Summary
Creator Dan Kieft presents a full course on AI video editing, covering six methods for speeding up post-production while leaving creative decisions to a human editor. The tutorial uses GPT Astra inside the ChatGPT desktop app or Codex as the core model, and covers trimming, multi-angle footage, motion graphics, VFX and sound design.
The toolkit includes the Higgsfield MCP and API for generations, the DaVinci Resolve MCP, and Premiere Pro or After Effects. Kieft explains why the desktop app matters, since the cloud version caps uploads at 512 MB, which is too small for typical camera files. He demonstrates a universal prompt that trims raw footage, transcribes it, removes dead space, adds editor notes as adjustment layers, marks spots for B-roll, and outputs a finished Premiere Pro or DaVinci timeline.
He is candid about limits: the agent takes over the PC, some trims took around 14 minutes, and the process still needs monitoring. The course also shows how to build a character sheet from reference photos using a skill file, then generate B-roll video of yourself with Higgsfield, and previews more advanced motion graphics workflows.
📺 Source: Dan Kieft · Published September 29, 2026
🏷️ Format: Course Lesson
