Is ChatGPT Ready for Video Editing?

Posted on by Larry

Recently, I asked readers of my newsletter to describe how they are using ChatGPT in their work. Their replies ranged from “Not using it at all.” to “Using it extensively.”

John Ryan sent a detailed answer that I want to share with you. I ESPECIALLY recommend his Reddit link at the end of his comment. As well, Daniel Larson explains how he uses ChatGPT for prep for interviews.

(Image courtesy Pavel Danilyuk)


I saw your call to email you if we’re using AI tools in our editing, and I gotta say, for me, I keep looking out for some AI tool that will take away some of my pain points in editing (anything that takes me too long, that clients don’t want to pay as much for!). I just haven’t tested that one yet.


I have found that in two cases, I am able to use Chat-GPT on the business side of my editing. That would be for things like, having read a book and made tons of highlights on a Kindle reader, I then send it to myself as a PDF. Once I break this down, it produces a nice outline for onboarding new clients. The problem is, it was 14 pages long! I had ChatGPT condense that into 1 page (would have taken me forever) which I added into my process.

In a similar vein, I recognized the need for creating a Client Prep Doc, full of humorous little anecdotes on client video mishaps that caused the project to go over-budget and just generally cause me to have to create some warnings for future clients. “Don’t send me new footage to use, when we’re on our refined or final cut!” and “too-many-cooks-in-the-kitchen” type disasters that had run me ragged. When I was finished typing that up, it was 10 pages, ChatGPT got it down to less than 1. I added back in some visuals and some color, but – it would have taken me hours and many drafts to condense that on my own.

It’s riveting to watch the process get spat out at the speed of a printer.


But inside of the edit? Other than some neural stuff that Resolve now says is running some new features or that my Mac Studio says is happening in the background, I don’t employ anything outside of the NLE.

Well… maybe that’s not entirely true. I did get a lot of mileage out of the Whisper Transcription app. Where with 2 or 3 clients now, I said “Hey, the interview was 42 pages. I’ve uploaded it to a Google Doc, let’s start by cutting there and then once we’ve got this thing down to a manageable size, I’ll put a timeline together.” Now Resolve has that feature built into the editor itself. (Right click on a clip and transcribe – then use it on the spot or export and work with it there.) However, when editing with it, it’s not smooth, it took much massaging to get the cuts to actually be where an editor would want them.

Still, I am eager to try these two out:

AudioShake: breaks full songs down into their stems. This would save me some hassle, where most songs I have downloaded from Epidemic, or Soundstripe, I did NOT get the stems at the time.

Stable Audio: Generative AI for music & sound FX – 4 out of 5 that I listened to sounded…ok. But the example I heard where they tried to emulate a voice was really terrible. Humans can easily tell, for example, a country song with an actual singer and lyrics that go together vs a country-sounding mish-mash.

(Image courtesy Kevin)


The other limitation that I am seeing, is simply “how many monthly subscriptions can us editors really afford?” We’re already paying for cloud storage, music and SFX libraries (3-4) stock photos (1-2-3 sites) stock videos (1-2 sites) add ons and plugins (MVFX for example at $100/each), billing services, calendar appt services, CRMs, web builders, landing page builders, canvas, youtube so we don’t have to watch ads while learning something,, writing programs, advertising, domain services, MASV for transfers, hosting, Vimeo…’s a LOT already!


Here’s a post in the editor forum on Reddit that I thought I would call to your attention. There are some really strong opinions for and against, and some interesting limitations for those in larger studio/production house jobs too.

Link: Redditors Comment on AI in Media Editing

– – –

Larry adds: John, these are excellent thoughts, thanks for sharing. The Reddit link is an excellent resource! I also agree that the proliferation of subscription services is not sustainable. I suspect a shakeout is coming. What I especially appreciate are your specific examples of where ChatGPT is useful and where it is not.


I’ve started to use ChatGPT when writing documentary interview questions and researching the general topic. I pose the question to ChatGPT and get a response that can tell me if the interview question needs work and may also reveal more questions to ask.

I’ve also been able to expand the topic in directions that previous research didn’t reveal. With ChatGPT I can do a pre-interview and be better prepared for the in person recording.

Larry adds: Daniel, this is an interesting way to use ChatGPT for research. Thanks for sharing your comment.

Bookmark the permalink.

One Response to Is ChatGPT Ready for Video Editing?

  1. Mike Janowski says:

    The only AI feature I use with any regularity is Premiere’s transcription feature. It works reasonably well (still not great at getting the language correct, but its close enough that we can figure it out). And as that feature is in its infancy, I can see how it will become more powerful.

Leave a Reply

Your email address will not be published. Required fields are marked *

Larry Recommends:

FCPX Complete

NEW & Updated!

Edit smarter with Larry’s latest training, all available in our store.

Access over 1,900 on-demand video editing courses. Become a member of our Video Training Library today!


Subscribe to Larry's FREE weekly newsletter and save 10%
on your first purchase.