This is "Eve & Adam"
Tips below on how I made this using Gemini, JSON prompts (and why), Runway Aleph, and Veo3.
Spoilers below. Watch before reading đ¤
Eve and Adam is a story about a brother and sister who are part human, part machineâthe first of what could be the next evolution of the human species.
It all started with an idea of someone shooting at the moon. I asked myself, why would anyone do that? Each answer led to another question, and the story of "Eve and Adam" began to unfold.
I wanted to capture the nostalgia of films like Close Encounters of the Third Kind and Escape to Witch Mountain, so I used Gemini to research the equipment those films were shot on and to help craft prompts that would maintain that classic look. I used Veo3 text-to-video and image-to-video for the visuals.
This project marked my first experience with JSON prompting. While I skipped the social media debate about whether JSON prompting gives better results âafter three years of crafting natural language prompts in paragraph form, the structure this format provided simply made sense to my brain.
First, I described the shot I wanted to create to Gemini, which then gave me a prompt in JSON format that I would feed into Veo3. I'd run an initial generation without even reading the prompt, just to see the output. Once I saw how far off the initial generation was from my vision, I'd then read the prompt and find the exact category I needed to tweakâbe it lighting, composition, or the subject itselfâthanks to the prompt's structured format. To me, this was far easier than revising a block of natural language text.
Another advantage: I could copy and paste the entire JSON prompt into Veo3 due to its large token limit, whereas other platforms would hit their character count maximum.
Example prompt:
{
"shot_name": "Aerial Shot (AS) - Car on Dirt Road into Forest",
"camera": {
"type": "Wide Shot",
"movement": "Aerial shot, car turns right off highway onto dirt road, following the car then craning up at the moon in the sky above the forest trees",
"lens": "Panavision lenses",
"focus": "Car and forest in focus, then car fading",
"lighting_direction": "Backlit by evening sky"
},
"setting": {
"environment": "Highway and a Narrow dirt road leading into dense, overgrown Pacific Northwest forest",
"time_of_day": "Night",
"atmosphere": "Shifting from isolated to mysterious and ominous"
},
"subject": {
"main_subject": "Late 1970s muscle car",
"details": "Kicking up significant dust, becoming a tiny, fading speck, tail lights receding"
},
"visual_style": {
"genre": "Action/Sci-Fi Drama",
"film_stock": "Kodak 35mm film (Eastman 100T 5247)",
"color_grade": "Gritty, natural with deepening purple shadows, cinematic",
"rating_tone": "PG-13",
"overall_feel": "Cinematic masterpiece"
},
"composition": {
"elements": "Car swallowed by immense, looming trees; road disappearing into wilderness"
},
"implied_elements": {
"sound": "Crunch of tires on dirt, rustling leaves, growing silence"
}
}
]I did this shot by shot, in a linear fashion. I added clips to my timeline in Premiere until the entire story was roughly laid out. On this first pass, the quality of the shots wasn't the priority; I just needed placeholders to begin understanding the film's flow, pacing, and tone.
Then I went back to the start of the timeline and regenerated and refined each shot. I did this hundreds of times.
I used Runway Aleph to fix inconsistencies in certain shots, as well as change lighting from day to night featuring the same characters. It came out right as I was nearing the end of making this, so I definitely need to spend more time with it, but this kind of natural language in-painting is amazing.
Veo3 provides an 8-second clip with sound effects, dialogue, and sometimes even baked-in music. For better sound design and mixing control, we need to export individual stems, so I used Lalal to split the audio.
Veo3âs audio often reflects the sceneâs toneâfor example, tense scenes get tense musicâbut it cuts off abruptly at the 8-second mark. One trick I used was to upload the 8 second clipâs audio to Suno and use it as inspiration for a longer sound design or music track. This gave me better better audio consistency when cutting between shots in a particular scene.
The main theme music is an original song I wrote years ago when I was in a band. I uploaded the recording to Suno and created variations in different styles that serve as Adam's theme song. I love how this tech isn't just helping me visualize new ideas; it's also breathing new life into old ones.
I don't think of this as a finished short film, but instead as a starting point. I can pitch this concept to a studio as a TV series or a feature and show this as a teaser. If it gets picked up, we can shoot it for real.
This is exactly why this technology needs to be taught in film schools today. Students can pre-visualize their scripts with AI, see whatâs working and what isnât, then shoot it with real actors on set for intimate scenes, while leveraging AI for any big-budget special effects shots.
If you're currently in film school and these tools aren't being taught, I have two words for you: Curious Refuge.
One question I know I'll be asked is, "How long did this take to make?" The misleading answer is three weeks. But the real answer, as anyone who has worked in production knows, is years. Years of film school and working as an editor on traditional productions cannot be discounted or ignored, because while generative AI excels at creating visuals, it's currently terrible at storytelling. It's not great at writing scripts and definitely can't edit well yet, so those aspects were done traditionally using the skills I learned from a career in post-production.
The good news is you don't have to spend years learning every aspect. There are plenty of knowledgeable people in this space looking to connect and collaborate. Team up.
Lastly, I don't believe AI cinema will replace traditional cinema; instead, they will co-exist as separate mediums or merge as an evolution of VFX. Weâre still going to need actors, writers, editors, and VFX artists, likely even more than before, given the increased demand for productions that these tools will facilitate.
Whether it's AI, CGI, live-action film, or theater, at its core, it's all just storytelling. But to me, having this level of visual effects available and affordable is a dream come true. So, experiment, ask questions, and most importantly, have fun! âď¸
Tools used:
Gemini, Veo3, Freepik, Runway, Magnific, Midjourney, ElevenLabs, Suno, Kling, Hailuo, Dreamina, Luma, Topaz
Link to 4K video on Youtube: https://t.co/zude1OuhuK
