This Fruit Was Grown in Dan Saelinger’s Imagination and Created Using AI

Photography and Director Dan Saelinger is known for bringing bold, conceptual ideas to life with color, bold visuals, and precision. This summer, he revisited one of the many creative test ideas he'd jotted down over the years: isometric fruit. Inspired by the geometric perfection of Japanese cubed melons, Dan imagined a surreal series of animations in which everyday fruits were rendered in impossible shapes, and then deconstructed in unique ways. 

Instead of turning to CGI, Dan leaned into his generative AI experience, influencing the program by uploading his own work, so the generated images match his distinctive lights and photographic style. The resulting short-form animations blend conceptual art, technical skill, and storytelling. In a visually compelling and forward-thinking way, it's a glimpse into what's possible when you let your imagination take control. An exemplary showcase of how AI can open up new ways for artists to ideate and create and expand what is possible.  

What was the inspiration behind this project?

I am always looking for an opportunity to create some of the projects that live in my brain. This idea of isometric fruit came from the Japanese cubed melons, which are sold for hundreds of dollars. I thought the concept was really interesting and not dissimilar from my work in general. The tension between something organic and something manipulated was really interesting to me. I wanted to explore the idea further and create animations. 

Why did you end up using AI?

I knew that I wanted specific lighting, textures, and movement in these animations, but the logistics of creating these objects in real life and then manipulating them using CGI would be costly and time-consuming. I have been experimenting with AI tools and training programs to my photography style and taste. Over time, I’ve built a model that reflects the aesthetic decisions I make in my photography, so it made sense to use that toolset here. The goal wasn’t to simply make something using AI, but rather to use AI to solve a visual problem. How could I render a surreal, impossible fruit idea that still feels like my work? 

What was the process like to create these animations? 

This process, like any AI project, certainly was lengthy with many, many iterations. I generated thousands of potential images until landing on the ones I ultimately wanted to use. As I mentioned, I have trained the AI software I use to mimic my personal work’s lighting style, composition and color preferences. I then spent a lot of time refining prompts. You can’t just type in random words; you have to almost art-direct the software through language. Once I had a few stills that hit the right notes visually, I ran them through some light retouching to tweak the tones and backgrounds like I would a raw file from a camera. 

Then, I moved on to the animation portion. This was the most time-consuming part, especially the pineapple. Since there was no external object destroying the fruit, like in the peach and watermelon, the computer has a hard time understanding the physics of it. Even directing it to have the peach pop was challenging. These aren’t actions that clearly exist in its training data; it does not know how a peach deflates since that isn’t a real-life application. AI is not wired for surrealism, so I really had to keep pushing it. 

Why did you decide to go one step further and create animations?

Once I started working with the imagery and honing in on the concept of perfection and destruction, I knew it needed to move. These fruits aren’t real, they’re digital ideas, so I wanted to emphasize that by destroying the fruit. The animations show the objects breaking apart, collapsing, and unraveling. It’s a visual cue to the viewer: this isn’t nature, it’s a simulation. 

How did this project fit into your thoughts on AI?

In my opinion, AI is not “doing the work” for us. It needs us to input the proper prompts and really train it to get the outcome we desire. It’s no different than working with CGI or an editing program; it’s another tool. You still have to bring a concept, refine it, make decisions, and put in the time. 

Ultimately, I think AI is opening up new ways for artists to create things we couldn’t make otherwise. It’s not replacing creativity, but expanding what’s possible, especially for short-form conceptual work. I appreciate the ability to create new, interesting things, and AI allows the impossible to be created. That’s exciting.