Search This Blog

Thursday, May 18, 2023

ChatGPT: Making YouTube videos

A while back ago I wrote a blogpost in which I laid out a structure for a five part series of YouTube videos that I choose to call PowerShell for Beginners. 

These videos have now been published and looking back at the experience I wanted to write a few lines about it.

First of all, the plan was to set a goal that I knew was possible to reach. One series containing five videos was reasonable and achievable. I also didn't want to let AI do all the creative work so to speak, I still wanted to record a voiceover, create slides and write the manuscript myself.

Instead I let ChatGPT create the topics for each video, with only the content keywords that I then could build manuscripts around.

The process of creating the first video was slow since I was unused to it and was figuring out the way. At the end when I had figured and polished my workflow it looked something like this:

1. Read the chapter guidelines to find out what it needs to encapsulate.

2. Divide the content into chapters while creating the manuscript, so that it is clear what slide contains what information. Make sure to include an intro and an outro text.

3. Create slides using PowerPoint, boiling down the text from the manuscript. Also a chance to make final corrections in the manuscript text as you step through it. Export each slide as a .PNG. At this stage I had created a simple PowerPoint theme to easily reuse the colors and fonts that I like.

4. Record voiceovers, one recording per slide, by reading from the manuscript and make sure it matches the slides. Make any final corrections if you notice any errors. By having one recording per slide you get the chance to rerecord any faulty files without having to redo everything.

5. Using OpenShot video editor I added the images and the sound recordings. Then after some editing I exported it as a video .

6. Upload to YouTube Studio and use the right settings for your content.

No comments:

Post a Comment