In this video, I test the new "Video to Video" feature from Runway Gen3 and ask a question that could shake up the film industry: Can we create a movie without actors?

My experience:
I put together a simple project: one actor, multiple roles, a white wall, and a few props, all transformed by Runway's AI. The results are impressive but come with some limitations. Here’s what I found:

→ The system generates visuals based primarily on shapes.
→ Source videos can include traditional video, 3D, augmented reality, or AIgenerated videos.
→ The white background stays static, 100% Matt painting.
→ Lip sync needs improvement, and changing skin colors naturally is tricky.
→ Facial expressions are impressive, but lighting and special effects sometimes lack realism.
→ Clothing transformations are limited (example: tshirt → shirt).
→ AI can generate sets and props, even in cardboard, to enhance films.
→ A simple smartphone video can achieve professional film quality with this tool!

⚠ Current limitations:

It’s difficult to create a complete film without recurring characters or stable locations.
No features like Cref or Sref yet to refine the creative process.
Despite these challenges, I’m convinced the future of filmmaking is headed toward more options and creativity with AI. The potential is huge!

Sound design: ElevenLabs.

What do you think? Are we on the verge of a cinematic revolution without actors? Let me know your thoughts in the comments!