One of the great upsides of AI art is the ability for artistically unskilled people to create their own artwork. Instead of spending years learning the craft of how to paint or draw, you can instantly generate a skilled painting or drawing. The most artistic part of the process in AI art becomes choosing which image to create. For that image to be artistically meaningful, prompters must learn how to think like an artist. 

Writers and online content creators who previously would use stock art or public domain images can now create custom artwork for free with AI image generators. I have great respect for human artists who can create great artwork with their craft of drawing, painting, or digital design, but the average blogger (like myself), cannot afford to pay such artists to create a custom image for each post we write. Now with generative art we can create custom artwork instantly on demand. While it might not be as good as a professional human artist, it is better than the stock art and public domain images that bloggers would otherwise use.

I like to think that I have decent Photoshop skills and was able to create artistically original images for my blog posts pre-AI art. Some are certainly better than others. I spent much more time on my book covers, so I would hope they are more artistic and striking, but even they were limited to the public domain stock art I could find. I would search for images related to the story in some way. My options were limited, but I would try to do my best within those constraints. 

But now with generative AI, I have much more freedom. The possibilities are virtually limitless. Anything I can think of (just about) I can create an image of. So now the task at hand is to create the most artistically creative image for the project. To successfully use AI-generators, writers must learn to think like visual artists. 

So how do visual artists think? Specifically the type of graphic design artists who create book covers and movie posters. The artist must think of a single image that captures the essence of the textual content, be it fiction or nonfiction. You must think of a single image to represent the entire literary work. Thinking like an artist means considering what image would best convey visually what you spent so many words describing in the essay or book. 

For fiction and nonfiction it is slightly different. A fiction book cover depends on the genre, as different genres have different customs and tropes, but ultimately the cover should elicit a certain mood. A science fiction book cover will use different techniques than a horror book, a romance book, or a literary fiction book. Each genre uses different colors, typefaces, images, and styles. 

Sometimes the tone of the book can be captured entirely through fonts and colors. Within each genre, you have the freedom to create whatever image best resonates with your particular story. This might be a character (usually the protagonist), or it might be a graphical image of an object or scene, ideally in a visually striking way that captures the specific angle of your book. It might involve combining different images, which is also something I would do with stock art, but now I can customize further with AI to produce even more unique art. 

For nonfiction it’s less about conforming to genre expectations and more about trying to convey the central thesis of your essay/book. Take this blog post for instance, about thinking like an artist. After trial and error through dozens of prompts, I settled on the image above. It portrays an artist thinking about what to create, imagining a vivid scene in his mind before painting it on the canvas. Hopefully that single image successfully conveys what this blog post is about. Like fiction, a nonfiction image can represent a certain mood to let the reader know whether the following piece is humorous or serious, speculative and philosophical or dry and scientific. Different images capture different tones, and you can use AI generators to specify which tone you want to elicit.

With AI image generators, more people can create more art. But for that art to be any good, prompters need to think like artists. While it’s not ideal that more and more human art will be replaced with AI art, I’m less worried about stock art and public domain images being replaced by AI because those images would just be reused and replicated over and over again anyway. True original human artists are not being replaced by AI. They will still be hired by those who can afford to do so because great human artists can make art better than AI—for now and likely forever.

But for those who could not afford to hire human artists to create original art and were just using stock art, public domain images, or no images at all, I think it’s much better to have mediocre AI art that is at least original. Plus AI art has the chance to rise above mere mediocrity—if the prompter learns to think like an artist. Writers and content creators need to become more creative in the AI art they generate and develop their digital graphic design skills to further improve upon the AI image. If more content creators think like artists, the world will have much better AI art to look at, and the world is always better with better art.