapt them to help artists be inspired, save time, or sell more art.
I became a programmer after the 1st technology life changing experience I had in the tweet at the top, seeing a microcomputer for the first time. I realized the power of computers, and taught myself to program.
FASO itself was born after the 2nd technology life changing experience I mentioned in that Tweet - the birth of the Internet. It truly changed everything and FASO answered the question “how can I use my programming skills to allow artists to show and sell their art on the Internet?”
For example, we integrated blogs into every FASO site in 2006, and FASO artists were able to catch the wave of the golden age of blogging (approx 2007-2012), which had passed before most other services were able to finally catch up to us.
And, in 2009, after the 3rd life-changing experience I had with technology, the iPhone, we made a commitment to make every FASO site display well on smartphones. Our customers’ were ready, just as the iPhone was starting to gain massive adoption. Again, this proved to be useful for our customers as, even today, FASO websites look better on smartphones than any other service that I see artists hosting with. Good thing too since most people today browse on mobile devices and not desktop computers.
Which brings me to my point: Now that I’m seeing the 4th game-changing technology shift of my lifetime, I can’t shake the feeling that we have a responsibility to use AI technology if, for no other reason than to learn about it to determine if it might be useful for our customers!
Consider this: What if we were able to allow you to train an AI on all YOUR OWN ARTWORKS?
And then you could use that AI to create Print on Demand artworks that were based ONLY ON YOUR OWN IMAGES? No copyright issues. And you could put out new, saleable works at a pace you can only dream of at the moment. Or, if nothing else, create images you could use to inspire paintings? Would that be of interest? Wouldn’t it be great to not be “out of stock” of new artwork to sell?
Or how about this? Artists often tell us how much they hate writing their “About the Artist” page. Or their “Statement.” What if you could just type in a couple of sentences and then AI could complete the whole page for you? Would that be of interest?
What if AI could help you write better copy for your newsletters?
What if AI could be fed your artwork, and then create a description to help with SEO and human readers?
Shouldn’t we at least learn about these technologies to determine if they are viable, and, if so, use them to generate inspiration about how we might adapt them? And the best way to learn any technology is to use it for real world applications.
Our Duty to use Technology Responsibly
There very well may be more lawsuits coming in the future given the uncertain legalities around some of these technologies. It’s even possible there will be legislation regarding what is allowed and disallowed when it comes to AI technology. We are very keenly following all of this, and although we are experimenting and learning different aspects of this technology, we are also being careful not to fully integrate it yet, or get in a position where our customers rely upon it until we understand more fully the legal implications of doing so. I am hopeful that we will begin to see more clarity around this issue in 2023.
AI Makes Human Created Art MORE Valuable
While many seem concerned that AI will take over creative jobs, I don’t think that’s true in the case of truly unique creative endeavors, such as fine artists.
Sure, an AI might be able to replace a low level marketing copywriter or intern. But real human creativity and uniqueness…even weirdness….cannot be programmed.
And, in an algorithmically driven, AI generated, digital, instant world the opposite of machine-made, instantly available will become more valuable.
In other words, in a world of high volume machine created digital art, hand-made original art becomes more sought after and more valuable. Much as art prints lead to many collectors ultimately to purchase original art, or just as the Apple watch led to a resurgence in demand for fine swiss hand-made Rolexes, it's possible, likely even, that easy on-demand art images will spur interest in real, original art, made by real humans with real stories.
Remember that art marketing is about context and human created vs. machine created is a huge and important difference in context.
AI cannot Notice Beauty & Create Ideas, But Humans Can
Which brings me to the main point behind all of this: Real artists live inside the beauty of the universe more than the rest of us because they NOTICE beauty everywhere, in the mundane. In the little things. In abstract connections between seemingly disparate ideas. Making these connections, noticing beauty, following ideas - this is what makes the artist unique.
And while AI may get better and better and helping us explore those connections and ideas, I don't believe it will ever be able to independently notice beauty, have ideas, and make connections entirely on its own, for those are things granted to the mystery that is human consciousness, and how can we recreate consciousness in a machine when the best minds of humanity don't understand where it comes from even in us? It is the greatest mystery of science and we certainly won't accidently imbue machines with it.
So yes, I will, as a creator, as an artist, use all forms of non-conscious tooling that I have access to, and that are interesting to me, and that I have the skill to utilize, in the service of the most noble and holy thing a human can do which is this: Create.
I hope you will too.
Clint
PS - I am sure this article won’t be the final word on this topic, and I am sure the debate will continue for some time to come.
And, as always, we are truly interested in your point of view, your thoughts about what we’ve written here, or the topic of AI images/text as a whole. I would just ask, before providing feedback that you explore inside yourself what you really feel. Don’t just parrot what you’ve seen on social media or heard via hearsay or the media. Think it through from first principles and beliefs.
Are you using AI in some fashion, even if only for inspiration? Do you plan to experiment with AI as a medium? Have you tried any of the image generators or played with ChatGPT? Are you completely against the concept? If you’re against the concept, have you tried it? What do you think about visual artists who are already using AI as a creative outline and who are making additional income from this source? Will you use AI if the legal issues are clarified and resolved? Should we add an AI art category to the BoldBrush contest? Have you seen AI image gallery sites?
There are so many questions about this brand new technology and we are open to all points of view because ultimately, we are here to be part of this great journey with artists, which is, creation.
Please feel free to reply to this email, or to leave a comment at:
Clint, thanks so much for your thoughtful and informative exploration and modeling of this fascinating new technology. I for one cant wait to try it and to learn more.
Clint- I’m so grateful to you and your team for creating FASO and keeping ahead of all the new technologies. Yes keep looking ahead and incorporating the best into your platforms that are helping all artists to succeed if they use them. Greatly enjoyed this article on AI while using Deltas new free Wi-Fi above in the skies.