A collection of aesthetic art pieces I created with Midjourney, along with some reflections on the creative process and how I’m thinking about AI art so far.
Individual Project
Midjourney
2025
To this point, I have been familiarizing myself with the capabilities of Midjourney and getting the hang of how to produce high-quality AI-generated pieces. I started to think about how to move beyond pure aesthetics to give my creations deeper meanings and more cohesive styles.
A theme I landed on was childhood — things, places, and feelings that can only exist in the past.
I grew up in a country that’s changed way too fast, and now I live far from it. I decided to capture my nostalgia through art.
It forces me to do a deep emotional search for some feelings residing in the back of my mind. When I try to visualize the scenes, it’s almost impossible to get them to look exactly how I want — and honestly, a lot of the time I’m not even clear about how I want them to look. The final results feel like a mix of me and the AI meeting halfway. But I’m actually pretty happy with how they turned out.
As a designer and artist, I’ve spent years sharpening my craft, so I was hesitant to embrace AI art for a long time. Honestly, I’m late to the party. AI visuals have often been dismissed for looking unnatural, soulless, and riddled with quirks, which discouraged me from exploring the space—I assumed the limitations would outweigh the possibilities.
Recently, though, I came across a few online creators producing stunning, highly aesthetic art with AI. They shattered my preconceptions of what AI could achieve, inspiring me to give it a try myself.
In just a few days, I found myself hooked, using every spare moment to create something new.
This process feels entirely new to me. While the output is visual—something I’m deeply familiar with—it demands a very different skillset.
I’ve learned that crafting prompts requires precision and balance. The instructions need to be vague enough to allow creative freedom, but not so vague that the AI takes a completely unintended direction. There’s also a technical side: understanding how parameters influence the output and searching for reference codes that shortcut my way to the styles I envision.
It’s both exciting and challenging—a blend of creativity and problem-solving unlike anything I’ve done before.
Moderately.
AI makes it remarkably easy to generate plausible art, but creating something with soul and meaning is another matter entirely.
As a product designer, my work goes beyond making things visually pleasing. I aim to express ideas and emotions through art, and handing over full creative control to AI feels unfulfilling to me.
That said, if my role were more heavily reliant on visuals alone, I’d likely feel more pressure. Even then, I’ve found AI tools aren’t quite ready for prime time in a commercial setting. Fine-tuning the visual direction and details is far more painstaking than I anticipated. Often, I want to tweak an image in a specific way, only to find the tools can’t accommodate my needs—tasks I could easily handle if I created the art myself. Still, I’m sure these gaps will close as the technology evolves.
My relationship with AI art will evolve, just as the technology itself will.
I believe in the “know your enemy” philosophy—not that AI is an enemy, but whether I choose to embrace it or keep my distance, understanding it deeply is essential. Ultimately, it’s up to me to decide whether AI-generated art is something I want to claim as my own and be proud of.
*Drafted by me, polished with ChatGPT.
