Disclaimer: The original script and storyboard are not from me. They are belong to a production house and its talents crew. I created these to examine the possibilities of AI as an in-house creative and innovative R&D department lead. This was produced parallel with traditional production process and not affect the final result of the project.
I used to work on stop motion and film production projects before as a Writer, Director, Art Director, Animator and also Producer role. I constantly explore new tools and innovative ways of making films, AI is one of it.
By combine the knowledge of creative, production with an undeniable powerful tool, I believe we can do what we can’t in the past.
Treat this as a prototype for the final produced film. The footages I generated are up to 4K, but I only used 1080p for editting.
Like any films. It need to have a character design sheet, a storyboard, style frames. Key frames are very crucial for an AI film production. I spent time with Google Flow, Higgsfield using mostly Nano Banana model to create all of static pictures necessary, adding more insert scenes. Kling is awesome in keeping detail consistent and simulate stop motion’s authenticity well: stuttering motions, while Veo3 is make things smooth like a CGI Pixar’s film, but saving more credits and rendering time.
When I have the key frames. I animated them, by telling how all the elements in photo to ‘act’ as I want it. There is no shortcut. A lots of trial and fail, but once you know how it works, you will lessen the errors and having a better sense about the outputs (and saving credits too!).
Unfortunately, until now, a lots of people are still thinking that AI is cheap, everything created from AI are rubbish, trash, flop and people that exploring AI like me are *ssholes, betrayals, blah blah blah. But, well, you feel it cheap because someone not you putting up to $1 trillion into AI and make it accessible. Do you think AI is still cheap? And the skill of the one is controlling it was NOT.
I will go down this rabbit hole and see where it goes anyway.
