I have been experimenting with generative AI in creative contexts from the very beginning. While the discussion about the benefits and dangers of generative AI in general is a complex one, I have personally approached it with hands-on curiosity and considered it as a new set of tools to bring my creative vision to life.

One example is the short film “Sol” I created in the beginning of 2024, which would not have been possible without these advancements in technology.

The idea for the film has been in my “sketch book” for many years, but was obviously never realized due to the considerable production costs the shoot would entail. Generative AI now made it possible to produce a short film version of the idea with very little expenses and as a one-person project.

While I wrote the whole script by myself, I did use ChatGPT to further refine the backstory (e.g. questions like “How does a modern woman with no outdoor experience survive on a deserted island?”).

All images were generated in Midjourney or Runway and most of them retouched using Photoshop including generative fill and generative expand. I then used Runway’s image-to-video (still in Gen-2 at that time) to bring the images to life. The resulting clips were further upscaled with the Borix FX AI plugin suite for AfterEffects (the whole “postproduction” was performed with Adobe CC and color correction in DaVinci Resolve).

In the beginning, I worked with a voiceover generated with Elevenlabs, but I realized very quickly that I wasn’t too satisfied with the result. I thus contacted and worked with the actress and voiceover artist Sydney LaFaire to replace the AI voice with a human one.

The final film was submitted to several short film festivals focused on artificial intelligence.