AI and Art: Exploring the Creative Intersection

Once upon a time, the concept of “artificial intelligence” seemed a little far-fetched. Computers were mechanical, electronic devices, inanimate objects, machines that could only do what they were programmed to do, right? Computers couldn’t learn to think, could they?

Today’s computers not only think — or at least make it appear that they’re thinking — but they also write prose and poetry, offer critiques of our work, and create works of art on their own. How do they do it? I have no idea, really.

A lot of people have very negative feelings about artificial intelligence. AI “steals” ideas from others. They make use of other people’s art and literature. These are claims I’ve often heard. And, if true, then I would quickly agree that using “artificial intelligence” as part of a creative process would indeed be unethical.

Curiosity led me to an AI “chat-bot” where I posed this question directly:

It’s claimed that AI “steals” ideas and art from humans. Is this true? Are “AI-image generators” for example using work that’s created by real artists as the basis of their imagery? When you “compose” a poem or write an article are you “stealing” ideas from real people?

Here is the response I received:

This is a complex and controversial topic, and I’m glad you brought it up. AI, including language models like me and image generators, are trained on large datasets of human-generated content, such as text, images, and even code. This means that our responses and creations are influenced by the patterns and ideas we encountered during our training process.

While we don’t “steal” content in the sense that we directly copy-and-paste existing work, our output is based on the data we were trained on. So, in a way, we are indeed influenced by real people’s ideas and creations. This is why it’s so important for AI developers and users to be ethical and responsible when it comes to data collection and usage.

As for composing poems or writing articles, I draw upon the patterns and structures I learned during my training, but I generate new content based on those patterns. It’s a bit like how humans learn from the work of others to create their own unique pieces.

This particular bot does not create art, nor does it “view” art or critique it, although along with Theobot — discussed previously — there are many image generators and “critique” bots. Artificial intelligence is everywhere around us. It’s part of Facebook/Meta. Microsoft includes an AI “Co-Pilot” and Google offers its own AI-based services.

Word Press, too, features an AI “assistant” who will perform a number of tasks, from helping with writing, choosing appropriate titles, and generating a “feature image” based upon the content of a blog post. This image feature is the one I used to create this:

This was “generated” based on the text in my post titled “We Must Believe…” Whenever the WP “generator” creates a featured image, it provides a bit of text describing its creation. Here is the description AI provided to itself to convey the ideas expressed in the article I’d written:

Create a highly detailed, high-resolution image that visually represents the concept of “belief in success” in the context of learning and creating art. The main subject should be a thoughtful female artist at a canvas, surrounded by art supplies, appearing reflective yet determined. Soft, warm natural light should gently illuminate the scene, creating a serene, inspirational atmosphere. The artist should be captured in sharp focus, with a blurred background showcasing colorful artwork that symbolizes the growth and potential in artistic endeavors. Emphasize the blend of confidence and creativity in the image.

I did have to step in and make one slight change here. I had to add “female”. Seems like AI might be a bit gender-biased toward males since that was its original creation.

Of course, I’m making no claim to this image as “art”, nor am I claiming to have created it. I simply made use of a tool provided by Word Press. This is how I see “artificial intelligence” — a tool that has much to offer in terms of convenience. I see artificial intelligence as a great help for education and for research purposes, although there’s some disagreement there.

Just yesterday my husband and I watched a news segment about an honor student who received a “D” grade because he used artificial intelligence in writing a paper. Now, did he “use” artificial intelligence, or did he “copy” material written by artificial intelligence and claim it as his own work? These are two very different things. Using artificial intelligence as a research aid or even as a tutor to help us understand a subject better would be quite legitimate, in my opinion. Of course simply doing a “copy-and-paste” of material generated by an AI-bot would be a sort of “electronic plagiarism” — definitely wrong.

But this is an art blog, so let’s talk about artificial intelligence in that context. Can any AI-generated image ever be considered “art”. I don’t think so, really, although it would be tempting to say otherwise.

While playing around with Canva’s “AI-Image-Bot”, I came up with this moonlit landscape scene:

I would not, of course, claim this as my own “art” — although I might use it as inspiration for a landscape.

But let’s back up a moment here to look more at “digital art” as a whole — including software used to “manipulate” images. This was one of my first ventures into “art”. I’ve done a lot of photography over the years, and since I had no real “artistic talent” of my own, I enjoyed using early versions of photo-editing programs to create my own versions of “digital art”, such as this image — which I still use here on the blog:

Do I claim this work as my own? Yes, although with the caveat that it was created with a digital program. The original photograph was my own, one I’d taken during a visit to Powell Gardens. I then used the editing program to create the colorful “sun-filled” image shown.

For a long time, I frowned a bit at the whole idea of “digital art”, but that was before I really understood it. I wrote a blog post about it, in fact, and as I explored digital art more and more, I came to see that it truly can be called art.

Of course, using a software program still requries an artist’s input, an artist’s eye, an artist’s use of imagination and inspiration. These programs are “tools” — but really, aren’t all art supplies “tools” in a sense?

So what of art that’s wholly created — generated — from artificial intelligence? It’s not really devoid of input. It does require a prompt. An artist — or any user — must tell the program what to do, what colors to use, what mood to create. Those are all components of “art”, right? Many AI-art programs also include a variety of “styles” from which we can choose. Do we want line art, watercolor emulation, cartoon art… or what?

Just for fun, let’s see what we come up with. How about “ducks on the water”? No idea why I’m choosing this. It just popped into my head.

Here is an image created by Google’s “Imagen 3” using a “soft watercolor” style.

And Imagen 3 also created this “work of art” using a painterly style with textured brushstrokes:

Is either of these a “work of art”? No, of course not. Of course, both of these “paintings” are much better than any art I could create of this same subject!

It’s been said many times that “beauty is in the eye of the beholder”, and we can say that “art itself is often in the eye of the beholder.” Maybe there can be something beautiful in AI-created images. I know some individuals who work with image generators to create exquisite images, going back and refining and re-defining, making alterations until they have an image that’s exactly what they want.

I don’t have that sort of patience, nor do I have the technical knowledge or skills to do it, but in a way, that almost seems to approach some threshhold at which artifically-generated images could be likened to art.

All I do is play with AI-generated images, and I don’t even pretend at any level that I’m creating art. I’m having fun, creating images for my own amusement primarily. I think it’s entertaining to give an image generator a ridiculous-sounding prompt just to see what its “bot-brain” will come up with! But that’s me.

And these are simply my thoughts on artificial intelligence and art. Maybe your thoughts are different. And maybe the future will bring changes. Maybe artists will learn to work with artificial intelligence in much the same way they developed photo editing and manipulation tools and software programs designed to create real art.

I’m fascinated by artificial intelligence, and I see both good and bad within its capabilities, or perhaps I just see both good and bad in the way people might choose to use AI. But however we feel about it, one thing is certain. Artificial intelligence is here to stay, whether we like it or not. I’d love to hear your opinions on this controversial topic, especially as it applies to the realm of art.

NOTE: Yes, once again, I asked AI to generate its own image for this post. Here is the prompt it created:

Generate a highly detailed, high-resolution image that captures the intersection of art and artificial intelligence. The main subject should be a vibrant and imaginative representation of an artist’s studio, filled with digital art tools, glowing screens displaying AI-generated artworks, and traditional art supplies like paintbrushes and canvases. Natural light should filter through a window, casting warm tones across the scene, emphasizing the harmony between human creativity and AI technology. The style should blend realism with a touch of surrealism to reflect the complex relationship between artists and AI. Ensure a sharp focus on the details to create a captivating visual narrative.

I also allowed AI to choose the title for the post. To me, the use of AI is obvious in both the image and the title, yet at the same time both do serve the purpose for which they were intended.

How do you see AI? A useful tool? A certain step in the decline of civilization? A dangerous approach to education? And specifically, what are your thoughts about artificial intelligence as applied to the world of art?

Let’s talk about it!