Skip to content

AI Video Editing: How to Create Content Without Editing Skills?

12 min read
AI-видеомонтаж: как создать контент без навыков редактирования? — illustration 1

AI Video Editing: How to Create Content Without Editing Skills?

AI video editing is not just a trend; it’s a revolution that allows for scaling video content production to previously unimaginable volumes. If you find that video editing is a bottleneck preventing you from publishing more videos on YouTube and social media, this article is for you. Learn how to use Claude Code or ChatGPT Codex to edit videos, add motion graphics, and select music without learning complex software.

In this article, we will reveal how to set up an effective AI video editing workflow that transforms raw footage into a finished video with motion graphics, titles, and music. This will significantly increase your AI vs. real UGC efficiency and allow you to create hundreds of videos without spending weeks on editing.

Accelerating Video Production with AI: Justin Brown’s Experience

Justin Brown, who has been running his YouTube channel for 14 years, used to spend 14 to 22 hours creating a single video weekly. With the implementation of an AI video editing workflow, his team can now publish videos in less than an hour, and even faster when no additional polishing is required. Longer videos go live in about 40 minutes.

According to Justin, the biggest change is the lowered barrier to entry. Creators no longer need to learn editing software like Adobe Premiere, Final Cut Pro, or DaVinci Resolve, or understand sound levels and music composition. Instead, they can focus on prompting and the human side of video: storytelling, messaging, and communication.

Why AI Editing Works: From Simple Requests to Complex Instructions

Many mistakenly believe that AI is poor at video editing. Justin argues that this is due to attempts to get a finished edit with a single request like “edit my video,” hoping for the best. However, just like a human editor, AI needs clear instructions and guidance to succeed.

Clear instructions allow for faster good results, without spending tokens, time, and money on additional iterations.

Setting up the AI Video Editing Workflow

The foundation for effective AI video editing is a desktop application with access to files on your computer. This means downloading the Claude app using Code or the ChatGPT app with the Codex option, which now runs within the same application. These tools can manage a folder on your hard drive, so editing happens locally, not in the cloud, which uses fewer tokens.

  • Tool Selection: Justin gets excellent results from both ChatGPT and Claude. His current favorite is the latest ChatGPT model in Codex. He also notes Google Antigravity, though he considers it significantly behind.
  • Getting Started: You don’t need to be an AI expert. The best questions to start with, according to Justin, are “How can I do this?” and “What’s possible?”. Asking AI to “make a good video” won’t work, because “good video” means something different for every niche and every creator.
  • Training the AI: Justin starts by teaching the AI what a good video looks like in his eyes, based on the video’s goal: awareness, leads, subscribers, or sales. Each video is a test. If he’s unsure what exactly to test, he provides the AI with examples of outstanding videos from niche leaders who are already achieving that goal. The AI analyzes what works: pacing, number of titles, what’s happening on screen, use of sound effects and music. This analysis happens before the AI views his own material.

Optimizing Shooting for AI Editing: “Shoot for the Edit”

AI video editing begins at the shooting stage. Justin calls this “shooting for the edit”: thinking about the finished product before recording, so the material is easy for AI (or a human) to cut. The material doesn’t necessarily need to be shot sequentially if the prompt describes how the video should play out.

  • Continuous Recording: Justin records one continuous file, without stopping and starting the recording. This eliminates the risk of forgetting to hit record again and gives the AI one file to process. AI can handle large files, and since editing happens locally on the computer, nothing needs to be uploaded to the cloud.
  • Last Take by Default: Justin’s rule, for both human editors and AI, is to always use the last take. Moving on to the next section signals that the previous attempt was successful. If another take is better, he says it out loud during recording, for example: “Actually, use the first take.” Human editors often missed his claps and raised hands, but AI reliably follows these verbal instructions.
  • Pauses Between Statements: A short pause between statements gives the AI a chance to make a clean edit. Going straight from one thought to the next makes it difficult to time correctly.
  • Restarting Full Sentences: When a phrase doesn’t work, Justin restarts the entire sentence instead of picking it up halfway through. This allows him to give the AI a clear instruction: don’t cut sentences in half unless absolutely necessary.
  • B-roll and Visual Engagement

    A-roll is the main camera track: the primary shot of the person speaking. B-roll is anything layered over that main track, including graphics, animations, photos, and other footage. It shows what is being talked about, which makes the video more understandable and engaging.

    AI Video Editing: How to Create Content Without Editing Skills? — illustration 2
    • For short videos and reels, shooting B-roll that is planned to be layered over the A-roll gives the AI material to work with. Still photos also work, which can be given a slow Ken Burns-style zoom.
    • When Justin is unsure what B-roll a video needs, he uses AI as a source of ideas. AI can suggest what to shoot or create graphics itself for any moment where showing something will help the viewer understand.
    • Pro Tip: An AI editor can point out when a video lacks a strong “hook.” It can then suggest that the creator record several variations of “hooks,” labeled as take one, take two, and take three, and add the best one to the finished video.

    Organizing Files: The Path to Seamless AI Editing

    It’s important to have all files in one place so the AI knows where everything is. Keeping everything together prevents files from being scattered and ensures the AI doesn’t miss anything.

    • Justin recommends one folder for each video or video series, with subfolders for raw camera footage, B-roll (e.g., product shots or landscapes), music, photos, and graphics. A series can also have one folder for common assets, such as licensed music and the show’s intro.
    • Pro Tip: Before the AI starts editing, Justin suggests it first review everything in the folder and assess what’s there. Modern models often do this on their own, but he wants to be sure the AI has looked at every file.

    Accurate Transcription: The Foundation of Perfect Editing

    After the AI reviews the project folder, it transcribes the material. Justin calls this the most important step in the process. If the transcription is inaccurate, the editing and timing will be off.

    • For editing, timing in transcription is more important than the words themselves. AI needs to know exactly when each word starts and how long it lasts so it can make edits between words, rather than cutting off the beginning or end of one.
    • Justin argues that many editing tools, including AI features in DaVinci Resolve, struggle because their transcription timing is inaccurate. Accurate timing is also crucial for removing filler words like “um” and “uh” from a perfect take.
    • Long Videos: For short-form content, transcription built into ChatGPT or Claude is usually accurate enough. In a longer video, a one- or two-second timing error accumulates over time, so edits at the end of the video are severely misaligned, and audio loses lip-sync. Justin warns that AI can “gaslight” the user, claiming to have transcribed everything while subtly skipping sections.
    • Recommended Transcription Tools: Wispr, which AI can use behind the scenes, is one of the most reliable transcription models. Gemini API credits are another option. For long-form content, Justin has had the most success with Deepgram, a cloud-based service that charges a small fee.
    • Pro Tip: To test this workflow, perform a simple edit on existing footage and provide the AI with a transcription of that podcast episode as a guide. The AI will walk you step-by-step through setting up tools like Deepgram and connecting to Google.

    Two-Stage Editing: Story and Polish

    Justin edits long videos in two stages: a story pass that builds the A-roll, followed by a polish pass that adds captions, graphics, and music. This allows for AI efficiency versus real UGC and significantly speeds up the process.

    First Stage: A-Roll Assembly and Story Check

    For long-form content, Justin’s first editing stage builds the A-roll story, which he calls the most important part of any edit, whether done by a human or AI. The AI removes bad takes and mistakes, then assembles the core content before adding any captions, graphics, or other polish.

    A quick way to check the story is to ask the AI to create a timed script, generated from the transcription, that shows what it cut and what it kept.

    • The script lists each speaker, timestamp, and what is said. It costs almost nothing to generate, and reviewing it is a quick way to remove sections, bring in other people, or identify problems before adding any polish.
    • Shooting rules pay off here. Without instructions like “always use the last take” and “don’t cut my sentences in half unless absolutely necessary,” AI tends to stitch together individual words from different takes. The result is a jumpy edit almost every word.
  • When an edit goes wrong, Justin attributes it to one of two reasons: instructions that weren’t clear enough, or material that was shot in a way that made it difficult to make a decision. Each correction becomes part of a rule set that builds and improves over time.
  • Pro Tip: For short-form videos, he combines the A-roll pass and the B-roll pass (described below) into one step because there’s so little material that the story comes together right away. Removing flubs, ums, and ahs creates a clean video.
  • Stage Two: Polishing with Graphics, Music, and Effects

    B-roll, motion graphics, titles, sound effects, and music make a video engaging, and this is the polish a production company typically adds. Generating 100 videos with this level of quality is now possible with AI.

    • B-roll: AI can be trained to hit a specific B-roll ratio, such as 45% B-roll and 55% subject on camera, if it has a B-roll library to pull from. Without a set percentage, Justin tells it he wants the video to be engaging and clear wherever he’s explaining something. The inverse matters, too. When a video hits a key moment where the viewer should take action, it’s often best to cut back to the person on camera. AI knows best practices for storytelling and story arcs and can make that decision.
    • Motion Graphics and Titles: Claude now has a built-in feature that creates and animates graphics without needing anything else. Justin says a prompt such as “help me make this video engaging, add motion graphics, add titles, add captions” is enough to get started. Pointing to a brand book, PDF, or website allows graphics to automatically match existing brand fonts and colors.
    • HyperFrames: For more advanced work, Justin’s standout tool is HyperFrames, an open-source framework from HeyGen that works with various AI models. Codex or Claude Code can be instructed to use HyperFrames during editing to add motion graphics and give the video the look and feel of a high-end commercial. Justin doesn’t use HeyGen’s AI avatars. HyperFrames is strictly a tool to enhance real footage.
    • Placing Text Behind the Speaker: HyperFrames handles removing the background and placing titles behind the speaking person. Justin says text appearing slightly behind the speaker is a current trend that signals extra polish beyond a static title.
    • Generating Graphics from Transcription: In one video, Justin mentioned a Venn diagram. Without any instruction other than “using HyperFrames, make this engaging,” the AI created an animated Venn diagram that appeared behind him. This works because the AI has the transcription and knows everything that was said in the video.
  • Enhancing Tutorials with Screen Recording: For tutorials, HyperFrames can screen record websites and add scrolling footage as B-roll. When Justin appears picture-in-picture in the corner, it tracks the mouse and moves him if he blocks the action. It also automatically zooms in when text is too small or something on the screen is unclear to the viewer.
  • Highlighting Key Moments: The AI can slowly zoom in on the speaker when they make a key statement, similar to how a film director zooms in on a face during an important line. It can also create thumbnails.
  • Music and Sound Effects: Music and sound effects are key parts of an engaging video, but choosing the right track on stock music sites can take weeks. Leading music libraries like Artlist and Epidemic Sound now offer an MCP connection that links directly to the AI. Justin uses Epidemic Sound for his videos.
  • Music Timing: Once the AI has the transcription, it understands what the video is about and its mood. It then finds suitable music and sound effects and places them on the timeline without any manual work. If a track isn’t quite right, a prompt like “something more dynamic” results in a replacement that fits into the video. The music plays in the finished video, which makes a difference.
  • ChatCut: A Visual Interface for AI Video Editing

    ChatCut is a free tool for Mac and Windows that features a visual interface for AI video editing and viewing. It suits creators who have some editing experience and want to see a timeline of clips or edit video text like a Word document while the AI does the work. It integrates with Claude Code and Codex and connects to AI video generation models like Seedance to create new clips. Justin compares ChatCut to Adobe Premiere or Final Cut Pro, except the timeline is visible and doesn’t require touching.

    • For short videos

    ← Back to blog