Creating short-form video using artificial intelligence is a process that combines creativity with technology. This particular short was created as an experiment to show how a simple idea can evolve into a complete visual story – from still images to a dynamic video.
Where did the idea come from?
The concept was inspired by a classic nighttime scenario – stopping in an unexpected place and making a decision that may have consequences. The atmosphere was built around tension, ambiguity, and a subtle sense of unease.
A key element in achieving this look was the use of carefully crafted prompts that defined both the visual style and the camera perspective.
Image generation with Grok
The images were generated using Grok, which creates realistic visuals based on detailed textual descriptions.
Example prompt used for image generation:
Long distance candid shot, far distance zoom, peeper style, image taken from inside a car with window rolled down, voyeur style, a 30-year old sexy female Caucasian mom, pretty face, beautiful makeup, light burgundy lipstick, standing on low-class street, cityscape in backdrop, night-time, very dark environment, brunette hair, tanned skin, slightly wrinkled skin, tight top, belly is shown, sexy exposed thighs, showing a heavy amount of cleavage, large natural breast crease in between her chest, incredibly large natural breasts, natural breasts spilling out entirely, very huge natural macromastia, large natural buxom breasts, looking at the viewer and smiling with lips closed
Detailed prompts like this help the AI understand the scene composition, lighting, mood, and overall aesthetic.
Video generation with Grok
The next step was turning static images into short video clips. Grok was also used here to generate motion based on simple instructions.
The video prompt looked like this:
woman says: “[spoken line used in the video]”This added movement and a narrative element, giving each scene more depth and engagement.
Editing in DaVinci Resolve
The final step was assembling everything into a complete video.
DaVinci Resolve was used to:
- combine all generated clips into a sequence,
- adjust timing and pacing,
- add cuts and transitions,
- render the final video.
Summary
This workflow shows that AI video creation is not just a one-click process. It requires:
- well-designed prompts,
- the right tools,
- and final editing to bring everything together.
The end result is a collaboration between human creativity and artificial intelligence – where AI generates the assets, and the creator shapes the final story.