In early 2022, I started noticing AI-generated images on social media, but I didn’t think much of it. That changed later in the same year when I saw the first AI-generated and AI-manipulated videos. As a filmmaker and visual artists, that’s when I became curious and took notice and began my own research and experiments to know more about this emerging technology.

At that time, AI-generated video was still in its infancy. AI video tools were complex and inaccessible. Stable Diffusion was one of the few available options, but it was far from perfect. Using it required trial and error, tinkering with parameters and command lines, all with scarce online resources. I had to push myself beyond my comfort zone to understand how it worked.

Months of experimenting followed. What fascinated me most were the imperfections—the glitches, inconsistencies, and randomness that felt visually raw and unexpected, like a new kind of digital aesthetic. Over time, I learned to control AI’s unpredictability and started achieving artistic results.

This led to “I&AI”, a short series where I combined AI with my expertise in filming, rotoscoping, visual effects, animation, sound design, and acting. AI was just another tool in the mix. I used it like a visual effect, distorting parts of the image to create something unique, such as turning a part of the video into an oil painting effect. The lack of consistency was frustrating for many, but I loved it.