Descriptions:
Matt Wolfe delivers a practical, tool-by-tool walkthrough of how to create AI-powered visual effects for video content, covering the exact workflow he uses for his own YouTube channel’s signature intro segments. The core technique involves exporting two still frames from DaVinci Resolve — one without the presenter and one with them seated — then feeding those keyframes into Runway using the Seed Dance 2.0 model to generate a dynamic transition. Wolfe notes that Seed Dance 2.0 currently outperforms alternatives like Cling and Google’s VO3.1 for this type of keyframe-to-keyframe interpolation, though he points out the model is available through multiple platforms including Crea, Higgsfield, and Leonardo.
Beyond intro effects, the video covers using Fable and Claude Code (now Claude Co-work) to programmatically generate B-roll animations: given a URL and a description of what to highlight, the agent screenshots the page, writes the necessary code, and exports an MP4 with scroll, zoom, and highlight animations — without the user touching a video editor. Wolfe compares outputs between Claude Code and Fable, finding Fable produces slightly smoother highlight animations. Additional segments cover generating stock-footage-style corporate B-roll on demand and creating product mockup animations.
The video is a useful reference for content creators and marketers who want to produce polished video effects without dedicated motion graphics skills, and it doubles as a current snapshot of which AI video generation models are leading for specific use cases as of mid-2026.
📺 Source: Matt Wolfe · Published July 08, 2026
🏷️ Format: Tutorial Demo
