If your FYP looks like this, then you've probably thought about trying it out for yourself.

If your FYP looks like this, then you've probably thought about trying it out for yourself, but you weren't sure exactly how to do it or weren't sure that it would even work. Well, I'm gonna try it so you don't have to! And I'm gonna show everything. So you can learn from both my successes and mistakes. Welcome back to episode 5 of the YouTube automation series where I try and launch my own YouTube automation channel, from scratch, all on my own, with minimal financial investment to see if it really is possible. If you haven't seen the other episodes, then head over to my page to catch up, but today, we'll be talking about the exact process that I use to create these videos. It's a pretty simple 4 step process which includes preparation, image generation, video generation, and editing. But, let's start with preparation in step 1. So, once I've got my script, which I went over how to create in episode 2, the first thing I'll do is record the voice-over so that I have something to start working from. Then, I'll create a list of what shots I want to appear in the video, and once I have that, I'll open up a conversation with ChatGPT. Specifically this GPT is the one that I've been using, but you can use a different one if you've found something better that you like or just do this part entirely on your own. This GPT is a prompt generator, so we'll give it a description of the type of image we want for the shot, yes, image, not video. Don't worry, we'll get there in a second, then it fills in the blanks for us and gives us a much more polished and ready prompt to generate our image with. So, I'll do this as many times as I need to have a prompt ready description for every shot that I want to appear in the video. On to step 2, image generation. I then copy these prompts that we got from GPT, and take them over to Google Labs ImageFX, which at least right now, is totally free, and this'll spit out some options of images that I can download and use as the start frame of the video, when I take it over to our video generation AI. So again, I'll do this with every prompt that we got from step 1, until I have a starting image for every shot in the video. Now for step 3, video generation. The site that I use for this is called Higgsfield and there's a few different options for generating video here. I do end up using Kling 2.1 a lot more often because it costs significantly less credits, but, you can play around with these and see which one you prefer. Then, it's a process of generating the video with some simple instructions in the prompt, and repeating that for every image that we got from step 2. And finally, step 4, editing it all together. Once I have all my videos generated, I'll open my editing software of choice, which is Final Cut Pro, but you can use anything here like CapCut, which I'm pretty sure is free. I'll edit together my voice over and generated videos and add some sound effects, music, and potentially a few video effects, along with the subtitles before exporting it out and with that, we have a finished video ready to post. I don't claim to be a 'guru' and I'm not trying to sell you anything. I'm just a guy that wants to try out a side hustle, and this series will show everything that goes into it, so that if I find some success in doing this, then you'll know everything you need to know to try it out for yourself too. So if that sounds like something you'd like to see, then drop a like and comment for the algorithm, and press the follow button to follow along with the journey. In the next episode, we'll reflect on the first videos that we made on the channel to see what we could do better, AND, I'll give you all an update on how these videos performed. See you there!