AI Autofill (Autopilot)
Dal Nulla’s AI Autofill (also referred to as Autopilot) is a powerful feature that allows you to rapidly prototype an entire video concept from a single text prompt. Instead of building your storyboard scene-by-scene, the AI does the initial heavy lifting, generating a structured project for you.
How it Works
When you provide a high-level idea to Autofill, the AI performs several steps:
- Idea Analysis: It interprets your main concept and identifies key narrative beats or visual elements.
- Scene Breakdown: It breaks down your idea into a series of distinct Scene Nodes for the video timeline.
- Context Generation: It often creates Global Text Nodes for overall project context (e.g., “Global Context”, “Characters”) and Reference Image Nodes to establish visual themes or characters.
- Automatic Connections: The AI intelligently connects these generated nodes in the Graph Editor:
- Text Nodes might be linked to Scene Nodes as prompt sources.
- Reference Image Nodes are often linked to Scene Nodes as visual references.
Using AI Autofill
From the Dashboard:
- Enter your video idea into the main prompt input field.
- Select your desired Aspect Ratio and Generation Mode (e.g., “Split” for a multi-scene storyboard).
- Click “Generate Video”.
From the Editor:
- Click the Autofill Storyboard (sparkles ✨ icon) in the floating sidebar menu on the left.
- Enter a prompt describing the storyboard you want.
- The AI will then generate and populate your canvas with new nodes.
What You Get
After running Autofill, your Graph Editor canvas will be transformed with:
- Multiple Scene Nodes: Each representing a distinct part of your story, often with initial prompts already filled.
- Reference Image Nodes: Visual elements (like character concepts or key objects) that serve as inputs for your scenes.
- Text Nodes: Providing narrative context, character descriptions, or dialogue for your scenes. These may include Global Text Nodes that influence the entire project.
- Interconnected Graph: The nodes will be pre-connected, forming a coherent visual flow that you can then further refine and customize.
This provides a strong starting point, allowing you to iterate on the AI-generated structure rather than starting from a blank canvas.