Rather than prompting from scratch, I wanted to see what happened if I included my own work in the process.
Having developed a personal passion for both fine art and, more recently, generative image-making, I became curious whether I could find ways in which to combine the two. Rather than prompting from scratch, I wanted to see what would happen if I included a painting of my own in the process - using a skull I'd painted in oil as the foundation, and asking Midjourney to translate its texture, anatomy, and mood into real, tangible objects. Getting to these two final products took many failed attempts, prompt adjustments, and the occasional happy accident.
Why combine the two mediums?
The idea was that by grounding the AI in a painting I'd actually made, something of its character might carry over: the weight of the brushwork, the texture of the paint surface, the particular anatomy of the skull I'd spent time studying and rendering.
I chose this painting specifically because it felt like a strong visual foundation for product design. The skull form naturally suggests vessel-like objects, and the monochrome palette felt transferable to materials like chrome and ceramic. I wanted to see how much of what I'd put into the painting could survive being handed to an algorithm and translated into something entirely different — but still inspired by it.
I wasn't interested in generating something from nothing. I wanted to use AI as an additional tool — a way to take something I'd already made and explore what else it could become, while keeping my own work at the centre of it.
I had been actively learning Midjourney through taking courses, and this project became a way to test what I'd learned, and to push it somewhere more personal than a standard prompt.
