While learning about the implications of AI in the creative realm, I was particularly intrigued--and a bit worried. I've been doing photography since I was a sophomore in high school, and one of the biggest thing that drew me to it then (and still draws me to it now) is how much you can push the limits of creativity through a singular photograph. I especially enjoy photographing people; if you manage to align all of the variables in just the right way, you get to capture an image that invokes emotion: what I see as the ultimate goal of photography. So when I heard that AI was being used to do supposedly similar things that I do, I was immediately interested to see just how well these tools can execute an editorial photography concept.

The concept I chose to capture was the fine line between opulence and greed. I had two models pose as siblings caught in a money-and-envy-fueled battle, their uneasy relationship degrading even further with each photo at the hands of their own wealth. I wanted to channel an almost Succession-like sibling hatred, but with even more of an emphasis on money being the catalyst for their own downfall. In these photos, I used the color red to symbolize this greed and hatred, starting with splashes here and there and culminating in an inundation of scarlet.

Asking AI to create something like this was harder than I thought it would be. I used Wonder AI. Engineering the prompts to generate exactly what I was picturing took several rounds of trial and error, which proved difficult when it wasn't quite clear which parts of the prompts were causing which issues to arise in the image generation. Because AI is much like the black box of the human brain, we can't see what's going on inside of it. It's opaque in that way, and the only way to navigate it is to experiment, which takes some tweaking, but is sometimes worth it. Below is the series of AI photographs (with the prompt underneath) followed by my own photographs--which ones do you think execute the concept better?

"Siblings dressed in all black reaching for each other"

This was one of the earlier ones, before I realized "siblings" means children to Wonder

"Siblings dressed in all black sitting next to each other on a white couch with grim expressions"

This image surprised me as the man's attire seems rather unconventional

"Elegant woman dressed in black with pearls around her neck putting on red gloves angrily"

After the interesting fashion choices in the last image, I started to use the word "elegant" in the descriptor, which provided a more consistent wardrobe for my AI models

"Elegant man dressed in all black sparking a lighter close to his face and looking into the flame"

Words that can be taken symbollically (like "flame") are hyperbolized, and made to be the subject of the image

"Elegant woman holding a dollar bill dripping in red with red gloves on"

This was a deviation from the previous images, as it generated an abstract background

"Elegant man and elegant woman dressed in black half obscured by red curtains"

I tried to word this prompt several different ways, and for some reason, the AI could not obscure half of their bodies to achieve the image I achieved myself.

"$100 dripping with red on fire"

The flame is once again hyperbolized in almost a cartoonish way

"Adult brother and sister dressed in all red standing in extravagant hallway"

You can immediately tell that this one is AI-generated with one look at the woman's hands.

"An elegant man and an elegant woman dressed in black laying lifeless on an all red carpet"

This one made me laugh; it generated such uncanny valley-like celebrities that seem like you should be able to recognize them, but you can't. It seems that "lifeless" was lost in this prompt.

Just Out of Reach

Posed

A Taste of Power

Someone to Lean On

The Spark of Envy

Blood on your Hands

The Distance Between Us

Curtain Call

Bridges Burn

Bloody Together

My Thoughts

This project reassured me that I'm safe in my creative abilities for now (at least, when I'm up against the basic AI tools I had on hand). Even if the same prompt words are used each time, its consistency abilities are next to zero. And abstract concepts aren't something AI can handle either--the program needs tangible, descriptive words to generate something of substance. However, my experience is just the tip of the iceberg.

Opinions on AI art are mixed to say the least. Some, like an article by Forbes, are enamored by the quickly-developing accessibility of AI art tools. In the article, they pitch it as an exciting new opportunity into the art world for people who maybe aren't as creative and have always seen that as an obstacle to pursuing creative endeavors. It claims AI is "democratizing" creative enterprise, but is that really what we want? This was an interesting take to me, especially because I see art as an extension of talent of the individual. The unique personal edge is what I love most about art; a person's inside thoughts become outside expression in ways that emphasize their idiosyncracies, whether they intend to or not. And the different mediums different artists choose to express this through only enhances this uniqueness. For me, my passion for photography is what allows me to create my photos in the most "me" way possible. And while creative AI might be a start for people that haven't considered doing art before, I believe that when scaled up, it strips art of its true essence.

But the biggest problem with this technology of them all is what it does to artists. While some people argue AI isn't much different from new creative technologies of the past (like the camera, or like Photoshop), I disagree. These technologies still depend on the artist to generate the base content--even if they use editing technology, they still did the brunt of the work. AI is different in this way. But not only does it make art maybe a bit too easy, it also occasionally steals content from real artists. Midjourney, the AI tool used in a New York Times article about an AI art piece winning 1st place at an art fair, is actually trained on current working artists, meaning their art loses value as AI art based on their work gains value. This is unfair to the original artists, and as a digital artist says in the article, "This thing wants our jobs, it's actively anti-artist". For a long time, it was belived that creative jobs were some of the only positions that would be safe from AI taking them over. But according to digital artists like this, this isn't the case anymore.

So what's the conclusion on AI art? Public opinion is up in the air, according to another New York Times article about an art professor trying to get his students to embrace AI as a part of their artist toolkit. "This class is about daring students to embrace the machines," says the professor. And to an extent, he's right. We can't ignore the exponential growth of AI in all aspects of our daily lives, especially in the past few years. Ignorance is the worst position to choose in this case; it's almost a guarantee that you will become obsolete in the face of this new technology. It's necessary to embrace technological change as it comes, especially one as big as this. But as an artist, I sympathize with those being taken advantage of for this technology to be successful. However, as a student (and especially after taking this class), I can also respect the magnitude of this technology, and am excited to see what its future holds. In my eyes, the best solution here--the only solution here--is to be skeptically optimistic.

References

Wonder AI: https://wonderai.app/home

Forbes's "How AI is Changing the Future of Creative Enterprise": https://www.forbes.com/sites/forbestechcouncil/2023/03/27/how-ai-is-changing-the-future-of-creative-enterprise/?sh=210ccea55e61

New York Times's "An AI-Generated Picture Won an Art Prize. Artists Aren't Happy.": https://www.nytimes.com/2022/09/02/technology/ai-artificial-intelligence-artists.html

New York Times's "An Art Teacher Says AI is the Future. It's the Students that Need Convincing.": https://www.nytimes.com/2023/05/01/arts/design/ai-art-class.html