Building AI video tools for creators. Cinematography + psychology for better AI videos. Shipping what I learn along the way.
Joined November 2023
- Tweets490
- Following62
- Followers25
- Likes35
Two weeks of GPT Image 2 + Seedance 2.0 demos hit Twitter.
Everyone showed what the workflow looks like. Almost nobody named why it works.
The storyboard isn't the workflow. The anchor frame is.
🤖 Made with AI
Same prompt, two passes.
GPT Image 2 generated the still. Seedance 2.0 brought it to life.
That's the workflow. That's all of it.
🤖 Made with AI
OpenAI launched GPT Image 2 two weeks ago.
API live on VicSee this week, paired with Seedance 2.0. storyboard to animation, both models native, no agent chain.
what surprised me wasn't the integration. it was how fast the workflow narrative locked in around this combo on twitter.
full breakdown, from reference portraits to the image legend naming system that actually compounds across shots: vicsee.com/blog/ai-character…
AI filmmaking tip most people skip: reference images solve character generation. that's the easy part.
the editorial layer is what actually holds sequences together. B-roll at transition points, cutting on action instead of mid-pose, reducing visual axes between shots. without that, consistent characters still feel like a slideshow.
🤖 Made with AI
"character sheets solve AI consistency"
wrong. sheets give you a face lock, not a performance lock. the model can reproduce the same face in one batch. maintaining that face across different lighting, expressions, and environments is where it breaks.
same prompt, two generations. two different people.
the #1 character consistency mistake in AI video: using your coolest scene as the reference image.
neutral lighting, front facing, no hard shadows. that's what scales. a dramatic scene carries its lighting and camera angle into every generation after it.
generate the reference first. test it. then build scenes from it.
Kling 3.0 tip that took a full production to figure out:
lock your shot vocabulary before you open the generator. wide to close, action to reaction, movement match. the sequence grammar matters more than what each individual prompt says.
same three actions in different shot order = completely different emotional register.
full breakdown with all five transition types:
vicsee.com/blog/kling-3-mult…
Wrong. “sadness” gets melodrama. “quiet resignation after a long day” gets something closer to human. Same model.
Getting the Ghibli look took 1 session. Keeping it consistent across 19 episodes took a system.
Most AI video creators solve the wrong problem.
Interior designers are using Nano Banana 2 as a floor plan reader, not an image generator.
Input a floor plan, get a rendered room. Iterate furniture placement without re-rendering from scratch. The spatial reasoning holds across edits.
