Big Idea

The argument in one line.

ChatGPT and Claude can now edit a full YouTube video end to end, cuts, layouts, zooms, and motion graphics, when paired with a one-time-purchase desktop editor that exposes an AI connection, cutting editing time from hours to minutes while still needing manual polish for the last fifth of the cut.

Who This Is For

Read if. Skip if.

READ IF YOU ARE…

You record talking-head YouTube videos solo and spend hours cutting dead air, false starts, and zooms by hand.
You already use the ChatGPT or Claude desktop apps and want them doing more than writing and research.
You're open to a one-time-purchase desktop editor instead of a subscription NLE, as long as AI can drive it.

SKIP IF…

You edit in Premiere, DaVinci, or Final Cut and aren't looking to switch editors just to get AI assistance.
Your videos are scripted, multi-camera productions where scene-based solo recording doesn't apply.

TL;DR

The full version, fast.

ChatGPT's Remotion plugin now generates rendered motion-graphic videos from a text prompt, and a $79 desktop editor called Borumi exposes an AI connection so ChatGPT or Claude can actually run the edit: removing false starts and dead air, choosing camera/screen layouts, adding zooms, and dropping in the motion graphics. The workflow is record in short scenes with spoken inline instructions, send one prompt describing the edit, then manually fix the roughly 20% the AI gets wrong, such as duplicate or mis-aimed zooms. Feeding those fixes back and asking the AI to update its reusable prompt is what makes each future edit land closer to the creator's taste.

Free for members

Chat with this breakdown — free.

Sign in and you get 23 free chat messages on us — ask for the hook, quote a framework, find the exact transcript moment, generate a markdown action plan. Bring your own key when you want unlimited.

Create a free account →

Chapters

Where the time goes.

00:00 – 01:09

01 · Intro

Cold open reveals the video was AI-edited, and states the two recent changes (better motion-graphics generation, easier AI-to-editor software) that made this possible.

01:09 – 02:42

02 · Remotion Plugin

Installs ChatGPT's Remotion plugin and generates a rendered five-second title card from a single prompt.

02:42 – 03:58

03 · Meet Borumi

Introduces Borumi, the $79 one-time-purchase desktop editor he's used for a year, now with an AI connection so ChatGPT or Claude can drive it.

03:58 – 07:33

04 · Borumi Workflow

Walks the Borumi app: scene-based recording, the editor timeline, layout choices, and speaking inline editing instructions ('Hey Borumi...') during the raw take.

07:33 – 08:56

05 · Connect ChatGPT

Points ChatGPT at the Borumi project folder, names the target video, and sends a saved 'skill' prompt describing the desired edits.

08:56 – 10:29

06 · First AI Edit

Reviews the finished AI edit: runtime cut from 2:45 to 1:26, false starts and dead air removed, layouts and zooms chosen automatically, motion graphic added.

10:29 – 12:09

07 · What It Misses

Audits the roughly 20% the AI gets wrong, duplicate zooms into the same spot, a zoom aimed slightly off-target, and how he manually fixes each.

12:09 – 12:32

08 · Editing Tips

Notes motion graphics are optional and slow the process down, and describes feeding manual fixes back to the AI to improve its saved prompt over time.

12:32 – 13:35

09 · Claude Alternative

Shows the same workflow running through Claude Code instead of ChatGPT, and gives his head-to-head preference for ChatGPT on this task.

Atomic Insights

Lines worth screenshotting.

ChatGPT can now drive a real desktop video editor end to end, cutting a raw talking-head recording from 2:45 down to 1:26 with one prompt.
The AI editing pipeline needs two separate tools: a motion-graphics generator (ChatGPT's Remotion plugin) and a video editor exposing an AI connection (Borumi).
A one-time $79 purchase, not a subscription, buys lifetime updates to the AI-drivable editor used in this workflow.
Recording a video in short, separate scenes means one bad take only costs you that scene, not a full re-record.
Spoken inline instructions during recording, tagged with a wake phrase, get stripped from the footage and executed later by the AI editor.
An AI edit currently lands about 80% finished; the remaining 20% is precision fixes like duplicate or mis-aimed zooms.
Feeding the AI a list of your manual post-edit fixes, and asking it to update its prompt, is how the automation improves per creator over time.
Skipping AI-generated motion graphics meaningfully speeds up an AI edit, since motion graphics are the slowest part of the pipeline.
The same AI-editor connection works with either ChatGPT or Claude by pointing the assistant at the identical project folder.
In a same-footage side-by-side test, ChatGPT outperformed Claude on video-editing decisions, though the creator expects that to keep shifting as models update.

Takeaway

How AI now edits your talking-head videos for you

AI VIDEO EDITING

ChatGPT and Claude can now run an actual desktop video editor end to end, cutting a solo creator's editing time from hours to minutes, but the automation currently gets you about 80% of a finished cut.

02Remotion Plugin

ChatGPT's Remotion plugin turns a plain-English request into a rendered motion-graphic video file, not just a static image or a code snippet.
A five-second animated title card took one prompt and about a minute to generate, with an editable side panel if you want to tweak it further.

03Meet Borumi

AI chat tools can't edit video directly; they need a purpose-built editor exposing an AI connection, which is the role Borumi plays here.
Borumi is a one-time $79 purchase with free future updates, not a subscription, which matters if you're trying to avoid recurring software costs.

04Borumi Workflow

Recording in short, separate scenes instead of one continuous take means a single flubbed section only requires re-recording that scene, not the whole video.
You can speak inline editing instructions during the raw recording itself, tagged with a wake phrase, and the AI strips those instructions out and follows them later.
A camera-only layout, a screen-only layout, and a picture-in-picture layout are all choices the AI can make on its own once it understands the footage.

05Connect ChatGPT

The AI editor needs to be pointed at a specific project folder on your computer before it can find and touch your video files.
A saved, reusable prompt, what the creator calls a 'skill', describing what edits to always make means you never retype the same instructions per video.

06First AI Edit

One prompt cut a 2 minute 45 second raw recording down to 1 minute 26 by removing false starts, repeated takes, and dead air automatically.
The AI also picked layouts and added zooms without being told which ones, and built a motion graphic from the inline instructions recorded earlier.

07What It Misses

About 80% of an AI-edited video comes out usable; the remaining 20% is precision work like fixing a zoom aimed at the wrong panel or one that fires twice in a row.
Redundant back-to-back zooms into the same spot read as jarring rather than intentional, so a human pass to consolidate or delete them is still worth the time.

08Editing Tips

Motion graphics add real time to an AI edit; skipping them, and building them separately in ChatGPT or Claude Design, makes the edit noticeably faster.
Telling the AI what you manually fixed after each edit, and asking it to update its skill or prompt, is how the automated cut gets closer to your taste over time.

09Claude Alternative

Claude Code can drive the same Borumi workspace instead of ChatGPT, using an identical folder-and-prompt setup.
In this creator's side-by-side test, ChatGPT outperformed Claude on the same edit, though he expects that gap to shift as models change.

Glossary

Terms worth knowing.

Remotion: A React-based video framework, and the name of a ChatGPT plugin, that turns a text prompt into a rendered animated video file rather than a static image.
MCP: Model Context Protocol, the connection standard that lets an AI assistant like ChatGPT or Claude control a separate desktop application instead of just chatting.
Borumi: A paid, one-time-purchase desktop video editor built for solo talking-head recording, with an AI connection that lets ChatGPT or Claude perform the actual edits.
Scene-based recording: Splitting a video script into separate chunks that are each recorded, and re-recorded, independently instead of filming the whole video in one continuous take.
Motion graphic: A short animated video clip, as opposed to a static image, used for titles, data visualizations, or transitions inside a larger video.

Resources

Things they pointed at.

01:09toolRemotion (ChatGPT plugin)

03:18toolBorumi ↗

08:23link"Edit Borumi Video" free prompt/skill ↗

12:37toolClaude Design

Quotables

Lines you could clip.

00:00

“Everything you're watching right now was edited by ChatGPT.”

cold-open hook that reframes the whole video as a live case study, no setup needed→ TikTok hook↗ Tweet quote

09:08

“It cut down the video from two minutes and forty five seconds down to a minute twenty six.”

concrete before/after number that proves the claim→ IG reel cold open↗ Tweet quote

10:35

“ChatGPT can successfully edit about 80% of the video, but I still have to go in and fine tune it and finish that last 20%.”

honest limitation that keeps the whole pitch credible→ newsletter pull-quote↗ Tweet quote

The Script

Word for word.

Read-along

Don't just watch it. Burn it in.

See every word as it's spoken — crank it to 2× and still catch all of it. The same dual-channel trick behind Amazon's Kindle + Audible.

metaphor

Everything you're watching right now was edited by ChatGPT. The video of me is real and I made all these screen recordings myself, but ChatGPT chose what cuts to make, all the layout choices, and when to zoom in.

It even created all of these motion graphics. A month ago, this was impossible. But thanks to some new software and better models, you can now have ChatGPT or Claude edit an entire video just like this one.

And I'm not exaggerating when I say that this has taken my editing time from hours down to minutes. Up until recently, AI was really bad at editing videos. I tried pretty much every AI video editing tool out there, and none of them actually saved me any time.

But recently, that all changed for two reasons. First, because ChatGPT and Claude Design have gotten really good at motion graphics. Those are these animations that you may have seen in some of my videos.

And the second reason is because we finally have easy software that ChatGPT and Claude can plug into to actually do the video editing for us. I wanna start by showing you how you can enable ChatGPT to create motion graphics, and then I'll show you the new editing software that ChatGPT and Claude can control. To get started with this, come into the ChatGPT desktop app and click on the left here where it says plug ins.

The plug in that we're looking for is called Remotion. This is it right here. I already have it installed, but if you don't have it installed yet, there'll be a button here to install it.

And that's it. It's that easy to set up. Now when you come over to a new chat, you can type in the at symbol, type in Remotion, and ask ChatGPT to create a motion graphic for you.

So for instance, here, I'll ask it to create a simple motion graphic that I can use as a title card for a video about Remotion. Alright. So ChatGPT is now finished, and it put the video for us right here in the chat so we can watch it here.

There is the title card, and that's perfect. That is exactly what I wanted. So now what I can do is click right here where it says download the m p four if I wanna use that video.

If you're curious, though, you can click on the top right here where it says toggle side panel. And this is the Remotion UI. So if you ask ChatGPT to make any changes to this motion graphic, it will do it in here.

And then you can always click right here where it says render to actually export the video, or it may just send you it in the chat like it did for us here. There's a whole lot you can do with these motion graphics. They're really good for visualizing data, for instance, but we're not gonna spend a lot of there or really any more time talking about how we can prompt ChatGPT to create motion graphics with Remotion because we're just going to leave that at the discretion of ChatGPT while it's editing the video.

So let's move over now, and I'll show you the editing software that I've been using with ChatGPT and Claude. So ChatGPT cannot edit videos directly or at least not very well.

You really need a third party tool that is built for video editing that ChatGPT can then control. Like I've said, I've tried a lot of them, and the one that I found that's actually good is called Burumi. I've actually been using Burumi for almost a year now, well before they added any AI capabilities because it's just really good for creating these talking head style YouTube videos.

But recently, they added MCP support so Claude and ChatGPT can edit the videos for you inside of Broomy. Now I wanna be upfront that Broomy is not free to use. However, I think it's pretty reasonable.

It's just a one time payment of $79, and then you get full future updates included with your purchase. So you never have any sort of upcharge for it.

It looks like that price may go up eventually, but at least as of right now, it's only $79. By the way, Broomy is not sponsoring this video. This is just genuinely the actual tool I've used to record and edit all my YouTube videos on my channel for the past year.

But before deciding if you wanna purchase it yourself, let me show you what this can do and how you can use it with ChatGPT or Claude. So this is the Burumi desktop app, and when you open it up, you can create these different video projects for each video that you're working on. Here's an example of one I started earlier today where I was making a video all about how to use the AI capabilities inside of Google Drive.

There are three tabs down here at the bottom. The first one is where you put in your script. And what I love about Broomi is that they have scene based recording where you can break up your video into different chunks so you're not forced to record the entire thing at once.

And then this tab is where you actually record it. So it records your camera, your microphone, and your screen. And then over here on the right, these are each of the scenes that we just put together inside the script section.

So that way, if you mess up one of the scenes, you don't have to start the entire video over again. You can just start over on that one scene. It makes it a lot easier to get through the content.

Now once you've recorded everything, you can come over to the editor, and this is gonna look very similar to a lot of other video editors. We're down here.

We have the timeline with each of the different scenes labeled. You could see the bottom row down here.

This is my talking head, my face, and the top row is the screen. And they are kind of linked together so that you don't lose your place and they're matched up. And within this editor, there's a lot we can do.

So let's start by playing the video and taking a look at it. Today today, I today, I wanna show you how to use all the AI features inside.

So notice how in the beginning, there's a lot of mistakes that I made. It took me a couple of tries to actually get started. So, normally, what I could do is just drag over the start so that it gets rid of all those bad takes.

Very easy to do in Broomy. But as you see right here, there's a lot of blank spaces like here. There's another false start here.

There's a couple of false starts here, more blank spaces here. So even though it's easy to do it for one, doing it for the entire video, especially for longer videos, can get quite tedious. So we're gonna have ChatGPT do that for us.

Another thing you can do inside of Broomy is set up these layouts. So when I click on layout, I can choose to just show my face. I can choose to show myself in the corner as a round circle in the corner.

I can do just the screen. So there's all these different layout choices. Again, pretty easy, but this is also something that ChatGPT can choose for us.

And there's also one section right here where I talk about AI adoption rates. And this video couldn't have come at a better time because right now, the rate of adoption for the AI features inside of Drive is just taking off.

So I'm hoping that that will be a place where ChatGPT will add a motion graphic showing the adoption over time.

And in fact, what I've done is right here, I've actually given instructions to ChatGPT. We can listen to that. Hey, Burumi.

When I'm talking about the AI adoption, create an infographic for that. It doesn't need specific numbers, but just something showing adoption over time going up, like a bar graph or a line graph.

So what I've done is I've recorded the video. And then while I'm recording the video, I'm actually giving instructions to ChatGPT on how I want the video edited.

That way, I don't have to type it out. And ChatGPT knows that when I say, hey, Burumi, it's going to remove those instructions and follow those instructions. So let's stop playing around inside the editor now, and let's go over to ChatGPT and have it edit the video.

Back inside the ChatGPT app, you wanna make sure you have either ChatGPT work selected or you are inside of codex. Then click right here where it says choose project and new project, and you wanna select a very specific folder on your computer.

So click here where it says add folders, then click on your username on the left, open up documents, and click on the folder that says Broomi AI workspace, and then click on open.

As long as you have that folder selected, then ChatGPT will be able to edit any of your videos.

You then wanna give ChatGPT the name of the video. The video I want you to edit is called how to use Google Drive, and then you'll just describe what edits and changes you wanted to make. What I've done is I've created a skill called edit Burumi video.

And in the description down below, I'll a link where you can grab that prompt free, and you could just copy it and throw it into the chat. That will tell it to remove all the pauses, all the false starts, and to follow your instructions anytime you say, hey, Burumi, as well as a few other things.

So with all that in place, I can now send it along. And now we'll just sit back and let it work. This can take a little bit of time.

So usually what I do is I go grab some lunch or do some other tasks. And then when I come back, the video is edited. Alright.

I finished my lunch and ChatGPT finished editing the video, and we can see right here what it did. It cut down the video from two minutes and forty five seconds down to a minute twenty six, and it did that by removing all the false starts, the repeated takes, my instructions, and all the dead air.

It also created a motion graphic. It struggled a little bit here. It originally tried to create an image instead of the motion graphic, and I told it, no.

I wanted to create a motion graphic instead. I'll make sure to update that for the prompt that I give all you so that shouldn't happen again. And then it made a few other edits, changing the layouts, and adding all the zooms.

So let's go over now to Broomy and see what it did. So as you see here, this looks much different now. We have all this formatting.

All the dead space has been removed. They added a bunch of zooms. So let's give this a quick watch, at least part of it, and see what ChatGPT did.

Today, I wanna show you how to use all the AI features inside of Google Drive. And this video couldn't have come at a better time because right now, the rate of adoption for the AI features inside of Drive is just taking off.

So if you don't understand how to use these, you are going to get left behind. Now inside of each of the different tools like Docs, Sheets, Slides, and Vids, there are different capabilities. So we're gonna dive into each of these.

Yep. So not bad. That's exactly the way that I would have edited the video, and all of that was just done with one prompt inside of ChatGPT.

Now I've been using this for a few weeks now, and what I found is that ChatGPT can successfully edit about 80% of the video, but I still have to go in and fine tune it and finish that last 20%.

For instance, right here, I'm showing Google Docs and ChatGPT correctly zoomed in on the Gemini bar on the bottom, but then it zoomed out only to zoom right back into the same spot, which is a little bit nauseating to go in and out like that.

So I would fix that. I would get rid of the second zoom and just make the first one longer so it stays zoomed zoomed in on that one spot.

There's another spot here where it zooms in on the document where it really doesn't need to, so I would delete that zoom. But then it correctly changes the layout to just my face and then switches back to a split view of the screen and my face and zooms in on the right section.

So that all looks great. But right here, it's supposed to zoom in on the panel on the right, but you can't actually see the right thing. So it zoomed in correctly at the right time, but the zoom is a little off.

So I can just use Broomy to fix that, and then it is perfect. So, yeah, just like those little minor adjustments, every time I use it, it gets better and better though.

So what you can do is after you finish the final polish of your video, you can come into ChatGPT and just tell it, I finished making the edits to the video. Look at what I did and update your skill or prompt so next time you get me closer to the final version. I do this every time and the skill gets better and gets me closer to that end result every time that I use it.

Couple other tips for you. The motion graphics are definitely optional, and they will definitely add a lot more time to the editing process.

Without those, it can edit the videos much faster. So what you could do is just ask it to edit the video for you and then create the motion graphics using either ChatGPT or Claude design and just drop them in yourself. Speaking of Claude, you can have the Claude desktop app edit the videos for you instead of ChatGPT, and it works the same way.

You just have to switch over to Claude code, and then make sure you select that same folder, the broomy.ai workspace, and then ask it to edit the videos for you. I personally prefer to use ChatGPT for this.

I've experimented with the same video, giving both of them to ChatGPT and one version of it to Claude, and I found that ChatGPT did a much better job. Of course, that may change as new models are released.

So, yeah, that's how I edit my videos or how I don't edit my videos now using ChatGPT and Burumi. Now we talked a bit in this video about motion graphics.

I have a whole video on that right here. I show you how you can create motion graphics with Claw Design, and I still think that Claw Design is a little bit easier to create motion graphics than ChatGPT, and usually the results I get out of it are better.

So if you wanna see all the cool things I've been making with motion graphics and cloud design, check out this video right here, and I'll see you over there in one second. Bye for now.

The Hook

The bait, then the rug-pull.

The video opens by undercutting its own credibility: everything on screen, the cuts, the zooms, the motion graphics, was assembled by ChatGPT working through a paid third-party editor called Borumi, not by the creator's hand. What follows is a full accounting of what that setup can and can't yet do for a solo talking-head channel.

Frameworks

Named ideas worth stealing.

12:24model

Record, AI Edit, Manual Polish, Feedback Loop

Record in scenes with spoken inline editing instructions
Prompt the AI editor to cut, layout, zoom, and animate
Manually fix the roughly 20% it gets wrong
Tell the AI what you fixed and ask it to update its skill/prompt

The four-step loop the creator repeats on every video, where each round of manual fixes is fed back in so the next AI edit lands closer to final without retyping instructions.

Steal forany AI-assisted creative workflow you want to keep improving without rewriting the prompt from scratch each time

CTA Breakdown